The Oxford Handbook of Experimental Semantics and Pragmatics 9780198791768, 0198791763
This handbook is the first to explore the growing field of experimental semantics and pragmatics. In the past 20 years,
135 5 9MB
English Pages [687] Year 2019
COVER
THE OXFORD HANDBOOK OF EXPERIMENTAL SEMANTICS AND PRAGMATICS
COPYRIGHT
CONTENTS
ACKNOWLEDGEMENTS
LIST OF FIGURES AND TABLES
Figures
Tables
LIST OF ABBREVIATIONS
THE CONTRIBUTORS
CHAPTER 1: INTRODUCTION
CHAPTER 2: LANGUAGES COMPREHENSION, INFERENCE, AND ALTERNATIVES
2.1 THE STUDY OF INFERENCE IN LANGUAGE COMPREHENSION
2.2 DEVELOPMENTAL EVIDENCE: SUCCECCES AND 'FAILURES'
2.3 SCALAR IMPLICATURE
2.4 QUANTIFIER SPREADING
2.5 AN ALTERNATIVE
2.6 CONCLUDING REMARKS
CHAPTER 3: CONSTRAINT-BASED PRAGMATIC PROCESSING
3.1 THE PROBLEM
3.2 CONSTRAINT-BASED APPROACHES TO LANGUAGE
3.3 LEADING EXAMPLE: SCALAR IMPLICATURE, THE DROSOPHILA OF EXPERIMENTAL PRAGMATICS
3.4 CONSTRAINT-BASED APPROACHES TO PRAGMATICS
3.5 CONSTRAINTS INVOLVED IN PRAGMATIC PROCESSING
3.5.1 Question Under Discussion
3.5.2 World knowledge
3.5.3 Properties of the observed utterance and its alternatives
3.5.4 Properties of the speaker
3.5.5 Common ground
3.6 CONSTRAINT-BASED APPROACHES TO SCALAR IMPLICATURE PROCESSING
3.6.1 What’s wrong with informational privilege accounts?
3.6.2 Beyond informational privilege accounts
3.7 CONSTRAINT-BASED APPROACHES IN OTHER AREAS OF PRAGMATICS
3.7.1 Perspective-taking
3.7.2 Lexical precedents
3.7.3 Learning: Adaptation, generalization, and speaker-specificity
3.8 INFORMATION INTERGATION
3.9 QUO VADIS, CONSTRAINT-BASED PRAGMATICS?
3.10 CONCLUSION
CHAPTER 4: SCALAR IMPLICATURES
4.1 INTRODUCTION
4.1.1 Overview
4.1.2 Pragmatics, communication, and a Gricean system
4.2 SCALAR IMPLICATURE
4.2.1 Domain of inquiry
4.2.2 Scalars as Gricean phenomena?
4.3 EXPERIMENTAL RESEARCH AT THE LANGUAGE-GRICEAN INTERFACE
4.3.1 Default implicatures: A testable proposal about the language-pragmatics interface
4.3.2 Counting the cost of scalar implicatures
4.3.3 Experimental research on potential scalars
4.4 HOW A GRICEAN SYSTEM MIGHT INTEGRATE LINGUISTIC AND NON-LINGUISTIC FUNCTIONS INTO UTTERANCE INTERPRETATION
4.4.1 Literal first or free-for-all?
4.4.1.1 The integration of a Gricean System in online utterance processing
4.4.1.2 Literal first in scalar processing?
4.4.1.3 ERP studies on the time course of scalar implicature
4.4.2 Beyond time course studies
4.5 CONCLUSION—BACK TO THE ORIGINS OF A GRICEAN SYSTEM
ACKNOWLEDGEMENTS
CHAPTER 5: EVENT (DE)COMPOSITION
5.1 INTRODUCTION
5.2 DECOMPOSITION AND EVENT STRUCTURE
5.3 THE MANNER AND RESULT OF EVENTS IN THE ENCODING OF VERBAL ROOTS
5.4 THE TEMPORAL BOUNDARIES OF EVENTS AND THE ENCODING OF VERB PHRASES
5.5 PARTICIPANTS IN EVENTS AND THE (IMPLICIT) ENCODING OF ARGUMENTS
5.6 SUMMARY
ACKNOWLEDGEMENTS
CHAPTER 6: PRESUPPOSITIONS, PROJECTION, AND ACCOMMODATION
6.1 CURRENT ISSUES IN PRESUPPOSITION THEORY
6.1.1 Basic phenomena
6.1.2 Semantics vs. pragmatics and explanatory challenges
6.1.2.1 The triggering problem
6.1.2.2 The projection problem
6.1.3 Distinguishing types of presupposition triggers
6.2 PRESUPPOSITION INTERPRETATION IN EXPERIMENTAL TASKS
6.2.1 Detecting presuppositions experimentally
6.2.2 Reflexes of the status of presuppositions
6.2.3 ‘Novel’ presuppositions: To ignore, accommodate, or cancel?
6.2.4 The time-course of presupposition interpretation
6.2.5 Interim summary
6.3 THE INTERPRETATION OF TRIGGERS IN EMBEDDED ENVIRONMENTS
6.3.1 Identifying what projects
6.3.1.1 Projection from the scope of quantifiers
6.3.1.2 Projection from sentential operators
6.3.2 Local readings
6.3.3 Differences in projection between triggers
6.3.3.1 Soft vs. hard triggers and local accommodation
6.3.3.2 Local readings reconsidered: Entailment vs. local accommodation
6.3.4 Summary on embedded triggers
6.4 PRESUPPOSITIONS IN DISCOURSE
6.4.1 Resolving presuppositions in the discourse context
6.4.2 Factives, prosody, and discourse
6.4.3 Obligatory triggers
6.5 CONCLUSION AND OUTLOOK
ACKNOWLEDGEMENTS
CHAPTER 7: SPATIAL TERMS
7.1 INTRODUCTION
7.2 LOCATION TERMS
7.3 MOTION TERMS
7.4 FRAMES OF REFERENCE TERMS
7.5 DOES SPATIAL LANGUAGE AFFECT SPATIAL COGNITION?
7.6 CONCLUDING REMARKS
ACKNOWLEDGEMENT
CHAPTER 8: COUNTERFACTUALS
8.1 HOW ARE COUNTERFACTUAL SITUATIONS ESTABLISHED?
8.2 COUNTERFACTUAL REASONING PATTERNS
8.3 HOW MIGHT COUNTERFACTUALS BE REPRESENTED DURING COMPREHENSION?
8.3.1 Mental spaces
8.3.2 Situation models
8.3.3 Mental models
8.3.4 Suppositional theory
8.4 EMPIRICAL EVIDENCE FOR MULTIPLE REPRESENTATIONS OF COUNTERFACTUALS
8.4.1 Evidence for rapid access to the explicit counterfactual world
8.4.2 Evidence for rapid access to the implied factual world
8.4.3 Both worlds are represented simultaneously
8.5 COUNTERFACTUAL THINKING AND SOCIAL COGNITION
8.6 SUMMARY
ACKNOWLEDGEMENT
CHAPTER 9: DISTRIBUTIVITY
9.1 INTRODUCTION TO DISTRIBUTIVITY
9.2 LEXICALLY-ENCODED DISTRIBUTIVITY
9.3 DEVELOPMENT OF THE UNDERSTANDING OF THE DISTRIBUTIVITY OF EACH
9.4 REAL-TIME BEHAVIOURAL MEASURES OF DISTRIBUTIVITY AND EACH
9.5 CONCLUSIONS
ACKNOWLEDGEMENT
CHAPTER 10: GENERICITY
10.1 INTRODUCTION
10.2 WHAT IS GENERICITY?
10.2.1 Main definitions and facts
10.2.2 Main characteristics of generics
10.2.2.1 Temporal unboundedness
10.2.2.2 Law-like or nomic character
10.2.2.3 Association with dispositions
10.2.2.4 Resistance to contextual restriction
10.2.2.5 Exception tolerance
10.2.3 Genericity: A mixed bag or a single phenomenon?
10.3 ACCOUNTS OF GENERICITY
10.3.1 Formal semantics
10.3.1.1 Carlson’s monadic generic operator
10.3.1.2 The modal approach, aka the received view: Krifka et al. ()
10.3.1.3 The probabilistic approach: Cohen
10.3.1.4 Recent proposals questioning the quantificational view
10.3.2 Cognitive psychology
10.3.2.1 The Generics-as-Default view
10.3.3 Philosophy
10.3.4 Summary and some considerations
10.4 EXPERIMENTAL APPROACHES TO GENERICITY
10.4.1 Experimental psychology
10.4.1.1 Off-line methods: Adult judgement data
10.4.1.2 Online methods: Real-time processing and processing in the brain
10.4.2 Experiments with generics in linguistically-driven research
10.4.2.1 Typical adult populations
10.4.2.2 Bilingual and second language adult acquisition
10.4.3 Generics in child language acquisition studies
10.5 REFLECTION ON THE EXPERIMENTAL LITERATURE ON GENERICITY
10.6 CONCLUSION
ACKNOWLEDGEMENTS
CHAPTER 11: MODIFIED NUMERALS
11.1 INTRODUCTION
11.2 IGNORANCE INFERENCES
11.2.1 Experimental support for the semantic account and the class A/B distinction
11.2.2 Evidence for a pragmatic account
11.2.3 (More) direct investigations of ignorance inferences
11.3 VARIATION/DISTRIBUTIVITY EFFECTS
11.4 DISCUSSION/CONCLUSION
11.4.1 Probing pragmatics
11.4.2 The comparative–superlative (A–B) distinction revisited
ACKNOWLEDGEMENTS
CHAPTER 12: NEGATION
12.1 NEGATION PROCESSING: THE EXTRA EFFORT AND THE ROLE OF THE POSITIVE ARGUMENT
12.2 THE REPRESENTATION OF THE POSITIVE ARGUMENT: WHEN IT HAPPENS AND WHEN IT DOESN'T
12.3 NEGATION IS NOT DIFFICULOT WITH CONTEXT
12.4 NEGATION HAS RICH PRAGMATIC EFFECTS
12.5 ACCOUNTS OF NEGATIVE SENTENCE PROCESSING
12.5.1 Rejection accounts
12.5.2 Contextual approach
12.5.3 Dynamic pragmatic account
12.5.4 Evaluating theories of negation processing against empirical findings
12.5.5 General implications and future directions
12.6 CONCLUSION
CHAPTER 13: PLURALITY
13.1 INTRODUCTION
13.2 THE IMPLICATURE APPROACH
13.2.1 The implicature approach and its predictions
13.2.2 Experimental work
13.2.2.1 Sauerland et al. (2005)
13.2.2.2 Pearson et al. (2011)
13.2.2.3 Tieu et al. (2014, 2015, 2017)
13.2.2.4 Patson (2016)
13.3 AMBIGUITY-BASED APPROACHES AND THEIR PREDICTIONS
13.3.1 Farkas & De Swart (2010)
13.3.2 Grimm (2013)
13.4 GENERAL DISCUSSION
13.4.1 Sensitivity to monotonicity
13.4.2 Suspension and context-dependence
13.4.3 Children’s development of plural meanings
13.5 CONCLUDING REMARKS
CHAPTER 14: QUALIFICATION
14.1 INTRODUCTION: WHAT IS QUANTIFIER SCOPE?
14.2 FACTORS INFLUENCING THE GRAMMAR AND PROCESSING OD QUANTIFIER SCOPE
14.3 THE DIFFICULTY OF PROCESSING INVERSE SCOPE: IMPIRICAL FINDINGS
14.4 PROCESSING THEORIES OF QUANTIFIER SCOPE
14.4.1 Underspecification theories vs. hierarchy-based accounts
14.4.2 Cost of inverse scope as scope revision
14.4.2.1 Logical form revision
14.4.2.2 Discourse model revision
14.4.3 Distinguishing between logical form and discourse-model theories
14.5 CONCLUSION: BRINGING IT ALL TOGETHER
ACKNOWLEDGEMENTS
CHAPTER 15: QUANTIFIER SPREADING
15.1 INITIAL DESCRIPTIONS OF ERRORS WITH UNIVERSAL QUANTIFIERS
15.2 LOGICAL ERRORS IN SYLLOGISTIC REASONING
15.3 STRUCTURE-NEUTRAL REPRESENTAIONS OF SENTENCES WITH UNIVERSAL QUANTIFICATION
15.4 DISTINGUISHING THE MEANINGS OF ALL AND EACH
15.5 THE EVENT QUANTIFICATION HYPOTHESIS AND RELATED SYNTACTIC ACCOUNTS OF QUANTIFIER SPREADING
15.6 CHILDREN'S ERRORS MAY REFLECT TASK DEMANDS, NOT FAULTY GRAMMAR
15.7 EYE-TRACKING STUDIES OF QUANTIFIER SPREADING
15.8 CONCLUSION
ACKNOWLEDGEMENTS
CHAPTER 16: ADJECTIVE MEANING AND SCALES
16.1 INTRODUCTION
16.2 REINFORCING INTUITION-BASED DATA
16.2.1 Scale structure and the absolute/relative distinction
16.3 GOING BEYONG INTUITIONS
16.3.1 Relative gradable adjectives and context-dependence
16.3.2 The logic of vagueness
16.4 MAPPING THE EMPIRICAL LANDSCAPE
16.4.1 Ordering Subjectivity
16.4.2 Adjectives and implicature
16.5 NOTES ON METHODOLOGY
16.5.1 Design
16.5.2 Task
16.5.3 Question/answer wording
16.5.4 Stimuli
16.6 CONCLUSIONS
ACKNOWLEDGEMENTS
CHAPTER 17: IRONIC UTTERANCES
17.1 GRICE: BRINGING AN ATTITUDE
17.2 THE PSYCHOLINGUISTIC APPROACH TO IRONY
17.2.1 Direct Access
17.2.2 Graded salience
17.2.3 A middle way
17.2.4 Interim conclusions
17.3 THEORY OF MIND
17.4 REINTRODUCING THEORY OF MIND TO LANGUAGE PROCESSING
17.5 RECONCILING THEORETICAL PRAGMATIC APPROACHES WITH PSYCHOLINGUISTIC METHODS: DOING EXPERIMENTAL PRAGMATICS
17.5.1 Links between Theory of Mind and irony processing?
17.6 CONCLUSIONS
CHAPTER 18: METAPHOR
18.1 INTRODUCTION
18.2 METAPHOR AND LITERAL MEANING
18.2.1 Theoretical approaches to metaphor
18.2.2 Literal first hypothesis
18.2.3 What place for the literal meaning?
18.3 HOW ARE METAPHORS PROCESSED?
18.3.1 Class-inclusion model
18.3.2 Career of Metaphor hypothesis
18.3.3 Aptness and conventionality
18.4 VARIOUS METAPHORS, VARIOUS TROPES
18.4.1 Are all metaphors processed in the same way?
18.4.2 Metaphors vs. other tropes
18.5 HOW CHILDREN INTERPRET METAPHORS
18.5.1 How early do children understand metaphors?
18.5.2 Which factors influence children’s understanding of metaphors?
18.6 CONCLUSION
ACKNOWLEDGEMENTS
CHAPTER 19: METONYMY
19.1 TYPES AND FUNCTIONS
19.2 MEANINGS ALTERNATIONS
19.2.1 One word, multiple meanings
19.2.2 One word, multiple senses
19.3 MULTIPLE SENSES VS. EXTENDED SENSES
19.4 TOWARDS AN ACCOUNT OF MEANING CONSTITUTION
19.5 METONYMY AND LANGUAGE ACQUISITION
19.6 METONYMY AND LANGUAGES DEFICITS
19.7 FUTURE DIRECTIONS
CHAPTER 20: VAGUENESS
20.1 INTRODUCTION
20.2 THEORIES OF VAGUENESS
20.2.1 Valency two: Epistemicism
20.2.2 Valency three: Super-/subvaluationism, strict/tolerant truth
20.2.2.1 Super-/subvaluationism
20.2.2.2 Strict/tolerant evaluation
20.2.3 Valency greater than three: Fuzzy logic
20.3 EXPERIMENTAL RESEARCH ON THE BORDERLINE
20.4 BEYOND BORDERLINE CONTRADICTIONS
20.4.1 Imprecision and a typology of vagueness
20.4.2 Round numbers
20.5 CONCLUDING REMARKD
ACKNOWLEDGEMENT
CHAPTER 21: VERBAL UNCERTAINTY
21.1 INTRODUCTION
21.2 OVERVIEW OF METHODS
21.2.1 Mapping verbal probabilities onto a numerical probability space
21.2.1.1 Point numerical values
21.2.1.2 Numerical intervals
21.2.1.3 Membership functions
21.2.2 Mapping verbal probabilities onto an outcome distribution space
21.2.3 Verbal probabilities, spontaneous production, and selection
21.3 OVERVIEW OF FINDINGS
21.3.1 Key results on numerical interpretation
21.3.2 Directionality
21.3.3 Continuous outcomes
21.4 CONCLUSION
CHAPTER 22: WORD SENSES
22.1 THE STRUCTURE OF LEXICAL FLEXIBILITY
22.1.1 Explaining patterns of flexibility
22.1.2 Explaining generalization
22.2 LEXICAL REPRESENTATION OF WORD SENSES
22.3 WHY DO WORDS HAVE DISTINCT SENSES?
22.3.1 Functional pressures and lexical flexibility
22.3.2 Mechanisms and constraints for learning senses
22.4 CONCLUSION
ACKNOWLEDGEMENTS
CHAPTER 23: ANTECEDENT-CONTAINED DELETION
23.1 INTRODUCTION TO THE PHENOMENON
23.2 INVESTIGATIONS OF ACD IN CHILD LANGUAGE ACQUISITION
23.2.1 Evidence for QR in child grammar
23.2.2 Evidence for the landing site of QR in child grammar
23.2.3 Evidence for multiple landing sites for QR in child grammar
23.3 INVESTIGATIONS OF ACD IN ADULT GRAMMAR
23.3.1 ACD as evidence against clause-boundedness of QR
23.3.2 ACD and possible evidence for theoretical frameworks
23.4 CONCLUSIONS AND FUTURE DIRECTIONS FOR RESEARCH ON ACD
ACKNOWLEDGEMENT
CHAPTER 24: EXHAUSTIVITY IN IT-CLEFTS
24.1 INTRODUCTION
24.2 THE THEORETICAL OPTIONS
24.2.1 At-issue content
24.2.2 Conversational implicature
24.2.3 Presupposed exhaustivity
24.2.4 Indirect exhaustivity
24.3 EXPERIMENTAL EVIDENCE
24.4 DISCUSSION
24.5 CONCLUSION
ACKNOWLEDGEMENTS
CHAPTER 25: FOCUS
25.1 ATTENTION, MENORY, AND DEPTH OF PROCESSING
25.2 FOCUS AND SYNTACTIC AMBIGUITY RESOLUTION
25.3 FOCUS AND INTERPRETATION
25.3.1 Focus as cue to discourse structure
25.3.2 Anaphoric dependencies
25.3.3 Focus and reference resolution in visual displays
25.3.4 Inferring implicit alternatives
25.4 CUES TO FOCUS STRUCTURE
25.4.1 Conflicting cues to focus
25.4.2 Cues to focus projection
25.5 FUTURE DIRECTIONS
25.5.1 Different focus operators
25.5.2 Contrastiveness and the sources of contrast sets
25.5.3 Cross-linguistic variation
CHAPTER 26: NEGATIVE POLARITY ITEMS
26.1 INTRODUCTION: WHAT IS AN NPI?
26.2 LINGUISTIC PROPERTIES
26.2.1 Downward entailment
26.2.2 Non-veridicality
26.2.3 Domain widening and exhaustifying the alternatives
26.2.4 Summary
26.3 PROCESSING NEGATIVE POLARITY ITEMS
26.3.1 Basic sensitivity to the licensing condition
26.3.2 Illusory licensing effect
26.3.3 Summary
26.4 ACQUIRING NEGATIVE POLARITY ITEMS
26.4.1 Children’s knowledge about NPI licensing
26.4.2 What is in the input?
26.4.3 Summary
26.5 CONCLUSIONS AND FUTURE DIRECTIONS
CHAPTER 27: PRONOUNS
27.1 ONE PRONOUN, MANY FACTORS DIFFERENT MODELLING APPROACHES
27.1.1 Does syntax matter?
27.1.2 Does thematic role matter?
27.1.3 Does distance matter?
27.1.4 Does topichood matter?
27.1.5 Prior models of pronoun use
27.2 STOP LOOKING AT PRONOUNS TO UNDERSTAND PRONOUNS
27.3 LIKELY MESSAGES AND LIKELY FORMS
27.4 PRONOUNS IN A GENERATIVE MODEL: A BAYESIAN APPROACH
27.4.1 Coherence-driven factors influence next mention
27.4.2 Topichood influences pronominalization
27.4.3 Model comparison
27.5 HOW DOES A GENERATIVE MODEL CLARIFY PRONOUN PUZZLES?
ACKNOWLEDGEMENTS
CHAPTER 28: REFERENCES AND INFORMATIVENESS
28.1 THE PROBLEM OF REFERENTIAL CHOICE
28.2 THEORETICAL APPROACHES TO THE PRODUCTION OF REFERENTIAL FORMS
28.2.1 Gricean maxims
28.2.2 Informativeness
28.2.2.1 What is informativeness?
28.2.2.2 Empirical investigations of informativeness in production
28.2.2.3 Empirical investigations of informativeness in comprehension
28.2.3 Discourse theories: Framework and empirical evidence
28.2.4 Comparing informativeness and discourse approaches
28.3 REAL-LIFE REFERENCE PRODUCTION: CONTEXTUAL CONSTRAINTS ON REFERENTIAL CHOICE
28.3.1 What’s in the context? Linguistic and non-linguistic sources of information
28.3.2 Constraints from interlocutors
28.3.2.1 Does common ground matter?
28.3.2.2 Social cues; gestures
28.3.2.3 Conceptual pacts
28.3.2.4 Speaker goals
28.3.3 Constraints from the referent
28.3.4 Pulling together contextual constraints on referring
28.4 FUTURE DIRECTIONS IN REFERENCE PRODUCTION
28.5 CHAPTER SUMMARY
ACKNOWLEDGEMENT
CHAPTER 29: PROSODY AND MEANING
29.1 INTRODUCTION
29.2 PROSODY: A BRIEF PRIMER
29.3 THE PRODUCTION OF PROSODICALLY REALIZED FOCUS
29.3.1 Prosodic realization of information-structural focus across languages
29.3.2 The prosodic realization of focus is context-dependent
29.3.3 Prosodic realization of other information-structural properties of foci
29.3.4 Prosodic marking of focus size
29.3.5 Experimental tasks to explore the prosodic realization of focus
29.4 THE PERCEPTION AND INTERPRETATION OF PROSODICALLY REALIZED FOCUS
29.4.1 The perception and interpretation of focus across languages
29.4.2 Contextual influence on the perception and interpretation of prosodically realized focus
29.4.3 Interpreting prosodic cues to foci that differ in other information-structural properties
29.4.4 Perceiving and interpreting the size of prosodically realized focus
29.4.5 Experimental tasks for the perception and interpretation of prosodically realized focus
29.5 CONCLUDING REMARKS
ACKNOWLEDGEMENTS
CHAPTER 30: POLITENESS
30.1 A BRIEF OVERVIEW OF PLOITENESS THEORIES
30.2 TESTS OF BROWN & LEVINSON'S MODEL
30.3 SOCIAL INTERACTIONAL DETERMINATIONS OF POLITENESS
30.4 POLITENESS, REASONING, AND THE COMMUNICATION OF UNCERTAINITY
30.5 PROCESSING POLITENESS
30.6 CONCLUSION
ACKNOWLEDGEMENTS
CHAPTER 31: THEORY OF MIND
31.1 INTRODUCTION
31.2 BRIEF REASONING: AT WHAT AGE, AND WHEN?
31.3 THEORY OF MIND, LANGUAGE AND PRAGMANTICS: RELATIONS AND PARALLELS
31.4 TWO-YEARS-OLDS: DO THEY KNOW BETTER NOW THAN THIRTY YEARS AGO?
31.5 THREE-YEARS-OLDS' UNDERSTANDING OF FACTIVITY: WE KNOW, OR WE THINK?
31.6 EARLY REFERENTIAL COMMUNICATION: ENGAGEMENT AND DISENGAGEMENT
31.7 PRESCHOOLERS' REASONING ABOUT OTHERS' PERCEPTIONS: FROM SEEING TO KNOWING
31.8 SUMMARY AND CONCLUSIONS
ACKNOWLEDGEMENTS
CHAPTER 32: TURN-TAKING
32.1 THE RULES
32.2 ISSUES AND CONTROVERSIES
32.2.1 The status of The Rules
32.2.2 The processing of The Rules
32.2.3 Cues for predicting the end of turns
32.3 RESEARCH METHODS FOR INVESTIGATING TURN-TAKING
32.3.1 Conversation analysis
32.3.2 Corpus analysis
32.3.3 Off-line experimental methods
32.3.4 On-line experimental methods
32.4 VALIDITY ISSUES
32.5 SUMMARY
REFERENCES
INDEX
Recommend Papers
- Author / Uploaded
- Chris Cummins
- Napoleon Katsos
File loading please wait...
Citation preview
EXPERIMENTAL SEMANTICS AND PRAGMATICS
OXFORD HANDBOOKS IN LINGUISTICS Recently published
THE OXFORD HANDBOOK OF DEVELOPMENTAL LINGUISTICS Edited by Jeffrey Lidz, William Snyder, and Joe Pater
THE OXFORD HANDBOOK OF INFORMATION STRUCTURE Edited by Caroline Féry and Shinichiro Ishihara
THE OXFORD HANDBOOK OF MODALITY AND MOOD Edited by Jan Nuyts and Johan van der Auwera
THE OXFORD HANDBOOK OF PRAGMATICS Edited by Yan Huang
THE OXFORD HANDBOOK OF UNIVERSAL GRAMMAR Edited by Ian Roberts
THE OXFORD HANDBOOK OF LANGUAGE AND SOCIETY Edited by Ofelia García, Nelson Flores, and Massimiliano Spotti
THE OXFORD HANDBOOK OF ERGATIVITY Edited by Jessica Coon, Diane Massam, and Lisa deMena Travis
THE OXFORD HANDBOOK OF WORLD ENGLISHES Edited by Markku Filppula, Juhani Klemola, and Devyani Sharma
THE OXFORD HANDBOOK OF POLYSYNTHESIS Edited by Michael Fortescue, Marianne Mithun, and Nicholas Evans
THE OXFORD HANDBOOK OF EVIDENTIALITY Edited by Alexandra Y. Aikhenvald
THE OXFORD HANDBOOK OF LANGUAGE POLICY AND PLANNING Edited by James W. Tollefson and Miguel Pérez-Milans
THE OXFORD HANDBOOK OF PERSIAN LINGUISTICS Edited by Anousha Sedighi and Pouneh Shabani-Jadidi
THE OXFORD HANDBOOK OF ELLIPSIS Edited by Jeroen van Craenenbroeck and Tanja Temmerman
THE OXFORD HANDBOOK OF LYING Edited by Jörg Meibauer
THE OXFORD HANDBOOK OF TABOO WORDS AND LANGUAGE Edited by Keith Allan
THE OXFORD HANDBOOK OF MORPHOLOGICAL THEORY Edited by Jenny Audring and Francesca Masini
THE OXFORD HANDBOOK OF REFERENCE Edited by Jeanette Gundel and Barbara Abbott
THE OXFORD HANDBOOK OF EXPERIMENTAL SEMANTICS AND PRAGMATICS Edited by Chris Cummins and Napoleon Katsos
For a complete list of Oxford Handbooks in Linguistics, please see pp. –
.........................................................................................................................................
EXPERIMENTAL SEMANTICS AND PRAGMATICS ......................................................................................................................................... Edited by
CHRIS CUMMINS and
NAPOLEON KATSOS
1
3
Great Clarendon Street, Oxford, , United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © editorial matter and organization Chris Cummins and Napoleon Katsos © the chapters their several authors The moral rights of the authors have been asserted First Edition published in Impression: All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press Madison Avenue, New York, NY , United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: ISBN –––– Printed and bound by CPI Group (UK) Ltd, Croydon, Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.
C
................................
Acknowledgements List of Figures and Tables List of Abbreviations The Contributors . Introduction
viii ix xi xiii
C C N K
. Language Comprehension, Inference, and Alternatives
D S D B
. Constraint-based Pragmatic Processing
J D M K. T
. Scalar Implicatures
R B
. Event (De)composition
S Y C E. M H
. Presuppositions, Projection, and Accommodation
F S
. Spatial Terms
M G A P
. Counterfactuals
H F
. Distributivity
K S
. Genericity
D L-C
. Modified Numerals R N, S A, Y MN
vi
. Negation
Y T R B
. Plurality
L T J R
. Quantification
A B J Dč
. Quantifier Spreading
P J. B O P
. Adjective Meaning and Scales
S S
. Ironic Utterances
N S I N
. Metaphor
N P G D
. Metonymy
P B. S
. Vagueness
S A U S
. Verbal Uncertainty
M J, M S, J-Fç B
. Word Senses
H R M S
. Antecedent-contained Deletion
K S
. Exhaustivity in it-clefts
E O
. Focus C S. K
. Negative Polarity Items
vii
M X
. Pronouns
H R
. Reference and Informativeness
C D J E. A
. Prosody and Meaning
J T
. Politeness
T H
. Theory of Mind
P R-F Á
. Turn-taking
J. P. R
References Index
A
..................................................................
W would like to express our gratitude to our colleagues who paved the way in experimental semantics and pragmatics and inspired us, among many others, to pursue research in this area. Through their work, they not only staked out a new field of enquiry, but created outstanding opportunities to bring scholars together to work in it, such as the XPrag conferences and the Euro-XPrag network (supported by the European Science Foundation). Our thanks go to the contributors to this volume for their enthusiastic support of this enterprise, their insight, and their professionalism, and for reminding us once again of the breadth as well as the depth of experimental semantic and pragmatic research. Thanks also to three anonymous reviewers for their important and astute comments at a very early stage that helped to shape the structure and content of this volume. Finally, our thanks go to the late John Davey who responded so positively from our initial contact with Oxford University Press, and to Julia Steer and Vicki Sunter, for their kind (and crucial) support and assistance throughout the project.
L ...................................................................................... Figures .. The average probability of fixating either the real-world or counterfactual-world object
..
Examples of collective and distributive pictures
..
Examples of collective, distributive, and exhaustive pictures
.. Examples of collective pictures showing sets with extra objects or extra actors
..
Grand average ERPs elicited by critical words
..
The average probability of fixating either the Reality or Belief location
.. Every farmer is feeding a donkey
..
Examples of distributive pictures showing sets with extra objects or actors
.. Standard semantic representation of the sentence ‘Every farmer is feeding a donkey’ ..
Event quantification representation of the sentence ‘Every farmer is feeding a donkey’
..
Fixations to the extra containers as a function of accuracy in sentence-picture verification
..
Sample stimuli from comparison class experiments
..
Schematic illustration of expected ERP effects
.. Example in tolerant/strict model ..
Visual context: ‘Contradictions at the Borders’
.. Visual context: ‘The Psychology of Vagueness’ ..
The point numerical estimates of seven probability expressions
.. Membership functions for two verbal probability expressions: ‘unlikely’ and ‘likely’
..
How much will the sea level rise?
..
Distribution of quantities provided to participants for different verbal probabilities
x
..
A sample discourse structure containing open positions to which subsequent clauses may attach
..
Implicit causality story continuation results: antecedent grammatical role, but not verb class
..
Voice manipulation story continuation results: antecedent grammatical role and voice
..
Voice manipulation story continuation results: voice and prompt type
..
Distribution of turn duration
.. Implicit causality story continuation results: verb class and prompt type
Tables ..
Four forms of categorical syllogisms with example problems
.. A summary of a paradigm .. Definitions of variables in metaphor processing research ..
Theory classification by valency and scope
.. Predictions of (global) super-/subvaluationist accounts for borderline cases .. Predictions of local super-/subvaluationist accounts for borderline cases ..
Predictions of strict/tolerant semantics for borderline cases
.. Predictions of the (global) strict/tolerant+SMH model for borderline cases ..
Predictions of the (local) strict/tolerant+SMH model for borderline cases
..
Predictions of standard fuzzy logic for borderline cases
.. Predictions of the ‘scaling’ fuzzy model for borderline cases
.. Examples of numerical meaning for the verbal probabilities likely and unlikely, with and without modifier
..
Examples of positive and negative verbal probabilities conveying a range of numerical probabilities
..
Examples of lexical flexibility in English
..
Choices between referential expressions
L ........................................................................ A&P ACD AOC AQ ASD B&L BMBF BOVW CA CLT CP DAAD DE DFG DP DRS DRT EE EEG ERP ESF ESRC FA FBT fMRI FN FoR FTO GaD GOG GQT GT IC II IPCC ISI LF LP
Alxatib and Pelletier () Antecedent-Contained Deletion Adjectives of Comparison Autism-spectrum Quotient Autism Spectrum Disorder Brown and Levinson () Bundesministerium für Bildung und Forschung Bonini et al. () Conversation Analysis Construal Level Theory (Trope & Liberman ) Cooperative Principle (Grice, ) Deutscher Akademischer Austauschdienst downward entailing Deutsche Forschungsgemeinschaft Determiner Phrase Discourse Representation Structure (Kamp, ) Discourse Representation Theory (Kamp, ) Embedded (Scalar) Enrichments electroencephalography Event-Related Potential European Science Foundation Economic and Social Research Council false affirmative False Belief Task Functional Magnetic Resonance Imaging false negative Frame of Reference Floor Transfer Offset (De Ruiter et al., ) Generics-as-Default Generic Overgeneralization Generalized Quantifier Theory Grammatical Theory (of scalar implicature) Implicit Causality Ignorance Inferences Intergovernmental Panel on Climate Change inter-stimulus interval Logical Form Logic of Paradox (Priest, )
xii
LSA MEG MPFC ms MTurk NP NPI NSF NWO PC PD PLA PP PSI QNP QR QUD RHD RSA RT SAT SHZ SMH SPM SS TA TCU TD TFA TMS TN ToBI ToM TPJ TRP TVJ(T) UE VP VPE ZAS
Linguistic Society of America magnetoencephalography medial prefrontal cortex milliseconds Amazon Mechanical Turk Noun Phrase Negative Polarity Item National Science Foundation Nederlandse Organisatie voor Wetenschappelijk Onderzoek precuneus Parkinson’s disease Pair-List Answer Politeness Principle (Leech, ); Prepositional Phrase Principle of Scope Interpretation (Tunstall, ) Quantified Noun Phrase Quantifier Raising Question Under Discussion right hemisphere damage Rational Speech Act Relevance Theory Speed-Accuracy Trade-off Serchuk et al. () Strongest Meaning Hypothesis (Dalrymple et al., ) Standard Pragmatic Model Straight Scalars true affirmative Turn Construction Unit (Sacks et al., ) Typically Developing Time Frequency Analysis Transcranial Magnetic Stimulation true negative Tones and Break Indices Theory of Mind temporo-parietal junction Transition Relevance Place (Sacks et al., ) Truth-Value Judgment (Task) upward entailing Verb Phrase Verb Phrase Ellipsis Leibniz-Zentrum Allgemeine Sprachwissenschaft
T
............................................................
Stavroula Alexandropoulou is a PhD candidate at Utrecht University. Her research lies in the area of semantics and pragmatics combining theoretical and experimental approaches, with a special focus on the pragmatics of scalar quantifiers such as modified numerals. Other topics she has worked on include bare singular count nouns and inalienable possession. Sam Alxatib is Assistant Professor in Linguistics at the Graduate Center of the City University of New York. He studied linguistics in Simon Fraser University in Burnaby, British Columbia, and in the Massachusetts Institute of Technology. He is interested in semantic, pragmatic, and syntactic theory, and in methodological issues in linguistic research. He has worked on vague phenomena, gradability and degree semantics, the semantics of evaluativity, focus, tense, and aspect, and the interaction of aspect with modality. Jennifer E. Arnold is Professor in the department of Psychology and Neuroscience at the University of North Carolina, Chapel Hill. She does research on the psychology of language, examining how the discourse context affects both language production and comprehension. Much of her research examines reference. When does the speaker say ‘she’ vs. ‘the woman’? How clearly are the words pronounced? Dr Arnold examines these questions as they relate to processes of utterance planning, disfluency, and the ability for speakers to model the knowledge and perspective of their addressee. David Barner is Professor of Psychology and Linguistics at the University of California, San Diego. His research spans topics including word learning, counting, grammatical number, quantifier acquisition, and pragmatic development, with a focus on cross-linguistic comparisons. Jean-François Bonnefon (PhD, cognitive psychology) is Research Director at the Toulouse School of Economics (France). He is affiliated with the Toulouse School of Management Research, the Centre National de la Recherche Scientifique, and the Institute for Advanced Study in Toulouse. He studies the rational mind in its various manifestations: reasoning, decision-making, and morality. His recent research applies the insights of moral psychology and behavioural economics to the new challenges of machine ethics and human-AI cooperation. Adrian Brasoveanu is Professor in the Department of Linguistics and Affiliated Faculty in the Center for Statistical Analysis in the Social Sciences, UC Santa Cruz. He is a formal semanticist and a computational psycholinguist. The main question driving his research is:
xiv
what is linguistic meaning and how does the human mind grasp it? He builds evidence-based mathematically and computationally explicit theories of natural language meaning (product) and interpretation (process) that try to answer this question. Richard Breheny is Reader in Linguistics in the Division of Psychology and Language Sciences at University College London. His work spans theoretical semantics and pragmatics, experimental linguistics, and psycholinguistics. Together with his then student Napoleon Katsos, he organized the first official ‘Experimental Pragmatics’ conference in Cambridge, . He was also a founder of the original ESF-funded XPrag Research Network. Patricia J. Brooks is Professor of Psychology at the College of Staten Island, City University of New York (CUNY) where she directs the Language Learning Laboratory. She serves as the Deputy Executive Officer of the CUNY PhD Program in Psychology (Area: Pedagogy) and Faculty Advisor to the Graduate Student Teaching Association of the American Psychological Association. Her research interests are in two broad areas: () individual differences in L and L acquisition; () development of effective pedagogy to support diverse learners. Sherry Yong Chen is a PhD student at the Department of Linguistics and Philosophy, Massachusetts Institute of Technology. She began her life as a linguist at the University of Hong Kong, and she was later trained at the University of Oxford. Sherry is broadly interested in issues related to syntax, semantics, and their interfaces, involving both theoretical and experimental inquiry. Her recent experimental work has a particular focus on language acquisition as well as the role of memory mechanisms during sentence processing. Chris Cummins is a Lecturer at the University of Edinburgh. He previously worked at Bielefeld University, having received his PhD from the University of Cambridge. His research interests include the pragmatic interpretation of quantity expressions, and how this bears upon reasoning and decisionmaking; he has also published on the topics of presupposition, speech acts, and metaphor. He is the author of Constraints on Numerical Expressions (OUP, ). Catherine Davies is Lecturer in Linguistics in the School of Languages, Cultures & Societies at the University of Leeds. Catherine uses experimental data (behavioural; eye-tracking) to investigate the development and processing of pragmatics, especially in the use of referring expressions. Her main area of research investigates how children and adults integrate information from communicative contexts into their referential choices in production and comprehension. She is also interested in psycholinguistic processing within its social context, for example how language environments interact with language processing and development. Judith Degen is Assistant Professor of Linguistics at Stanford University. Trained as a cognitive scientist at the University of Rochester and Stanford University, Judith is interested in the inference processes involved in language production and comprehension— how do speakers choose an utterance to convey an intended meaning? How do listeners
xv
arrive at interpretations that are often much richer and more detailed than the literal meaning provided by a sentence? She employs a combination of linguistic analysis, behavioural methods, corpus methods, and computational models to develop explicit theories of these processes and test them against behavioural data. J. P. de Ruiter is Bridge Professor in the Cognitive Sciences at the departments of Computer Science and Psychology at Tufts University. He was trained as a cognitive scientist at Radboud University Nijmegen, and developed a cognitive model of speech and gesture production during his PhD research project at the Max Planck Institute for Psycholinguistics. His primary interests are the cognitive foundations of communication, social robotics, gesture, turn-taking, pragmatics, and the intersection between scientific methodology and philosophy of science. He also co-produces a podcast on Bayesian statistics. Jakub Dotlačil is a postdoctoral researcher at the Institute of Logic, Language and Computation at the University of Amsterdam. He received his PhD in linguistics from Utrecht University. His main research interests are formal semantic theories of quantification and discourse, experimental research on processing, and cognitive modelling of processing and incremental interpretation. He investigates properties of symbolic systems, as developed, for example, in formal semantics, as well as how such symbolic systems are to be linked to human performance and behavioural data. Giulio Dulcinati trained in linguistics and experimental psychology at University College London. He completed a PhD in Linguistics in . He is interested in the experimental investigation of pragmatic inferences and how they are affected by the assumptions that we make about our interlocutors. Most of his experimental work focuses on the role of assumptions of cooperation among interlocutors in the derivation of pragmatic inferences in conversation. Heather Ferguson is Professor of Cognitive Psychology at the University of Kent. She trained in psychology at the University of Glasgow, then worked as a postdoctoral researcher in linguistics at University College London. Heather’s research focuses on the interface between cognitive processes and social interaction, specifically the way that we access and represent other people’s perspectives during communication. She employs a variety of techniques, including eye movements, neuroscientific methods, and behavioural techniques to examine the time-course of integration, the underlying neural mechanisms, and the extent to which constraints from world knowledge and context compete to influence social communication. Myrto Grigoroglou is a PhD candidate in Linguistics & Cognitive Science at the University of Delaware. She completed her undergraduate studies at the University of Athens, Greece, and has an MA from the University Paris -Sorbonne Nouvelle, France. Myrto studies how children and adults use language in communication with an emphasis on referential communication and pragmatic inference. She is particularly interested in how the
xvi
pragmatic system develops in children and how it interacts with the semantic system as well as other non-linguistic aspects of cognition (e.g. social or spatial cognition). Thomas Holtgraves is Professor of Psychological Science at Ball State University in Muncie, Indiana. He conducts interdisciplinary research into multiple facets of language and social psychology. He is the editor of the Oxford Handbook of Language and Social Psychology (Oxford University Press, ) and author of Language as Social Action: Social Psychology and Language Use (Erlbaum, ). E. Matthew Husband is Associate Professor in the Faculty of Linguistics, Philology, and Phonetics at the University of Oxford and Tutorial Fellow of St. Hugh’s College. Prior to his current post, Matt studied at Michigan State University, taught at Brown University, and was a postdoc at University of South Carolina. Matt has had a sustained interest in grammatically relevant elements of sentence meanings and how they are computed in real time. His research makes use of behavioural and neurophysiological techniques, from acceptability judgements and eye movements to electroencephalography and functional magnetic resonance imaging. Marie Juanchich (PhD, Socio-Cognitive Psychology) is Senior Lecturer in Psychology at the Faculty of Science and Health at the University of Essex. Marie trained as an experimental psychologist in Toulouse, France. Her research generally focuses on uncertainty and risk communication. She is particularly interested in the way people leverage the language of uncertainty to convey subtle information and how this subtle information shapes the judgements and decisions of recipients. One of the reasons this topic is of particular interest is its many potential applications (e.g. medicine, finance, climate change)—given that nothing is certain! Marie’s work draws from the literature in social and cognitive psychology but also from linguistics and mostly uses experimental methodologies. Napoleon Katsos is a Reader in Experimental Pragmatics at the University of Cambridge. He is interested in how we learn, process, and use the meaning of words and sentences, with emphasis on quantification and implicature. He draws relevant evidence from linguistic theory and experimental psychology, including sentence processing and typical or atypical language acquisition by monolingual or bilingual children. Christina S. Kim is Lecturer in Linguistics in the Department of English Language and Linguistics at the University of Kent. Since early in her studies, Christina has been involved in both linguistics and language processing, training at MIT, UCLA, the University of Rochester, and the University of Chicago. Christina is interested in how we use available linguistic and non-linguistic information to infer context-dependent aspects of meaning. She has worked on topics like presupposition, focus, ellipsis, and syntactic ambiguity resolution, using behavioural paradigms such as self-paced reading and Visual World eye-tracking. Dimitra Lazaridou-Chatzigoga is a postdoctoral researcher at the Humboldt-Universität zu Berlin and Affiliated Lecturer at the University of Cambridge. Dimitra trained as a linguist in Thessaloniki, Greece, and received a PhD in Cognitive Science and Language from the Universitat Autònoma de Barcelona. Her work investigates meaning and
xvii
variation in meaning as related to variation in form and use, with a focus on determiners, nominals, and quantifiers. She conducts theoretical and experimental research in semantics and pragmatics mainly on Greek and English and draws on data from mono- and bilingual adult and child populations. Yaron McNabb is a researcher and Lecturer at the Department of Languages, Literature and Communication at Utrecht University. Yaron received his PhD in linguistics at the University of Chicago and subsequently worked as a postdoctoral researcher at the University of Konstanz. Yaron investigates how form, structure, and meaning in the sentential and conversational context interact in instances of vagueness and (im)precision, namely in gradable adjectives, degree modifiers, and quantity expressions. Rick Nouwen is Associate Professor at the department of Languages, Literature and Communication at Utrecht University. He has worked on many topics in semantics and pragmatics, including scalarity, quantification, degree semantics, plurality, and pronominal reference. He owns a blue Volvo and is currently the editor-in-chief of the Journal of Semantics. Ira Noveck earned his PhD at NYU () in Experimental Psychology as a specialist of human reasoning. Through his work, he became interested in the semantic contributions of, and the pragmatic inferences associated with, logical terms and how such features could impact on reasoning performance. This led to a programmatic effort to experimentally investigate phenomena associated with pragmatics generally. He is currently a full-time Directeur de Recherche at France’s Centre National de Recherche Scientifique (CNRS) and works at the Institut des Sciences Cognitives—Marc Jeannerod in Lyon. Edgar Onea is Full Professor of Linguistics at the Department for German Studies of the University of Graz. After his PhD at the University of Heidelberg, Edgar Onea worked in Stuttgart and Göttingen. He has published on various topics at the semantics-pragmatics interface including information structure, it-clefts, discourse structure, and referentiality. His experimental research currently concentrates on the interpretation of marked structures and exhaustivity inferences. Anna Papafragou is Professor of Cognitive Psychology and Linguistics at the University of Delaware. She gained her PhD in Linguistics from University College London and received postdoctoral training at the University of California at Berkeley and the University of Pennsylvania. Her work focuses on the nature and acquisition of language, especially in the areas of semantics and pragmatics. She has also conducted extensive cross-linguistic research examining the relation between language and other cognitive systems, especially in the domains of space, events and theory of mind. Olga Parshina is a PhD student in the Psychology Department at the College of Staten Island and the Graduate Center, City University of New York (CUNY). Previously she received an MA in Linguistics and Cognitive Science at the University of Delaware. Her research interests include L and L acquisition, sequence learning, and bilingual language comprehension using the eye-tracking paradigm.
xviii
Nausicaa Pouscoulous trained in philosophy and cognitive science in France and received her PhD from the Institut Jean Nicod and the Institut des Sciences Cognitives. She then studied the development of pragmatic processes in children at the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany. Currently, as Lecturer in Experimental Pragmatics in the Linguistics Department at University College London, she continues to combine experimental and theoretical approaches in her research on pragmatic inferences. Hugh Rabagliati is Reader in the School of Philosophy, Psychology and Language Sciences at the University of Edinburgh. He studied psychology at Oxford and at New York University, where he received his PhD in . He is interested in how children learn, understand, and use language, with a particular focus on semantics. His current research topics include word learning, as well as the development of sentence processing and language production skills. Hannah Rohde is Reader in Linguistics and English Language at the University of Edinburgh. Her training in Linguistics and Computer Science at Brown University led to a PhD in Linguistics at the University of California San Diego, followed by postdoctoral positions at Northwestern and Stanford. She works in experimental pragmatics, using psycholinguistic techniques to investigate questions in areas such as pronoun interpretation, referring expression generation, implicature, presupposition, and the establishment of discourse coherence. Jacopo Romoli is Senior Lecturer in Linguistics at Ulster University and Director of the Masters in English Language and Linguistics. He is also Associate Investigator at the ARC Centre of Excellence in Cognition and its Disorders. Romoli received his PhD from the Department of Linguistics at Harvard University. Romoli’s research focuses on formal semantics and pragmatics for natural languages, language processing, and language acquisition. His current research interests are presuppositions, scalar implicatures, plurality, free choice inferences, neg-raising phenomena, assertability constraints, conditionals, and scope interactions. Paula Rubio-Fernández is a researcher at the Philosophy Department of the University of Oslo. Paula trained as a linguist at the Universidad de Oviedo in Spain and then specialized in experimental pragmatics for her PhD at the University of Cambridge. She is interested in how children and adults use their Theory of Mind in communication. She has designed Theory of Mind tasks to investigate the role of language processing and pragmatics in those paradigms, and also referential communication tasks that require perspective taking and epistemic inferences. Paula is currently researching the development of efficient communication across the lifespan. Uli Sauerland is Vice-Director and Head of the Semantics-Pragmatics department at the Leibniz-Centre General Linguistics (ZAS) in Berlin. His training was first in mathematics at the University of Konstanz, and then in linguistics at the Massachusetts Institute of Technology. He teaches regularly at the University of Potsdam and has taught at many universities as a visiting professor including the Universities of Vienna, Tokyo, Rio de
xix
Janeiro, Harvard, and Stanford. Uli’s goal is to understand the structures underlying human logical thought. To do so, he studies how thoughts are expressed in language, and how language is used to communicate thoughts. Petra B. Schumacher is Professor of Empirical Linguistics at the Department of German Language and Literature I at the University of Cologne, Germany. She received her PhD in Linguistics from Yale University. Her research focuses on processes at the interface of syntax, semantics, and pragmatics, with emphasis on reference resolution, information structure, and various aspects of meaning constitution in discourse. Her key interest is the time-course of language comprehension, which she investigates by means of on-line methods, in particular event-related brain potential studies, cross-linguistic investigations, and research with different speaker populations. Florian Schwarz received his PhD from the University of Massachusetts, Amherst, and is currently Associate Professor of Linguistics at the University of Pennsylvania. He studies natural language comprehension in context from both formal semantic and experimental perspectives. His main line of ongoing research investigates the on-line comprehension of presupposition in relation to current issues in presupposition theory, using a variety of empirical methods, including picture matching tasks, reading time studies, and visual world eye-tracking. Other topics of interest include definite descriptions, implicatures, and intensional semantics. Miroslav Sirota (PhD, cognitive psychology) is Lecturer in Psychology at the Faculty of Science and Health at the University of Essex. In his research, he is trying to understand how people estimate, judge, reason, and make decisions in situations of uncertainty and risk, how people do these ‘on their own’ and ‘in the presence of the others’, ‘in the lab’ and ‘in the wild’. His recent research focused on improving health risk communication between health carers and the public, and diagnostic and management decision-making of family physicians. Dimitrios Skordos is Assistant Professor of Linguistics at the School of Languages, Linguistics, Literatures and Cultures at the University of Calgary. Dimitrios trained at the University of Delaware and at UC San Diego. His main research interests include experimental semantics and pragmatics and the development of pragmatic inference in children with a focus on quantifiers and logical connectives. He investigates these topics using interdisciplinary research methods from theoretical linguistics and developmental psychology. Stephanie Solt is a researcher in semantics at the Leibniz-Zentrum Allgemeine Sprachwissenschaft (ZAS) in Berlin, Germany. She holds an AB in Mathematics from Bryn Mawr College and a PhD in Linguistics from the CUNY Graduate Center. Stephanie’s research focuses on how natural languages encode scalar meaning, and how speakers choose to communicate concepts of degree and quantity. She has worked on topics including quantifiers and quantity adjectives, approximation and scale granularity, adjectival vagueness, degree modification and subjectivity. Her most recent project, funded by the German
xx
Science Foundation (DFG), investigates the semantics of attenuating polarity items in the degree domain. Nicola Spotorno trained in philosophy at University San Raffaele in Milan and earned his PhD in cognitive sciences () at the Institute for Cognitive Sciences in Lyon where he worked in the framework of experimental pragmatics. He has more recently shifted his focus to the investigation of cognitive deficits in neurodegenerative conditions. Nicola is currently a postdoctoral fellow at the University of Lund where he uses neuroimaging in order to uncover biomarkers of neurodegenerative diseases. Mahesh Srinivasan is Assistant Professor in the Department of Psychology at the University of California, Berkeley. He received his BS in Symbolic Systems in from Stanford University and his PhD in Psychology in from Harvard University. Mahesh is interested in how linguistic and conceptual representations arise and interact with one another in human development. His research on language has focused on a range of topics, including word learning and pragmatic development. Kristen Syrett is Associate Professor in the Department of Linguistics at Rutgers, The State University of New Jersey—New Brunswick, with a co-appointment at the Rutgers Center for Cognitive Science (RuCCS). She is Director of the Rutgers Laboratory for Developmental Language Studies. Dr Syrett received her PhD in Linguistics from Northwestern University in , with a specialization in Cognitive Science. She held a one-year postdoctoral appointment at RuCCS, followed by a three-year NRSA postdoctoral fellowship through the National Institutes of Health in Psychology. Dr Syrett’s research focuses on how children acquire language and how we assign interpretations to words and sentences. She is especially interested in instances of ambiguity, and the interaction of syntax, semantics, and pragmatics in topics such as definiteness, predication, quantification, and measurement. Michael K. Tanenhaus is the Beverly Petterson Bishop and Charles W. Bishop Professor of Brain and Cognitive Sciences at the University of Rochester and Chair Professor, School of Psychology at Nanjing Normal University. His research has focused on real-time language processing, including issues in experimental semantics and pragmatics. In the s and early s, Tanenhaus’ research demonstrated that syntactic ambiguity resolution involves rapid integration of probabilistic constraints, including rich lexical representations, results that help form the foundation of current probabilistic approaches. In , Tanenhaus and his students pioneered use of a paradigm—which they dubbed the visual world paradigm— in which eye movements are monitored as participants perform actions as they follow spoken instructions or collaborate with interlocutors in collaborative tasks using natural language. This made it possible to study real-time spoken language processing at a fine temporal grain in natural tasks in rich contexts. Ye Tian is Computational Linguist at Amazon Alexa (UK), Honorary Research Fellow in the Division of Psychology and Language Sciences, University College London, and Associated Member of the Laboratoire de Linguistique Formelle, Université Paris Diderot. Her research area is semantics and pragmatics, with methods including formal, experimental, corpus, and computational.
xxi
Lyn Tieu is Senior Research Fellow at Western Sydney University, in the Education and Aspirational Change theme at the Office of the Pro Vice-Chancellor (Research & Innovation). She is also a member of the MARCS Institute for Brain, Behaviour & Development, Associate Investigator at the ARC Centre of Excellence in Cognition and its Disorders, and Honorary Senior Research Fellow in the Department of Linguistics at Macquarie University. Lyn received her PhD in Linguistics from the University of Connecticut. Her research investigates children’s language development, in particular their acquisition of syntax, semantics, and pragmatics. Judith Tonhauser is Associate Professor in the Department of Linguistics at The Ohio State University. She holds a Diploma in Linguistics from the University of Stuttgart and a PhD in Linguistics from Stanford University. Judith’s research explores universals and variation in how typologically distinct languages convey meaning; recent research areas include temporal interpretation, nominal reference, focus, and presuppositions. She collects primary data from Paraguayan Guaraní (Tupí-Guaraní) and English through one-on-one elicitation, experiments, and corpus work. Ming Xiang is Associate Professor of Linguistics at the University of Chicago. Her primary research area is psycholinguistics, especially sentence and discourse processing. Using a wide range of behavioural and electrophysiological methodologies, her research examines how linguistic structures are represented, stored, and accessed during language comprehension, and how language users make inferences based on both their linguistic knowledge and non-linguistic context.
......................................................................................................................
......................................................................................................................
B through the programmes of major linguistics conferences, and you will often find research on semantics and pragmatics that reports experimental data. In this respect, the field has changed dramatically in the past twenty years. The seeds of this change were not planted in a deliberate and concerted way, but neither is the present blossoming of experimental work a matter of chance. Instead, change came gradually, as pioneering individuals and then bigger communities and research groups came to recognize two facts: first, that linguists are increasingly short of robust and replicable armchair judgements about the acceptability or interpretation of expressions; and, second, that the predictions of competing theories often do not disagree about the acceptability or the interpretation of an expression, but do posit different processes and mechanisms underpinning its formulation, interpretation, or evaluation. Meanwhile, the proliferation of experimental methodologies from behavioural psychology and neuroscience, and their validation and standardization through repeated testing, has presented linguists with a suite of proven methodologies that are available off-the-shelf, and can readily be applied to linguistic questions. From this initial seed, experimental semantics and pragmatics is now flourishing, with an extensive literature in both linguistic and psychological journals. Lively debates in the field play out in a series of dedicated meetings and international conferences, most notably the biennial XPrag Conference which has proved to be a catalyst in establishing and promoting the new discipline. A major international network Euro-XPrag ( , chaired by Ira Noveck) brought young and established scholars from several countries together, while spin-off ongoing national programmes (including XPrag.DE, XPrag.IT, and XPrag.UK, in Germany, Italy, and the UK, respectively) and numerous ad hoc meetings offer ample opportunity for scrutiny of emerging research. The purpose of this Handbook is to capture a snapshot of the questions and methods used in the field, to synthesize the existing findings and to identify under-researched questions or under-utilized techniques. By doing so, we believe that the Handbook will be of use to those already using experimental research to inform theoretical linguistic debates, as well as those who are interested in doing so but haven’t yet made the attempt. We hope also to bring increased awareness of the benefits of experimental research to the wider linguistic community. While we certainly don’t regard experimental research as
essential to every linguistic enterprise, we feel that there are obvious benefits to making quantitative empirical data accessible and interpretable to a wide range of scholars. This is the first handbook of its kind, and it is also a long-awaited one. Although the burgeoning interest in experimental approaches in recent years make this enterprise especially urgent, similar methods have been used to inform and (dis)confirm semantic and pragmatic theories throughout the past half-century. To take a couple of important examples, Eleanor Rosch’s (e.g. ) experimental work on prototypicality introduced a new paradigm in the study of concepts and word meaning, while Herb Clark’s experiments in language use (e.g. Clark & Chase, ; Clark & Marshall, ) informed the study of reference, common ground, and speech acts, among other topics. What seems to be new about the experimental turn in semantics and pragmatics is not the practice itself but an increasing consensus about the relevance of experimental data to linguistic theory. We see this as an emerging understanding that techniques from experimental psychology, cognitive neuroscience, or computer science can fruitfully and legitimately be used as part of the process of answering theoretical linguistic questions. Of course, the label of ‘experimental linguistics’—of which ‘experimental semantics and pragmatics’ is a sub-domain—is also new, and has its own connotations. It suggests a certain separation from ‘psycholinguistics’, which is the term under which much of the earlier work has been categorized. To the extent that there is a real distinction between the two, it can perhaps be understood as follows: ‘experimental linguistics’ strives to obtain empirical data that are relevant to debates framed from a linguistic viewpoint, whereas ‘psycholinguistics’ seeks to understand how language is produced and comprehended in real time, and how these processes interact with other mental faculties such as memory and attention. Consequently, psycholinguists (and neurolinguists) do not always need to conceive of language in the same way as linguists, or to focalize the concepts and distinctions that linguistics draws. But, as experience from dinner-table conversations suggests, it is hard to explain to non-specialists how these two disciplines differ in principle, in that theories of how language is processed in the mind should ultimately be informed by theories of what language is, and vice versa. Indeed, many researchers on both sides would concur that their goal is to characterize linguistic competence in a psychologically and neurologically plausible way (see Ferreira, ; Marantz, ; among others). In many ways, the differences between linguistics and psycholinguistics could be seen as differences in perspective rather than in the ultimate goals of the enterprise. The recent ascent of ‘experimental linguistics’ and its sub-domains of ‘experimental semantics and pragmatics’ not only serves linguistics by providing relevant data, but also bridges a perceived gap between linguists and cognitive (neuro)psychologists (see Townsend & Bever, ; Pylkkänen et al., ). The emergence of experimental linguistics, as a trend in academic enquiry, has been triggered by various different questions within different fields of linguistics. In the area of syntax, the debate about gradience in acceptability judgements appears to have been especially influential in stimulating experimental work (Bard et al., ; Cowart, ). In semantics and pragmatics, the crucial debate that triggered a major outgrowth of experimental research was that concerning the derivation and distribution of scalar implicatures (see Noveck & Reboul, , for an overview of a decade’s research on this topic). However, this explosion in experimental research should not cause us to neglect the long line of theoretically motivated experimental investigations into semantic and pragmatic
topics that went before. Studies on concepts and word meaning, reference, speech acts, metaphor, irony, and figurative language comprehension have produced a wealth of findings for theoretical linguists. For example, in addition to the aforementioned work on concepts and on language use, research by Gibbs and Ortony and colleagues on indirect requests and metaphors (see Gibbs, , ; Ortony et al., ; among others) has produced evidence against the pervasive literal-first approach to meaning and ultimately against a distinction between literal and figurative language. We might ask, then, why the debate on scalar implicature (e.g. the inference from ‘some of the apples are red’ to ‘not all of the apples are red’) came to be such a seminal event in the development of experimental semantics and pragmatics. Several aspects of the phenomenon make it particularly amenable to experimental investigation. First, while there is widespread consensus about the interpretation of scalar terms in most cases, competing theories offer different accounts of the mechanism that is employed in the process of deriving implicatures: whether it is the grammar, pragmatics (as standardly construed), a specialized default-pragmatic system, or some combination of these. Second, there is little agreement about the interpretation of scalar terms in cases where they are embedded under other logical operators, or on whether this involves the same interpretative process as used in unembedded cases. The theoretical positions on this matter delineate two broad classes of approaches: those that postulate that pragmatic inferences operate at a ‘global’ post-propositional level, and those that postulate that the grammar is involved in deriving these inferences, and hence the inference can appear ‘locally’, embedded under other operators. The cluster of predictions that can differentiate between competing accounts is too fine-grained to be reliably adjudicated by the traditional tools of the armchair linguist, reflective introspection and intuition. In cases like this, some of the critical predictions are not just amenable to experimentation, but perhaps can only be addressed through experimentation. An additional factor that has sustained interest in experiments on scalar implicature is the centrality of this particular debate to questions about the semantic-pragmatic interface and the language system as a whole. Issues around the nature of implicature at once have repercussions for the role of the grammar and the lexicon, and for critical topics in psycholinguistics such as context in sentence processing, perspective-taking, speaker intentions, working memory, executive functions, and the overall architecture of the parser. The breadth of experimental perspectives from which scalar implicature has been studied attests to its potential importance for the (neuro)psychology of language: these include language acquisition and off-line processing by neurotypical populations (Noveck, ) and by populations with neuro-developmental disorders such as Autism (Pijnacker et al., ) or Specific Language Impairment (Katsos et al., ), on-line language processing (Huang & Snedeker, a,b), and neuroimaging, such as Event-Related Potentials (ERPs) (Nieuwland et al., ) and Functional Magnetic Resonance Imaging (fMRI) (Shetreet et al., ). However, even as that research continues to accumulate, the relevance of experimental data to other questions in pragmatics has become increasingly evident. This is strikingly true of presupposition, where experimental methods have been used to address open theoretical questions about the classification of presuppositions and their triggers, and the circumstances under which presupposed content projects or fails to project from under the scope of operators such as negation (Smith & Hall, ; Xue & Onea, ;
Cummins et al., ). Similarly, a substantial body of empirical work has addressed the interactions between operators, for instance between quantifiers and negation, as explored in the debate on isomorphism (Musolino et al., ; Gualmini, ). Once again, the resolution of these theoretical questions has potentially profound implications for our understanding of language in general, particularly with respect to the interfaces between semantics, pragmatics, and syntax. Selecting the topics within experimental semantics and pragmatics to be represented in the Handbook proved to be one of the most enjoyable aspects of the editing process. First of all, we had to decide which phenomena to include, as well as which methods fell within the scope of the project, and then to consider the structure of the chapters. Starting with the latter, we wanted the chapters each to be about distinct linguistic phenomena or debates, and not about specific research methods. This would allow the authors freedom to discuss what we can learn about a given topic using all relevant methods. Second, we wanted to capture the variety and range of debates that are taking place in the field, rather than to represent the full extent of the depth of investigations in a few topics. While the latter would certainly also be desirable, our priority in this volume is to showcase the wealth of topics that are amenable to experimental investigation. An interesting question was what precisely constitutes experimental research. On the one hand, a reviewer pointed out that some may consider introspective research and oneon-one elicitation as (mini-)experiments, with few items, few speakers (the researcher and her colleagues), and few or no fillers. On the other hand, another reviewer asked if we would include chapters on computational linguistics and especially modelling, corpus linguistics, clinical linguistics, or developmental linguistics. Here we have taken a broad view of experimental research, taking it to encompass all quantitative methods where inferential statistics are required for the interpretation of the data and where the data is not ultimately sourced from trained linguists themselves. However, we opted not to have dedicated chapters on broad methodological approaches, such as computational linguistics, preferring to structure the chapters with reference to linguistic phenomena rather than methods. Ultimately, we made choices made about which chapter titles to propose based upon our own experience, the advice of colleagues, and the programmes of recent conferences. We also received several suggestions from the reviewers of the proposal. In the end, we believe that the thirty-one chapters presented here represent a fair reflection of the direction of experimental research in semantics and pragmatics over the past two decades. The coverage is admittedly not exhaustive: a couple of topics have been omitted due to the unavailability of expert authors in the time available, although in some cases these topics are at least touched upon in other chapters. Selecting topics for the Handbook has been a fascinating experience, reinforcing our views about the impressive scope (and potential) of experimental work in semantics and pragmatics. Deciding upon a structure for the Handbook that incorporates these chapters, however, has presented a considerable challenge. Our initial plan was to divide the chapters first into the categories of semantics and pragmatics. However, for many phenomena, there is little consensus on whether they fall under the domain of semantics or pragmatics or both. In fact, this very question is articulated by many of our authors. Thus, we felt that readers would not benefit from our attempts to impose this kind of classification, and that we should select something more appropriate.
While handbooks in general are not meant to be read from cover to cover, we wanted to present the chapters in as coherent an order as possible, taking into account their conceptual relations to one another. In the end we decided to list chapters alphabetically within loosely defined themes. • Chapters –, on Alternatives (Skordos & Barner), Constraints (Degen & Tanenhaus), and Scalar implicatures (Breheny), discuss architectural issues in the processing of pragmatic phenomena and the semantics/pragmatics interface with reference to implicature. • Chapters – cover major topics in the field that are free-standing and not integrally connected with other chapters in the Handbook, namely Events (Chen & Husband), Presupposition (Schwarz), and Spatial terms (Grigoroglou & Papafragou). • Chapters – review research on quantifiers and operators: Counterfactuals (Ferguson), Distributivity (Syrett), Genericity (Lazaridou-Chatzigoga), Modified numerals (Nouwen et al.), Negation (Tian & Breheny), Plurality (Tieu & Romoli), Processing quantification (Brasoveanu & Dotlačil), and Quantifier spreading (Brooks & Parshina). • Chapters – are about resolving lexical ambiguities in a broad sense: Adjective meaning and scales (Solt), Irony (Spotorno & Noveck), Metaphor (Pouscoulous & Dulcinati), Metonymy (Schumacher), Vagueness (Alxatib & Sauerland), Verbal uncertainty (Juanchich et al.), and Word senses (Rabagliati & Srinivasan). • Chapters – are about disambiguation and licensing in discourse contexts: Antecedent-contained deletion (Syrett), Exhaustivity of clefts (Onea), Focus (Kim), Negative Polarity Items (Xiang), Pronouns (Rohde), Reference and informativeness (Davies & Arnold), and Prosody (Tonhauser). • Chapters – are concerned with interactional factors: Politeness (Holtgraves), Theory of Mind (Rubio-Fernández), and Turn-taking (De Ruiter). It will be immediately clear that these loosely defined themes reflect just one way of capturing the various interconnections between chapters. We chose not to mark these themes in the table of contents and rather leave these interconnections implicit, as an invitation for the readers to explore for themselves some of the other ways in which the chapters are related to one another. As regards methods, many chapters review data elicited from language acquisition studies and sentence processing, including reaction times and eye-tracking data. Some chapters review data using neuroscientific methods (for instance, Schumacher and Xiang). Some refer to corpus data (such as De Ruiter) or to computational modelling (such as Degen & Tanenhaus). While the vast majority of chapters reported data from neurotypical child or adult populations speaking or comprehending their dominant language, some also appeal to data from second language acquisition (e.g. Lazaridou-Chatzigoga) and from developmental disorders such as Autism Spectrum Disorder (ASD) (e.g Spotorno & Noveck). The chapter keywords often indicate some of the methods on which the chapter focuses, but a reader who is interested in examining the full breadth of a given experimental approach, in terms of its application to semantic and pragmatic research, is advised to make use of the relevant index entries. This volume thus strives to offer broad coverage of its field both thematically and methodologically. The idea is that the reader who picks up this volume will have a
state-of-the-art overview of the research in the topics that each chapter focuses on, but also an impression of why and how a particular experimental method is suitable for addressing a given theoretical debate. In this respect, this is intended to be a forward-looking overview of past and current research, an overview which actively seeks to identify questions that could be fruitfully combined in future experimental work. The intended readership of this volume comprises advanced undergraduate students, graduate students, and researchers in the areas of linguistics, psycholinguistics, computational linguistics, cognitive science, and neuroscience who have an interest in language meaning and use. The contributing authors were asked to make the chapters accessible to such an audience, and we believe they have written with clarity and structure while also addressing the issues at the cutting edge of research in their topics. The chapters can therefore be used as part of courses, for students who have attended introductory courses in semantics, pragmatics, and/or research methods in the psychology of language. The substantial and increasing volume of work on experimental semantics and pragmatics consists almost exclusively of primary research as reported in journal articles, books, or chapters in edited volumes. To the best of our knowledge, there is no handbook or any other comprehensive work whose aim is to showcase best practice in the field. The closest things to a handbook up to this point are perhaps the edited volumes from the proceedings of conferences in experimental semantics and pragmatics (Noveck & Sperber, ; Sauerland & Yatsushiro, ; Meibauer & Steinbach, ). Fortunately, at the time of writing, the first textbook in this field is in press, Experimental Pragmatics: The Making of a Cognitive Science by Ira Noveck, to be published by Cambridge University Press. Hopefully, this and other textbooks and introductory chapters and articles will serve as a first point of contact with experimental research in semantics and pragmatics for a wide audience, with the present Handbook serving as the go-to resource for those already conducting advanced studies or research in related fields.
......................................................................................................................
, , ......................................................................................................................
H have an astonishing ability to comprehend language. Comprehension begins when we perceive a continuous perceptual signal produced by another speaker, assign this signal a phonological form, and parse it into words. These words are arranged into a hierarchical syntactic structure and logical form which together permit the listener to recover the compositional semantics of the expression. The speed and efficiency with which we achieve these steps, with little apparent effort, belies the richness and complexity of the underlying system. However, as complex as these steps may be, our understanding of them far outstrips our current insights into the final step in language comprehension, in which we infer the speaker’s intended meaning from the utterance and the context in which it is used. In making this inference, we complete a direct bridge between the speaker’s mental states and our own. Each of these different steps of language comprehension involves a series of informed guesses based on sparse data. The sound signal is generally impoverished, and surface forms often admit multiple syntactic analyses. Similarly, the intentions of speakers are not always transparently reflected by the words that they choose. Instead, the literal meaning of an utterance is often consistent with a large space of possible intentional states, requiring the listener to make rapid abductive guesses about the intentions that best explain the speaker’s choice of words in a particular context. Consider, for example, a situation in which you walk into the kitchen area of your office building and overhear part of a discussion consisting of utterance () below. ()
‘I ate some of the cookies.’
This utterance might have several different intended meanings all derived inferentially from the same underlying semantic content, depending on what the hearer thinks the speaker intends to communicate. For instance, had you walked in a few moments earlier
you might have heard the other communicative partner ask a question that prompted the utterance. Depending on the question asked, the assertion of eating some of the cookies could take on multiple different meanings as in ()–(): ()
a. ‘Did you eat all of the cookies?’ b. ‘I ate some of the cookies.’ (I did not eat all of the cookies).
()
a. ‘Did you try the cookies and cake I made for you?’ b. ‘I ate some of the cookies.’ (I did not try the cake).
()
a. ‘Did you and Jane try the cookies?’ b. ‘I ate some of the cookies.’ (Jane did not eat any cookies).
How do humans make these inferences so readily? How do they restrict the hypothesis space of possible meanings in order to converge on the meanings that speakers have in mind? In the present chapter, we discuss this problem, and the fact that the specific inference derived by a listener in a context like this depends critically on the alternative utterances that he believes to be available to the speaker. These alternatives, in turn, depend on the ‘Question Under Discussion’ (QUD)—for example, which of the utterances in () through () best represent the question that the speaker intended to address with her utterance (Roberts, ). For instance, given the question in (a), the alternatives relevant to the conversation include the statement in (b), but not the alternatives in (b) or (b). Likewise, given the question in (a), the alternative statement in (b) becomes a relevant alternative, whereas those in (b) and (b) do not. Still, even given this framework, the question remains of how these alternatives are generated by the listener, and how listeners identify which alternatives are relevant to which questions. The inferences described in examples () through () constitute different examples of conversational implicature. Among these, the example in (b) is a classic example of what linguists call ‘scalar implicature’, which typically arises when an utterance contains a scalar term like some that is enriched via the negation of a stronger, alternative, utterance that contains a scale mate like all (e.g. Horn, ; Grice, ; Hirschberg, ). This stronger statement can be said to logically entail the original utterance by virtue of being ‘informationally stronger’—in other words, because the truth conditions of the original utterance are necessarily satisfied whenever the truth conditions of the stronger statement are satisfied. For instance, if it is true that ‘I ate all of the cookies’, it is necessarily true that ‘I ate some of the cookies’. Assuming Grice’s maxim of Quantity (Grice, ), which states that communicative partners are expected to be as informative as reasonably expected, the use of the weaker alternative containing some is taken to implicate that the speaker does not believe the stronger utterance containing all to be true, resulting in the inference ‘some but not all’. In this chapter, we first review studies of communicative inference in language comprehension and then focus on the development of these inferences in children, whose non-adult-like behaviours afford a window into the interaction between semantics and pragmatics. In particular, we will explore evidence from children’s difficulties with scalar implicature and Piagetian class inclusion errors, and argue that although children’s behaviour in these cases has been accounted for independently in the past, it is possible to provide a unified account of both sets of phenomena. We will argue that children’s logical
, ,
reasoning is limited by their ability to spontaneously access and reason about alternative linguistic forms and states of affairs.
. T
.................................................................................................................................. Inference in language comprehension has long been studied in adults, beginning with Grice (, ) and his proposal that communication is a cooperative endeavour between partners, mediated by common sense rules (maxims) that govern conversation. Specifically, Grice proposed that speakers should be truthful, clear, and relevant, as well as informative, by providing as much information as needed in an utterance but no more. Following Grice, subsequent approaches have reinterpreted his intuitions, resulting in a variety of distinct theoretical frameworks. First, one class of theories has attempted to reduce conversational implicature to an optimization process between the expected cognitive ‘gain’ of deriving an inference and the cognitive ‘cost’ incurred in the process of doing so (Sperber & Wilson, /; Carston, ; Noveck & Sperber, ). This approach is part of a ‘contextualist’ theory of inference that generally views inference as effortful, distinct from grammatical processes, and dependent on context. A second class of theories, dubbed neo-Gricean, is best exemplified by Horn (, ), Gazdar (), and Levinson (, a), and was later further developed by Sauerland (), Van Rooij & Schulz (), and others. This approach aims, by providing a formal characterization of Grice, to simplify the maxims in order to preclude inconsistencies or clashes between them, but also to explain the near automaticity of certain inferences like scalar implicatures. To do so, they assume the lexicalization of inferential components like the scalar alternatives (Horn, ), and, in some cases, the generation of scalar inferences by default (Levinson, a). Finally, a third class of theory, known as ‘grammatical’ accounts of scalar inference, treats some forms of communicative inference, including scalar implicature, as purely grammatical in nature. On these accounts, implicatures are triggered by a silent logical operator, akin to the English word only, which enforces exhaustification grammatically rather than pragmatically (Chierchia, ; Fox, ; Spector, ; Chierchia et al., ). In efforts to differentiate these accounts empirically, studies of mature, adult speakers have probed the time-course of implicature computation (Noveck & Posada, ; Bott & Noveck, ; Breheny et al., ; De Neys & Schaeken, ; Chevallier et al., ; Bonnefon et al., ; Huang & Snedeker, b, ; Grodner et al., ; Bott et al., ; Bergen & Grodner, ; Tomlinson et al., ; Breheny et al., ; Chemla & Bott, ). A preponderance of evidence supports the idea that deriving these implicatures requires time (and possibly effort) beyond what is involved in computing the core, semantic, meaning of an utterance. However, there is little agreement among researchers about the exact timeline and processes involved, and how these facts adjudicate between different theoretical views. More recent accounts (Degen & Tanenhaus, , ), as well as neurolinguistic evidence (Politzer-Ahles & Gwilliams, ; Spychalska et al., ), call for more nuanced approaches, where lexical, syntactic, and contextual factors must be integrated into a model of implicature (see Degen & Tanenhaus, Chapter in this volume,
for a review and discussion). Still, almost all theories can currently accommodate this range of facts, with minor modifications. Critically, scalar implicature is not the only test case for exploring how individuals represent and reason about linguistic alternatives in the service of pragmatic inference. For instance, formal semanticists have constructed theories of how language users comprehend sentential focus, which involves reasoning over unspoken alternatives, triggered in context by a focus element (Rooth, , ; cf. Kim, Chapter in this volume). For example, if one utters ‘Mary came to the party’, sentential focus on ‘Mary’ leads the hearer to consider alternative utterances that represent propositions about implicated individuals who did not actually come (e.g. John). In more recent work (Aloni, ; Simons, ; Alonso-Ovalle, ; cf. Rooth, for discussion), this type of analysis is extended to a much broader set of phenomena such as disjunction and conditionals, and prior work has made similar arguments for weak negative polarity items like any (Krifka, a). Finally, Chierchia () extends an analysis based on reasoning with alternatives to all polarityrelated phenomena, including strong negative polarity items and free choice inferences. Formal accounts of focus and topic and how these grammatical elements trigger the consideration of alternatives also share an affinity with the notion of the QUD (Roberts, , a). QUD accounts of pragmatics (Stalnaker, ; Roberts, , a), describe discourse as proceeding on the basis of conversational goals, primary among which is to discover the ‘current state of affairs’ with regards to what is being discussed, in other words, the topic of the conversation (cf. Hulsey et al., , also Conroy et al., , for discussion). To discover this information, communicative partners posit and answer explicit and implicit questions and sub-questions relevant to the hypothesized topic. An utterance is considered ‘relevant’ to the QUD if it provides an answer (full or partial) to it (Roberts, , a; Van Rooij & Schulz, ). It is not difficult to see how a QUD account of inference also requires reasoning with alternative propositions. For instance, if the QUD is ‘What did you eat?’, the utterance ‘I ate some of the cookies’ is likely to trigger consideration of alternatives that involve other foods, such as ‘cake’, ‘sandwich’, etc. If, however, the QUD is ‘Who ate the cookies?’ the same utterance would likely trigger consideration of alternatives that involve people, such as ‘John’, ‘Mary’, ‘Jane’, and so on. Although these accounts can be couched in the language of a Gricean theory of communication, it is important to note that, as Russell () points out, virtually all formal accounts of reasoning with alternatives, including theories linked to the QUD (Gazdar, ; Groenendijk & Stokhof, ; Sauerland, ; Van Rooij & Schulz, ), as well as those related to focus and ‘alternative semantics’ (Rooth, , ), do not actually require Gricean reasoning. Rather, the formal objects posited within these accounts—for example, exhaustivity operators—can also be thought of as grammatical representations that are independent of any epistemic reasoning that speakers might engage in during conversation. Still, some recent formal approaches motivated by computational linguistics and Bayesian inference incorporate epistemic considerations. These belong to the recently revived area of ‘probabilistic pragmatics’, in which hearers make sense of the speaker’s intended meaning by making probabilistic inferences that take into account alternative utterances the speaker could have used instead (Franke, , ; Frank et al., ; Jäger, ; Bergen et al., ; Degen et al., ; Goodman & Stuhlmüller, ; Bergen & Goodman, ; Bergen et al., ). According to such approaches,
, ,
language comprehension can be generally seen as a joint inference over (a) a set of linguistic expressions that could have been used, for instance ‘some of the cookies’ vs. ‘all of the cookies’ and (b) a set of possible meanings or intentional states that the expressions might convey, for example, ‘at least some of the cookies’ vs. ‘some but not all of the cookies’ (Frank & Goodman, ). Investigations into how adults make conversational inferences have undoubtedly yielded valuable insights into language processing and comprehension. That said, existing theories are often difficult to differentiate given adult behaviours, such that current debates focus on an increasingly narrow set of phenomena, sometimes disputing millisecond differences in processing of inferences as evidence for one account or another (Breheny et al., ; Huang & Snedeker, b; Grodner et al., ; Bott et al., ; among others). For this reason, researchers have also turned to populations other than typically developed adults, including individuals with linguistic and pragmatic deficits (Noveck et al., ; Loukusa & Moilanen, ; Chevallier et al., , Hochstein et al., ). One such population, which we explore here, is typically developing children, who often exhibit semantic competence, yet nevertheless fail to make certain forms of pragmatic inference, under a restricted range of circumstances, thus providing unique insight into the interface between semantics and pragmatics.
. D : ‘’
.................................................................................................................................. Despite appearing generally adult-like in their phonology, syntax, and semantics (Gleitman & Newport, ), older preschoolers and kindergarteners often have difficulties with inferences involving quantifiers (e.g. some, every) and logical operators like or and the conditional if . . . then. Inhelder & Piaget (a) first noticed such non-adult-like behaviours in simple reasoning tasks involving universal quantification, which they dubbed ‘class inclusion’. In a classic example, children were presented with a set of shapes consisting of red squares, as well as red and green circles. When they were asked, ‘Are all of the squares red?’ they often responded ‘No’, which led Inhelder & Piaget to conclude that children’s logical representations were not fully developed. Interestingly enough, some years later, researchers began noticing that children, if anything, responded more logically than adults in tasks that involved quantification and which typically elicited scalar inferences like those described earlier. Smith () for instance, pointed out that preschoolers tended to interpret some as ‘some and possibly all’, consistent with the logical interpretation of the existential quantifier. Paris (), along with Braine & Rumain (), showed that even kindergarteners preferred an inclusive/logical interpretation of disjunction as ‘p or q or both’ as opposed to the exclusive interpretation ‘either p or q, but not both’ preferred by adults. These early findings were confirmed and extended to modals such as might and must by Noveck (), who also highlighted their relevance to the study of inference in language comprehension. Children’s pattern of difficulties with these inferences is puzzling. Whereas they appear to exhibit logical failures when interpreting universal quantifiers, they appear to
be hyper-logical in other cases, like disjunction, or when interpreting existentially quantified statements. In these latter cases, children’s behaviour has been interpreted as a pragmatic failure precipitated by difficulties in Gricean epistemic reasoning. This approach, however, seems at odds with a wealth of recent evidence that much younger children have relatively sophisticated abilities to represent the goals and beliefs of other people. For instance, by months of age, infants engage in joint attention to monitor the interests of speakers as they speak (Corkum & Moore, ) and by months they can employ joint attention systematically when acquiring and interpreting language (Tomasello, ). Both - and -year-olds show evidence of grasping communicative intentions in non-linguistic tasks (Behne et al., ; cf. Tomasello et al., ), and -month-olds seem able to consider the epistemic state of others when trying to help them find items in a nonlinguistic game (Liszkowski et al., ). More generally, infants have been shown to understand the goals and beliefs of others in a wide variety of tasks (Meltzoff, ; Woodward & Markman, ; Gergely et al., ; Onishi & Baillargeon, ; Southgate et al., ; Buttelmann et al., ). Currently, there is no unified account of this pattern of behaviours, wherein children appear to be impaired or sophisticated logically, across different cases, or pragmatically, in other cases. Instead, for each set of phenomena, a different, isolated, account has been proposed, resulting in a disjointed portrait of inferential development in children. In section ., we propose that a unified account may in fact be possible, given recent developments in the study of pragmatic development, and in particular the case study of scalar implicature. We explore one proposal for why children struggle with implicature— namely the ‘access to alternatives’ account—and propose that this idea may also be profitably extended to other cases of logical reasoning, to explain why children struggle with universal quantification and the interpretation of conditional statements, while nevertheless having relatively strong epistemic reasoning abilities.
. S
.................................................................................................................................. As mentioned earlier, children struggle to compute scalar implicatures with the quantifiers some vs. all (Smith, ; Noveck, ; Papafragou & Musolino, ; Feeney et al., ; Guasti et al., ; Katsos & Bishop, ; Foppolo et al., ; Skordos & Papafragou, ). Moreover, their difficulties extend to other cases of scalar inference, including the singularplural distinction (Barner et al., ), disjunction (Paris, ; Braine & Rumain, ; Chierchia, Crain, Guasti, & Thornton, ; Hochstein et al., ; Singh et al., ; Tieu et al., ), and modals like might and must (Noveck, ; Ozturk & Papafragou, ). Unlike Piaget, most researchers have concluded that children have largely adult-like mental representations of the relevant lexical items and the necessary logical and syntactic competence, but lack other knowledge or abilities required to respond in an adult-like manner (although see Horowitz et al., in press). Most early accounts, for example, ascribed children’s difficulties to domain-general processing limits (Chierchia, Crain, Guasti, & Thornton, ; Reinhart, ; Pouscoulous et al., ), while other more recent studies attributed children’s apparent ‘failures’ to their reluctance to reject infelicitous statements that were otherwise logically true (Katsos & Bishop, ).
, ,
A more recent account argues that children’s problems with scalar inference stem from difficulties accessing relevant linguistic alternatives that are required to derive implicatures (Chierchia et al., ; Barner & Bachrach, ; Barner et al., ; Hochstein et al., ; Tieu et al., ; Skordos & Papafragou, ). As we saw previously, to derive the pragmatically enriched inference, ‘I ate some, but not all of the cookies’, from the utterance, ‘I ate some of the cookies’, one needs to consider and negate the relevant alternative, ‘I ate all of the cookies’. If children cannot access the relevant alternative during the computation, the inference cannot be made and children are left with the basic semantic content of the original proposition. In a test of this ‘access-to-alternatives’ account, a study by Barner et al. () showed -year-olds simple pictures and asked them questions that required scalar reasoning. In one condition, children were presented with a scene—for example, three animals including a dog, a cat, and a cow—and all three were sleeping. On critical trials, they were then asked either whether the cat and the dog were sleeping, or, alternatively, whether only the cat and the dog were sleeping. In this context, they answered Yes to the first question per cent of the time, since it was indeed the case that the cat and the dog were sleeping. Critically, they answered Yes only per cent of the time in response to the second question, which included the word only, since it was false that only the cat and the dog were sleeping. In stark contrast, children failed to differentiate trials that included only from those that did not when questions involved the scalar quantifier some. Here, on critical trials, where all animals were sleeping, children responded Yes equally often when asked whether some of the animals were sleeping and when asked whether only some of the animals were sleeping (in each case around % of the time). Barner et al. () reasoned that the best explanation of this pattern was that children failed to exhaustify only some utterances because they required access to stronger alternatives including the quantifier all, which was not contextually available. Crucially, they succeeded with the only a cat and a dog cases—which they called ‘ad hoc’ scales, because the relevant stronger alternative—in other words, the cat, the dog, and the cow are sleeping—could be constructed on the fly from the contextually available images. In this account it is access to alternatives, rather than a blanket pragmatic deficit or processing limitations, that explains children’s classic failures with some/all implicatures. In support of this, subsequent studies have reported that children successfully compute scalar implicatures if the alternatives are provided contextually, as is the case with so-called ad hoc implicatures (Hirschberg, ). First, in a study by Stiller et al. (), - and -yearolds were shown simple scenes consisting of three possible referents in the form of smiley faces. One of the smileys was bare, a second had glasses, and the third had both glasses and a hat. When told that the experimenter’s ‘friend’ had glasses and asked to guess who that friend was, even -year-olds were successful in picking out the smiley that had glasses but no hat, arguably because they were able to derive a scalar implicature from the contextual scale . In another study, Barner et al. (), presented - and -year-old children with two puppets, one who was described as silly, and who often said strange things, and the other who was described as smart. They then watched as a bear approached a pair of objects—for example, an orange and a banana—and either took one, or both of them. At this point, the experimenter told the child that one of the two puppets— the silly or the smart one—had described what had happened, and that they had to guess who it was that had provided the description. On critical trials, when both objects were taken, and the children were told, ‘Somebody said that the bear took an orange’, and were
asked to pick who had said it, they correctly chose the silly puppet, and did so to the same degree that they chose the appropriate puppet on control trials. Critically, in a different study using an identical design, but testing disjunction, children did not succeed. When shown a bear who took both an orange and a banana and told that ‘Somebody said the bear took an orange or a banana’ children systematically chose the smart puppet, despite choosing the appropriate puppet on control trials (Hochstein et al., ). Together, these results suggest that children can compute scalar implicatures when alternatives can be recovered contextually, as in the case of ad hoc scales, but not in cases like disjunction, which requires access to a stronger alternative involving conjunction that are not contextually available (e.g. The bear took the orange and the banana). In the Stiller et al. () study, as well as in ad hoc conditions of Barner et al. () and Barner et al. (), young children were successful in deriving scalar inferences when the necessary alternatives were made available to them. How exactly alternatives help children, however, was left open by these studies. One possibility is that children struggle with implicature because they lack strong associations between members of scales—for example, that in adults, scalar items like some and all become strongly associated over time, such that hearing some automatically triggers access to all. Also, relatively weak knowledge of scales might interact with processing difficulties—children with strong working memory capacities might be able to compensate for weak knowledge of scales in order to access scale mates, while children with weaker processing capacities might be most impaired by weak knowledge of scales. However, another possibility left open by these studies is that contextually available alternatives help children by hinting at the intended QUD: Children may fail to compute implicatures not because they lack the ability to access alternatives in general, but because they don’t know which alternatives are contextually relevant—in other words, they struggle to identify the QUD. Recall that if a speaker utters, ‘I ate some of the cookies’, they might intend to convey, ‘I ate some, but not all of the cookies’, or, ‘I ate some of the cookies, but no cake’, or even, ‘I ate some of the cookies, but Jane didn’t eat any’. Crucially, which of the logically possible inferences matches the speaker-intended meaning, is restricted by the alternatives that one considers (, , ). To address this idea, Skordos & Papafragou () conducted a set of studies with -yearolds. They used an acceptability judgement task (cf. Papafragou & Musolino, ), in which children were presented with simple visual scenes featuring novel creatures (‘blickets’) who possessed common everyday items. A puppet described the scenes with a statement featuring the quantifier some or all and children were asked to judge the puppet’s utterances. Statements with all were true or false and statements with some always true, but sometimes infelicitous (critical trials). In a first experiment, the authors tested whether the availability of the stronger scalar alternative all could help children derive scalar implicatures when judging utterances containing the weak alternative some. Indeed, when trials with all and trials with some were intermixed in the experimental battery so that all was contextually available, children derived scalar implicatures at rates comparable to adults. When, however, the trails with some preceded the trials with all in a blocked design (such that they never heard all before the some trials were completed), children’s performance dropped significantly. Having established that priming stronger alternatives helps children, Skordos & Papafragou () conducted a second experiment to further investigate why.
, ,
Their hypothesis was that priming alternatives helps by making it easier for children to understand the ‘conversational goal’ of the communicative exchange, a Gricean notion analogous to the idea of the QUD. In this experiment, test trials with the quantifier all always preceded trials with some so that the stronger scalar alternative would always be accessible to children. Critically, in one condition, the conversational goal (QUD) was about whether all or only some of the ‘blickets’ had an item that the puppet mentioned. This was done by manipulating the false all-trials so that the items the ‘blickets’ had were always those mentioned by the puppet and instead the truth of the statement hinged on the number of ‘blickets’ having an item (/ or /). This QUD was expected to prime the some-but-not-all scalar inference. In another condition, the test trials highlighted a different conversational goal (QUD), namely whether the ‘blickets’ had the item mentioned, or a different one instead. In this case, all four of the blickets always had a certain type of item (so no blicket was left without one) and the truth of the trials hinged on whether the ‘blickets’ possessed the item the puppet had mentioned (e.g. a crayon), or a different one (e.g. an umbrella), such that the QUD concerned the kinds of things that the blickets possessed, instead of how many blickets possessed an item. Skordos & Papafragou () found that despite access to the stronger scalar alternative all across the two conditions, children’s judgements were critically affected by the QUD. In the condition highlighting the ‘quantity’ QUD and priming the some-but-not-all scalar inference, children overwhelmingly rejected the statement ‘some of the blickets have a crayon’, when all four blickets had a crayon. When, however, the QUD was related to the type of item and consequently primed the crayon-but-not-other-item inference, children generally accepted the very same statement as perfectly fine. Even more strikingly, in a third experiment, children derived the some-but-not-all inference, even when the experimenters did not make the stronger scalar alternative all accessible, so long as the QUD primed a scalar inference, for example with the quantifier none in the place of all. Since making none accessible does not automatically provide access to all and since children nevertheless compute implicatures that require accessing all, children’s difficulty is likely not with the strength of associations between scale mates, but instead lies with their problems identifying the QUD and the inference to be made. Based on these findings, Skordos & Papafragou () concluded that children’s primary difficulty with scalar inferences is that they struggle to identify the conversational goal of these communicative exchanges, which can be straightforwardly translated as difficulty identifying the QUD. Together, these findings raise the possibility that when relevant alternatives are identifiable, children use them in a way similar to adults to make inferences. If so, then contrary to Piaget and some subsequent accounts in linguistics, the logical and computational abilities required by inferential reasoning may be in place by the preschool years (Barner et al., ; Stiller et al., ). Also, the results suggest that, like adults, children can use contextually accessible alternatives to infer the QUD and consequently the inference they should make among several possible inferences. In contrast to adults, however, children find it more challenging to spontaneously recover the QUD, perhaps because the hypothesis space of possible inferences is too large for them to restrict without help, or perhaps because their ordering of likely hypotheses differs from that of adults (Pierce, ; Schulz, ). When the hypothesis space is restricted as in Skordos & Papafragou (), they do much better, including spontaneously generating the relevant alternatives.
. Q
.................................................................................................................................. Consistent with the hypothesis that children’s inferences are limited by their ability to identify relevant alternative utterances, there is evidence that access to alternatives may also restrict children’s ability to make a range of additional logical inferences with quantifiers that are not pragmatic in nature, including the class inclusion examples originally studied by Inhelder & Piaget (a). According to some accounts, including that of Inhelder & Piaget, children’s difficulties with universal quantification reflect deficits in logical reasoning. Others, however, have argued that some form of pragmatic deficit may best explain children’s class inclusion judgements (e.g. Crain et al., ). Extending the original Piagetian findings, a study by Philip () showed children a scene in which three girls were each riding an elephant in the presence of a fourth, riderless, elephant. Surprisingly, when children were asked, ‘Is every girl riding an elephant?’ they responded ‘No’, sometimes explaining that the fourth elephant did not have a rider. According to some accounts, it was argued that such judgements arise because children interpret the universal quantifier every as applying at the sentence level, its scope spreading from the subject to the entire sentence including the verb object, resulting in a symmetrical interpretation equivalent to, ‘Every girl is riding an elephant, and every elephant is being ridden’. Based on this, the phenomenon was dubbed by some as ‘quantifier spreading’ (Roeper & De Villiers, ; cf. Brooks & Parshina, Chapter in this volume). Other accounts proposed that these judgements might arise because children quantify over events (e.g. instances of girls riding elephants) as opposed to entities (Philip, , ), or because children otherwise misidentify the syntactic features of the quantifier (Kang, ). Somewhat in the tradition of Piaget, others (Drozd & Van Loosbroek, ; Drozd, ) have suggested that children might have a different semantics for the universal quantifier. Errors arise because children treat every as a ‘weak’ quantifier, similar to many, and allow the quantifier to combine with an argument that can be determined contextually, sometimes switching to the verb object (Westerståhl, ). For example, consider the sentence in (): ()
Many Finns are World Rally Championship drivers. a. A large proportion of the Finnish population are WRC drivers. b. A large proportion of the WRC drivers are Finnish.
The canonical reading of () corresponds to (a), which we know to be false (it is not true that a large number of the set of Finnish citizens are WRC drivers). Yet people judge () to be true with a reading like (b), effectively allowing the quantifier many to combine with the object of the sentence as opposed to the subject. In this case, adults appear to disregard the ordinary ‘proportional’ meaning of the quantified sentence, as computed by syntax and semantics, and use their pragmatics to make an inference about what the speaker means (cf. Cohen, ), in a way not dissimilar to what children might be doing with every. Other accounts propose that children’s difficulty does not arise in semantics or syntax per se, but rather in their interface, when children attempt to map the complex semantics of quantifiers to their surface structure. For example, Geurts () argues that although children have an adult-like ‘strong’ semantics of every, processing difficulties associated
, ,
with the complexity of quantified expressions lead children to effectively treat them as weak quantifiers instead. Specifically, overwhelmed by the complexity of these sentences, children instead use context to restrict quantification, allowing the quantifier to combine with either the subject or the object of the sentence (e.g. girls or elephants), depending on which is contextually salient. Similar arguments that processing demands may impair children are relatively common in the literature (Freeman et al., ; cf. Brooks & Sekerina, /) and as we already saw appear frequently in developmental studies of scalar implicature as well. In both cases, however, no direct evidence was offered to support this possibility, and no detailed causal models were put forward to explain precisely how processing limits might cause quantifier spreading errors or scalar implicature failures.
. A
.................................................................................................................................. In recent work, we have begun to explore the idea that children’s problems with class inclusion stem from the same types of difficulties they experience with implicature. Specifically, children fail to interpret them in adult-like ways because they struggle to identify the QUD, and consequently do not access relevant alternatives that are required for interpreting the utterance. Further, the QUD can be identified in one of two ways: () via explicit mention of relevant linguistic alternatives, or () by mention of different contextual states of affairs that render specific linguistic alternative sets relevant. For example, in a context including a man, a man with sunglasses, and a man with a hat and sunglasses, then ‘the man with the sunglasses’ rules out reference to the man who is wearing both sunglasses and a hat. However, the same utterance is compatible with this man if there is no man who is wearing only sunglasses. In this context, ‘the man with sunglasses’ is compatible with the hat-wearing man, because now, due to context, the stronger statement involving a hat is no longer considered as a linguistic alternative. Such inferences can be modelled in a neoGricean epistemic framework, like the Rational Speech Act account (Frank & Goodman, ), but are also compatible with non-epistemic logics, so long as they incorporate models of both utterances and possible states of affairs / referential domains. Critically, our proposal is that by adopting this type of joint inference, the same framework which explains children’s difficulties with scalar implicature can be extended to explain quantifier spreading, too. First, consider the case of scalar implicature. In order to comprehend an utterance like ‘Some of the boys have bicycles’, one needs to consider both the alternative linguistic forms that the speaker could have uttered, but also the alternative meanings (or states of affairs) that the utterance is compatible with, which might be provided contextually. Critically, these are related. For example, in a context including a group of boys and a group of girls where all of the boys have bicycles and all of the girls have skateboards, salient alternatives to ‘Some of the boys have bicycles’ might involve skateboards (e.g. ‘Some of the boys have skateboards’). However, in a context in which some of the boys and all of the girls have bicycles, such alternatives might be less salient than an alternative such as ‘All of the boys have bicycles’. This is because in each case, the context provides a particular dimension of contrast (e.g. between kind of vehicle in the first, and number of vehicles in the
second). Similarly, the relevant alternatives can also be elicited via direct mention, making the QUD explicit—for example, if a speaker asks ‘What kinds of things do the boys have?’ or ‘Do all of the boys have bicycles?’ Context makes available different states of affairs, and therefore elicits different QUDs, resulting in the licensing of different linguistic alternatives. To see how this might work for children in the case of quantifier spreading, it is useful to consider how Crain and colleagues explain children’s judgements (Crain et al., ). According to Crain et al., children appear to have difficulties when asked to judge universally quantified sentences because the tasks used in the literature often fail to satisfy what they call the condition of ‘plausible dissent’. For a question like, ‘Is every girl riding an elephant?’ to be felicitous, the affirmative response (Yes) must have been in doubt (in the past) and not clearly incontrovertible. For this to happen, an alternative outcome (e.g. a girl riding a donkey) should be conceivable to the participant. If such an outcome is not accessible, children are left to infer that other interpretations of the question must be relevant and respond based on those instead (e.g. the numerical correspondence between girls and elephants). This explanation can be straightforwardly recast as children having difficulties identifying the QUD. In quantifier spreading cases, while adults identify the QUD as being about the number of girls riding elephants because they consider an alternative in which not all three girls are on elephants, children may fail because they cannot generate this alternative state of affairs (i.e. the possible intended meaning), leading them to instead base their judgement on another dimension, such as the mismatch between the scene and their prior expectations given the utterance—that is, there should be a set of girls riding elephants, but no extra characters, or extra animals. As evidence for the condition of ‘plausible dissent’, Crain and colleagues () show that when children are provided with a plausible alternative outcome (e.g. ‘some farmer feeds a dinosaur’ for ‘every farmer feeds a donkey’) in the form of short background stories, children’s responses become more adult-like. The notion of ‘plausible dissent’ captures an important intuition: The evaluation of statements inviting inferences in natural language requires more than merely processing the utterance and checking that it matches a semantic model. It addition, it also involves inferring the QUD and thus the dimension along which an inference should be made. Is it the number of girls riding elephants that matters, the kinds of things being ridden, or the number of elephants being ridden? Given clear contextual cues regarding a state of affairs that would make the utterance false (i.e. a basis for plausible dissent), children might identify the relevant dimension, and thus the QUD, thereby allowing them to make adultlike inferences. In support of this idea, a recent study by Skordos et al. (under review) attempted to address this question by priming different possible alternative states of affairs and their corresponding QUDs. In this study, preschoolers were shown simple cartoon pictures in which three girls were each riding an elephant (or driving a car, riding a bicycle, etc.) and an extra elephant (car, bicycle, etc.) was present in the scene. Children were asked to answer the question, for example, ‘Is every girl riding an elephant?’ In this baseline condition, children answered Yes only per cent of the time, a result consistent with prior work. In a second, experimental condition, the critical trials were preceded by scenes in which only two of the three girls were riding elephants, while the third girl was on foot and two more elephants were without riders. These trials were intended to provide children with alternative states of affairs that might indicate to children the QUD—in other words,
, ,
whether these three girls were riding elephants. When children were asked whether every girl was riding an elephant, they overwhelmingly answered No to the priming trials and subsequently answered Yes to the critical trials per cent of the time; a large and significant improvement compared to the baseline and much closer to adult performance. Still, even in this scenario appear to children differ modestly from adults, leaving open the possibility that other factors contributed to their difficulties. In the condition just described, children were assisted by being explicitly provided an alternative state of affairs that afforded a basis for plausible dissent. In the language of joint inference with which we began this section, these children were provided with a broader set of meaning hypotheses against which an utterance might be evaluated, which in this case played the role of indicating the dimension along which the speaker intended the utterance to be evaluated—in other words, whether all three girls were riding elephants. However, in the absence of such an explicitly provided alternative, children might generate their own, and might do so in ways that also contribute to their non-adult-like behaviours. One such possibility is raised by Drozd & Van Loosbroek (), who suggest that children may struggle with domain restriction. In particular, upon seeing a riderless elephant, they may infer that, since elephants in this world generally belong to girls, there must exist another girl who is not on her designated elephant, such that it is legitimate to conclude that not every girl is riding an elephant (since the missing girl is not). To test this, Skordos et al. (under review) tested children in a condition like the previous ones, but this time instead of priming children with a situation in which the utterance was false, they instead introduced children to three girls, by name, emphasizing that they would be the only three girls under discussion. At test, when shown, as before, the three girls riding four elephants, children judged that every girl was riding an elephant per cent of the time, which appears larger than the per cent of the baseline, but was not a statistically reliable difference. Consequently, we concluded that at least in this dataset, domain restriction did not play a reliable role in children’s judgements: Children did not primarily make errors because they overgenerated alternative states of affairs, imagining situations not intended by the speaker. A follow-up experiment, that combined the domain restriction cues (evidence that only three girls existed) with the false alternative cue (an example of only two girls riding) found that children performed near ceiling on the task (%), and thus almost never showed the type of error originally reported by Philip (). Together, these data suggest that a combination of plausible dissent and domain restriction can likely account for all of the variability that differentiates children from adults. Based on these recent findings, it seems plausible that the phenomenon of quantifier spreading can be given a treatment similar to that originally applied to scalar implicature. The difficulties children have when reasoning with quantified statements are in large part a result of their difficulty restricting the inferential hypothesis space in language comprehension. As suggested earlier, on this view, language comprehension involves a kind of abductive inference, wherein the listener attempts to identify the meaning of an utterance from what is initially a large (if not infinite) space of possible alternative meanings. The hypothesis space is only partially narrowed by the literal semantics of utterances, with the result that the listener must consider alternative linguistic forms that the speaker could have used but didn’t, as well as the possible states of affairs which the utterance might plausibly be taken to encode. Children’s difficulties in narrowing the inferential hypothesis
space can be mitigated by priming linguistic alternatives that are only consistent with a particular QUD, or by presenting alternative scenes that focus children’s attention to that QUD, by making a particular dimension of the scenes salient. In both the case of implicature and quantifier spreading, once children have identified the QUD, they readily narrow their inferences, access any required linguistic alternatives, and make judgements that are almost perfectly adult-like.
. C
.................................................................................................................................. In this chapter, we initially discussed the importance of pragmatic inference involving alternatives for language comprehension, highlighting that utterances are multiply ambiguous and can lead to different interpretations as a result of different inferences that can be made. We reviewed the problem of restricting this inferential hypothesis space, and presented a brief overview of theoretical and empirical work on adults that argues for the role of alternative linguistic forms and alternative meanings in restricting the hypothesis space and guiding both scalar inferences but also contrastive inferences triggered by focus elements. We then turned to developmental evidence and presented two characteristic case studies from language acquisition, scalar implicature and quantifier spreading. In both cases, children who are otherwise competent users of language appear to struggle when interpreting sentences including quantifiers. We argued that, in both scalar implicature and quantifier spreading, children’s problems are closely linked to difficulties in reducing the inferential hypothesis space, while matching what is said to what is meant. We argued that a substantial proportion of those problems is caused by the fact that children often misidentify the QUD, which leads them to consider irrelevant alternatives and make non-adult-like inferences. When relevant alternatives are made salient and the QUD is appropriately identified, children use it to make inferences in an adult-like manner. While alternatives can be used within the course of an experiment to lead children to the appropriate QUD and restrict inference, further work is needed to understand why it might be that children struggle to identify the QUD, forcing them to consider the wrong alternatives and make non-adult-like inferences.
......................................................................................................................
- ......................................................................................................................
.
. T
.................................................................................................................................. H do listeners make sense of language? Bottom-up information from the linguistic signal (whether spoken or written) is never processed in isolation. Higher level (top-down) expectations combine with bottom-up information to determine how incoming information is interpreted. The role of top-down information has become increasingly important within an emerging perspective that the brain is a predictive machine with expectations about the upcoming signal that the actual input may match to a greater or lesser extent (for a review and proposal, see Clark, ). While language processing in general addresses the question of how bottom-up information from the linguistic signal is integrated with top-down expectations generated from information about the discourse context (speaker identity, Question Under Discussion (QUD), common ground, utterance alternatives) and world knowledge, these questions take on a specific flavour in pragmatics, where the central questions focus on how context contributes to meaning. Theories of information processing often distinguish between automatic processes, which are fast, dumb, and are not resource intensive; and strategic processes, which are slow, flexible, and resource intensive (Posner & Snyder, ; Neely, ; Shiffrin & Schneider, ; see also Kahneman, , for a related framework.). Given the speed of online comprehension and dependence on literal meaning, it is often assumed that the contextdependent inferences, which are central to pragmatic processing, are strategic and, unless pre-compiled, must necessarily lag behind computation of literal meaning. In contrast we have argued that whereas unconstrained inference or inference in situations of cue conflict might be slow and costly, inference that is constrained by rich conversational context and natural use of linguistic forms can be remarkably fast and easy (Degen & Tanenhaus, ). Moreover, because language unfolds over time, just as lexical and syntactic expectations are available to predict and explain the linguistic signal, so are expectations from constrained contexts. From this perspective, it is unexpected linguistic forms in unnatural contexts that make language processing slow and costly.
.
In this chapter, we argue for a constraint-based account of pragmatic processing. We begin by describing central features of constraint-based approaches to language processing and then consider how constraint-based accounts have furthered our understanding in other domains, using syntactic ambiguity resolution as an example. We then explore the logical space of possibilities for accounts of pragmatic processing. Finally, we review one area of pragmatic processing in depth—scalar implicature—before discussing the application of constraint-based approaches to other areas of pragmatics.
. C-
.................................................................................................................................. There are four defining characteristics of constraint-based approaches. First, as an utterance unfolds, listeners rapidly integrate multiple probabilistic sources of information in a weighted manner. Second, listeners generate expectations of multiple types about the future, including the acoustic/phonetic properties of utterances, syntactic structures, referential domains, and possible speaker meanings. Third, speakers and listeners can rapidly adjust expectations to different speakers and different situations, etc. Fourth, explanations that depend upon architectural constraints (e.g. information-encapsulated modules, discrete sequential processing stages) are only considered as a last resort. Many early explanations of the speed and efficiency of language comprehension were inconsistent with a constraint-based approach. The speed of processing was attributed to encapsulated, autonomous subsystems or modules, which perform specialized computations, much like a worker on an assembly line (Forster, ; Fodor, ). Syntactic ambiguity was an important test domain: is ambiguity resolved by domainspecific processes or by appeal to multiple sources of information, including the general context of an utterance? The influential Garden Path model of syntactic parsing proposed that syntactic heuristics were privileged in processing compared to other information sources. Initial results suggested that listeners do indeed make initial commitments without taking context into account (Ferreira & Clifton, ; for a review, see Frazier, ). However, as ideas about probabilistic constraints, including those provided by context, became more refined, evidence that listeners rapidly integrate multiple constraints emerged (for proposals and reviews, see MacDonald et al., ; Tanenhaus & Trueswell, ; Jurafsky, ; Altmann, ; Gibson & Pearlmutter, ; Levy, ). Several factors were crucial for the emergence and convincing evaluation of successful constraint-based models. The relevant constraints needed to be identified and quantified. It was crucial to understand the central role of context and to develop methods that allowed language processing to be explored in well-defined contexts, which often required having clearly specified goal structures. Explicit testable linking hypotheses had to be articulated that mapped predictions of models onto dependent measures (e.g. reading times). Finally, it was important to appreciate the rich sources of information provided by the signal. In the following, we only implicitly address the state of pragmatic processing with respect to these five factors, but we include an explicit discussion at the end of the chapter.
-
. L : S ,
.................................................................................................................................. The literal content of the sentence in () is given in (a). However, an utterance of () is likely to give rise to the scalar inference in (b). Since Grice (), the accepted explanation is this: upon observing (), the listener reasons that, if the speaker was cooperative and knowledgeable, he should have produced the relevant stronger alternative in (c). Because he did not, he can be assumed to be implicating its negation, thus arriving at the inference in (b). ()
Some of the people in the White House are nuts. a. Some, and possibly all, of the people in the White House are nuts. b. Some, but not all, of the people in the White House are nuts. c. All of the people in the White House are nuts.
Questions around scalar implicature involve the regularity, probability, and strength with which the inference arises, and whether listeners necessarily pass through an informationally privileged stage of literal (or conventional) processing before integrating contextual information that results in enrichment of the literal meaning to yield the inference (or the cancellation of the inference to yield the literal meaning). In contrast to these informational privilege accounts, constraint-based accounts propose that both the probability with which the inference arises and the speed with which it is processed are functions of the contextual support for the implicature (i.e. that the speaker intended to communicate the negation of the stronger alternative).
. C-
.................................................................................................................................. Constraint-based approaches to pragmatics share the assumptions of constraint-based approaches laid out in section .. We focus here specifically on how the meaning of an utterance is processed. We do not commit ourselves to talking only about phenomena that fall solely within the domain of pragmatics; we, like the rest of the field, do not know where to draw the line between semantics and pragmatics (Szabó, ). Therefore, whenever we speak of ‘pragmatic processing’ we have in mind a listener processing information in service of inferring the speaker’s intended meaning. How are the multiple sources of information integrated in online pragmatic processing? Views of this process differ along at least two separable dimensions which span two extreme positions and many nuanced ones. One relates to when a piece of information is assumed to be processed compared to others; the other to how strongly it is weighted in the resulting interpretation. In the syntactic parsing example mentioned in section ., much of
.
the discussion focused on when contextual pragmatic information was used—that is, whether it was necessary to assume that there are distinct stages of processing. Researchers interested in the ‘when’ question typically use online measures of processing like eye movements in the visual world and, in reading, Event-Related Potentials (ERPs) and self-paced reading times. Some information about processing effort can also be gleaned from response times in truthvalue judgement tasks. Researchers interested in how strongly a piece of information enters the resulting interpretation typically use offline measures like judgement data from truthvalue judgement tasks, survey studies, and choices in interpretation selection tasks. The two extreme positions are: . Extreme informational privilege: some types of information are privileged over others. In online processing, they are processed earlier and weighted most heavily in the resulting interpretation. . Extreme parallelism: all available information is processed in parallel. In online processing, all available information is processed simultaneously and weighted equally in the resulting interpretation. Between these extremes lie other logical possibilities: for example, certain types of information may be privileged in processing but not in the resulting interpretation; or weighted more heavily in the resulting interpretation but nevertheless processed in parallel with other less heavily weighted types of information. Another intermediate position is that there is no principled default privilege of any type of information; how information is processed depends on how useful it is in the immediate context and how useful it has been in the listener’s linguistic experience. Intermediate positions that cluster close to caricature can be considered constraint-based theories. Positions that cluster close to caricature have different names depending on the type of information that is claimed to be privileged and the phenomena to which they apply. For scalar implicature, two extreme positions about time-course (close to caricature ) are the Literal-First hypothesis (discussed by Huang & Snedeker, a, ; Bott et al., ) and the Default hypothesis (Levinson, a). The Literal-First hypothesis proposes that an expression’s literal meaning is computed before pragmatic factors that lead to its enrichment. The Default hypothesis proposes that an implicature conventionally associated with an expression is automatically computed before additional pragmatic factors that lead to its cancellation to yield the literal meaning. In contrast, a constraint-based account claims that the speed and probability with which an implicature is computed is a function of the contextual support it receives (Degen & Tanenhaus, , ). The Literal-First and Default hypotheses share the assumption of informational privilege in the same way that the Garden Path model assumed that syntactic heuristics were privileged. Using the example of scalar implicature, we survey the evidence and conclude that the data are compatible with constraint-based theories but not with informational privilege theories, unless one makes ad hoc auxiliary assumptions. We begin by introducing factors that we assume are generally at play in pragmatic processing. Each may receive a different setting in a particular experimental context.1 1
We use ‘setting’ to refer to different values each factor can plausibly take on.
-
Researchers typically focus on only one factor at a time in their experiments. They rarely control for other factors, explore their effects, or speculate on their likely factor settings and how that might influence the results. Section . is intended to be both an overview of factors that affect pragmatic processing and a checklist for researchers to consider in designing experimental tasks and conditions, and in writing paper discussions.
. C
.................................................................................................................................. Constraint-based approaches are appealing in their broad coverage of phenomena yet are often criticized for being too permissive. How do we know which factors listeners are likely to be sensitive to in interpretation? How strongly should we expect these factors to matter? How can a coherent theory emerge from a framework whose main message appears to be ‘context matters’? We believe that the situation is much less dire in that, as for syntactic processing, a constraint-based approach to pragmatic processing lends itself to explicitly quantitative investigations in ways that informational privilege approaches do not. We now survey some factors that we propose are relevant for deriving any type of pragmatic inference and indeed, for processing any type of utterance, which include the QUD, world knowledge, properties of the utterance and its alternatives, properties of the speaker, and common ground. We distinguish these from more idiosyncratic inferencespecific factors, some of which are discussed in section ..
.. Question Under Discussion An important observation going back at least to Grice () is that language is interpreted with respect to a QUD (Roberts, a) or goal that listeners expect the speaker to be addressing. The QUD establishes states of the world worth distinguishing for interlocutors. It need not be explicit, and in fact often isn’t. For example, the precision with which speakers answer the overt question ‘Where are you?’ depends on the relevance of being maximally precise with respect to interlocutors’ joint goal (Potts, ). There may also be uncertainty about the QUD, which is humorously illustrated in a dialogue between Harry and Jess in the film When Harry Met Sally:2 ()
Jess: If she’s so great why aren’t YOU taking her out? Harry: How many times do I have to tell you, we’re just friends. Jess: So you’re saying she’s not that attractive. Harry: No, I told you she IS attractive. Jess: But you also said she has a good personality. Harry: She HAS a good personality.
2
This example was first made famous by Larry Horn ().
. Jess: [Stops walking, turns around, throws up hands, as if to say ‘Aha!’] Harry: What? Jess: When someone’s not that attractive they’re ALWAYS described as having a good personality. Harry: Look, if you were to ask me what does she look like and I said she has a good personality, that means she’s not attractive. But just because I happen to mention that she has a good personality, she could be either. She could be attractive with a good personality or not attractive with a good personality. Jess: So which one is she? Harry: Attractive. Jess: But not beautiful, right?
Harry’s explanation reveals that he and Jess were assuming that different QUDs were being addressed by ‘She has a good personality’, perhaps ‘What does Sally look like?’ or ‘What is Sally’s best feature?’, as opposed to ‘What are some general features of Sally?’ or ‘What is Sally’s personality like?’ Interpreting the utterance as the response to one of the former QUDs yields a very different interpretation than as a response to one of the latter. Although there are to date no good empirical measures of the QUD, researchers have had some success in manipulating the presumed implicit QUD via cover stories (Zondervan, , ; Degen, ; Degen & Goodman, ). In production tasks, manipulating the QUD has direct effects on the specificity of produced utterances (Potts, ). In visual world studies in particular, the QUD is often manipulated via the participants’ task; for example, referent identification by clicking, pointing, or performing an action makes salient the QUD ‘Which of the pictured images is the target?’ (Sedivy et al., ). The QUD is also increasingly serving as an explanation for why the prejacents of presupposition triggers project less frequently than previously assumed (Beaver et al., ); for instance, the Projection Principle formulated by Simons et al. () proposes that content only projects if it doesn’t address the QUD. While a systematic empirical exploration of this principle is still outstanding, it exemplifies how widely useful the QUD is as an explanation for listeners’ interpretation.
.. World knowledge World knowledge (i.e. prior beliefs) strongly guides utterance interpretation. For instance, the pronoun ‘they’ is generally interpreted to refer to the city council in (a) but to the demonstrators in (b) (Kehler et al., ; adapted from Winograd, b). ()
The city council denied the demonstrators a permit because a. they feared violence. b. they advocated violence.
Similarly, listeners use world knowledge about object affordances in reference resolution (Chambers et al., ) and the interpretation of declarative sentences (Hagoort et al., ).
-
.. Properties of the observed utterance and its alternatives Some aspects of the observed utterance related to the above factors are generally at play in pragmatic utterance interpretation: first and foremost, the cost and informativeness of an utterance compared to its alternatives (as estimated by the listener to be available to the speaker). The speaker’s utterance cost combines cognitive and production cost. The cost of an utterance is inherently difficult to determine and likely a function of non-independent factors (which may differ across the spoken and written modality): its frequency of occurrence, length, phonological and syntactic complexity, social value (e.g. taboo words likely have low cost in colloquial settings and high cost in formal settings), etc. How to best characterize the cost of an utterance is currently an open empirical question. While a sentence’s informativeness is often described in terms of its entailment relations compared to its alternatives, a more general notion of informativeness takes into account the QUD: what is often described as an utterance’s relevance can be interpreted as its informativeness with respect to a partition induced by the QUD (e.g. Russell, ; Degen & Goodman, ; Kao et al., ). In general, what an utterance’s alternatives are is an open question. As with the QUD, there are currently no good empirical measures for estimating alternatives. Some have proposed constraints on alternatives based on complexity (Fox & Katzir, ); others have found that explicitly manipulating the set of utterances within experimental settings changes expectations about the available alternatives and their subsequent pragmatic inferences (Degen & Tanenhaus, , ). The set of alternatives the listener assumes to be available for the speaker clearly affects the inferences drawn: Jess’s response ‘But not beautiful, right?’ indicates he assumed that ‘beautiful’ was an available alternative to ‘attractive’, and thereby drew a scalar inference from ‘attractive’ to ‘attractive, but not beautiful’.
.. Properties of the speaker A listener’s belief about the speaker’s epistemic state is likely to affect interpretation of any utterance. A listener should only expect the speaker to have intended, for instance, the negation of a contextually more informative utterance, if he can rely on the speaker being in an epistemic state fit to assess the truth of the more informative utterance. This assumption of speaker knowledgeability has been discussed mainly in the context of scalar implicature as the ‘competence assumption’ (Geurts, ) or the ‘epistemic step’ (Sauerland, ), but is relevant more broadly and has also been a main focus in reference resolution (Keysar et al., ; Heller et al., ). All of the above factors are directly related to the listener’s beliefs about the speaker’s cooperativity and reliability. Grice () takes the assumption of speaker cooperativity for granted. However, there are many ways in which speakers are uncooperative. In politics, for instance, speakers are often less informative than required and lying is not unheard of. In advertising, QUD-relevance is often exploited and outright violated (Sedivy & Carlson, ). The meat section at a grocery store will invariably contain a label that says ‘This meat is gluten-free!’, suggesting that perhaps other meat is not (however, meat is generally
.
gluten-free). Finally, children are less likely to learn from unreliable than from reliable speakers (Koenig & Harris, ; Gweon et al., ) and adults become less likely to draw contrastive inferences from utterances produced by unreliable compared to reliable speakers (Grodner & Sedivy, ).
.. Common ground An additional complicating factor is that for each of the abovementioned factors, what is presumed to be in common ground between speaker and listener can influence interpretation. Consider the exchange in (). ()
A: Why didn’t Jane get tenure? B: Because she’s a woman. a. . . . so she’s not good enough. b. . . . so the field is biased against her.
Depending on the beliefs shared by A and B, either one of the inferences in (a) or (b) may be drawn from B’s utterance. Mutual knowledge cannot be computed algorithmically without resort to heuristics, famously characterized by Clark & Marshall () as falling into one of three broad categories: Community Membership, Linguistic Co-Presence, and Physical Co-Presence. Each factor discussed in this section can in principle be characterized as belonging to one of these three categories. However, listeners can have uncertainty about any of the abovementioned factors themselves (e.g. what is the QUD?) and about which knowledge can be presumed to be in common ground (e.g. does my interlocutor share my beliefs about the QUD?). These complexities might seem to pose insurmountable challenges to the study of pragmatic processing. However, substantial progress has been made and can be made in the future by systematically investigating these aspects of interpretation and formalizing how they interact within a testable framework. Next, we survey the empirical facts about scalar implicature processing in detail from a constraint-based perspective and will show that all of the abovementioned factors play a role. These factors stand in contrast with inference-specific idiosyncrasies (e.g. lexical, syntactic), which we also discuss. Throughout, we contrast the predictions of the constraint-based and the informational privilege approaches.
. C-
..................................................................................................................................
.. What’s wrong with informational privilege accounts? Early experiments in scalar implicature processing were aimed at testing the informational privilege Default Hypothesis (Levinson, a)—the idea that, because certain scalar
-
implicatures appear conventionally tied to certain scalar expressions (e.g. ‘some’) and arise seemingly with great regularity, they should also be automatically computed in processing. Only if the context fails to support the implicature should it be cancelled, in a second step of processing, thus yielding the literal interpretation. A truth-value judgement task in which participants were asked to evaluate the truth of under-informative utterances like ‘Some elephants are mammals’ suggested that, in contrast to the predictions of the Default Hypothesis, scalar implicatures incur a processing cost (Noveck & Posada, ; Bott & Noveck, ): response times were slower for pragmatic FALSE responses (indicating that participants drew the implicature that ‘Some, but not all, elephants are mammals’) than for literal TRUE responses (indicating that participants interpreted the utterance as indicating that ‘Some, and possibly all, elephants are mammals’). These initial results were followed by a slew of studies replicating this ‘costly implicature’ effect in many different paradigms using many different measures, including reading times (Breheny et al., ; Bergen & Grodner, ), truth-value judgements (Degen & Tanenhaus, ), mouse-tracking (Tomlinson et al., ), speed-accuracy tradeoff paradigms (Bott et al., ), and eye movements in the visual world (Huang & Snedeker, a, ; Degen & Tanenhaus, ). In fact, so numerous are the contexts in which processing the implicature is delayed compared to literal controls that some have concluded that it at is this point ‘wellknown’ that ‘scalar implicatures are derived with a delay’ (Chemla & Singh, ). In light of the early evidence, a new informational privilege proposal was formulated for how scalar implicatures are processed: the Literal-First hypothesis (Huang & Snedeker, a). Under this account, a pragmatic processing stage follows a literal processing stage, thus explaining why scalar implicatures generally show increased processing times compared to literal content. This account is also linguistically appealing because it neatly divides semantics and pragmatics into two separate stages and treats semantics as prior to pragmatics in processing, mirroring the relative positions of semantics and pragmatics in the hierarchy of linguistic representation. There is here a clear analogy to the case of syntactic processing: the Literal-First hypothesis assumes an initial stage of context-insensitive processing of literal information, after which contextual, pragmatic information is integrated. However, the s saw the downfall of informational privilege accounts of syntactic parsing, as evidence increased that contextual information is integrated in syntactic processing as soon as it becomes available. In much the same way, evidence is now mounting that there are contexts in which scalar implicatures do not incur a processing cost. Most of this evidence has come from visual world eye-tracking (Grodner et al., ; Breheny et al., ; Degen & Tanenhaus, ). For instance, Degen & Tanenhaus showed that scalar implicature computation for ‘some’ was delayed when number terms like ‘two’ and ‘three’ were contextually available alternatives for the speaker, but not when the speaker could only use the quantifiers ‘some’ and ‘all’. There is also evidence from reading times showing that listeners exhibit no processing cost in reading quantifiers like ‘some’ in implicature-supporting compared to nonsupporting contexts (Politzer-Ahles & Fiorentino, ), suggesting that, in contrast to predictions of informational privilege accounts, contextual information can rapidly enter scalar implicature computation. Taken together, the empirical evidence suggests that the answer to the question ‘Do scalar implicatures incur a processing cost?’ is: ‘It depends’. That is, neither one of the two informational privilege accounts under consideration—the Default or the Literal-First
.
hypothesis—can capture the full range of data. Sometimes scalar implicatures are processed quickly; and sometimes they are processed slowly. Just as in the case of syntactic processing, this is where constraint-based accounts become a useful way forward, both for providing explanatory value as well as for generating new predictions.
.. Beyond informational privilege accounts How does a constraint-based account go beyond simply claiming that scalar implicature processing is context-dependent? First, it states explicitly that both the probability and the speed with which an implicature is derived is a function of the contextual support for the implicature. Thus, if one can identify the strength of different cues to interpretation, one should be able to predict, for a novel situation with different combinations of cues, what listeners’ likely interpretations will be. Identifying the set of cues that listeners use in interpretation is a daunting task and presumably why, despite the large amount of variability in implicature rates across experiments (from to %; see Degen, , and Dieussaert et al., , for overviews), there have been relatively few investigations into the different contextual factors that determine the probability of drawing the inference. However, one step in this direction has recently been taken by Degen (): by combining corpus analyses and web-based experiments through Amazon’s Mechanical Turk, she showed that for , naturally occurring instances of ‘some’ in the Switchboard corpus (Godfrey et al., ), not only was there a large amount of variation in the strength of scalar implicatures, but this variation was systematically predictable from contextual features, including syntactic features (e.g. whether the ‘some’-Noun Phrase (NP) was realized in the partitive or non-partitive form; or whether it occurred in subject position), semantic features (e.g. whether ‘some’ was relatively weak/indefinite or strong/quantificational), and pragmatic features (e.g. whether the embedded NP referent had been previously mentioned). Note than none of these cues are the general ones mentioned in section .—speaker epistemic state, QUD, world knowledge, properties of utterances and their alternatives, speaker cooperativity, and common ground. However, some of the investigated cues are plausibly mediators for these more general cues—for instance, speakers may be more likely to use the partitive when the stronger alternative is QUD-relevant. In general, the relation between bottom-up cues to speaker meaning available in the linguistic signal itself and top-down expectations that listeners bring to bear on utterance situations is still vastly under-explored. While a systematic investigation of the extent to which these cues matter is still lacking, their role in scalar implicature has been demonstrated in various paradigms via individual manipulations in controlled experimental settings. For instance, the speaker’s epistemic state modulates the speed with which listeners draw scalar inferences (Bergen & Grodner, ; Breheny et al., ) and the probability with which listeners draw the inference (Goodman & Stuhlmüller, ). The QUD modulates the probability with which listeners draw scalar inferences (Zondervan, ; Degen, ; Degen et al., ). World knowledge in the form of listeners’ prior beliefs about the effect of certain actions on objects modulates the probability with which listeners draw scalar inferences (Degen et al., ). Properties of utterances themselves—both the observed utterance and its
-
alternatives—modulate the probability of a scalar inference. For example, when listeners believe the speaker could have used number terms instead of ‘some’, implicatures are drawn less often (Degen & Tanenhaus, ; see also Skordos & Barner, Chapter in this volume, for the importance of reasoning about alternatives in children’s scalar inferences). The costlier the stronger alternative, the less likely the implicature is drawn (Rohde et al., ; Degen et al., ). Contrastive stress on the scalar increases the probability of the implicature (Cummins & Rohde, ; De Marneffe & Tonhauser, to appear). And while the investigation of speaker reliability effects on implicatures is even more sparse, there is evidence that in contexts in which the stronger alternative is face-threatening for the listener, implicatures are drawn less frequently (Bonnefon et al., ), suggesting that using the weaker alternative is also a device for speakers to avoid being impolite. In addition, resource limitations affect the probability with which scalar implicatures are drawn. An increasing number of studies have shown that listeners are less likely to respond pragmatically when under cognitive load (De Neys & Schaeken, ; Dieussaert et al., ; Marty & Chemla, ). While some see this as conclusive evidence that scalar implicatures incur a cost, the constraint-based perspective suggests an alternative explanation: each of these studies used the Bott & Noveck () style ‘Some elephants are mammals’ stimuli. For these stimuli, world knowledge does not support the implicature (i.e. we already know that all elephants are mammals), while the need to retain the assumption that the speaker is being informative does. This constitutes a situation of cue conflict. Under constraint-based accounts, the resolution of cue conflicts is one of the contributors to processing difficulty (Elman et al., ). It may thus be the resolution of this cue conflict in favour of speaker cooperativity and against world knowledge that is incurring the cost. The constraint-based approach predicts that if one were to set up situations where the implicature is strongly supported by the context, then increasing cognitive load should lead to more implicatures. The variation in probability and speed with which scalar inferences are drawn has various implications, both methodological and theoretical. Methodologically, it suggests that experimenters, when investigating a particular factor in a particular paradigm, should be vigilant of the settings the other factors are likely to take on in their experiment. Not manipulating a factor explicitly does not mean that participants don’t assign it a setting. For instance, when investigating the effect of the QUD on scalar inferences it doesn’t matter how strong the manipulation is; if participants believe that the speaker is unlikely to know about the truth of the stronger alternative (or if they have uncertainty about whether she is), then it will appear as though there is no effect of the QUD because participants will at most draw ignorance inferences. We believe that for the factors we have listed above, a researcher should have a good estimate of that factor’s setting in his experiment (or figure out a way to measure it). Otherwise, rather than evaluating hypotheses about naturalistic language interpretation, the results might reflect effects of complicated meta-linguistic reasoning. The above review suggests that informational privilege accounts like the Default or the Literal-First hypothesis are on a weak footing. Constraint-based accounts offer promise as a unifying framework within which to explore scalar implicature processing (see Breheny, Chapter in this volume, for a similar review and conclusion, though couched within a slightly different framework).
.
. C-
.................................................................................................................................. Discussing constraint-based approaches to all areas of pragmatics is far outside the scope of this chapter. Other chapters in this Handbook highlight the ways in which multiple constraints affect referential choice (Davies & Arnold in Chapter ); focus (Kim, Chapter ; Tonhauser, Chapter ); exhaustivity inferences associated with it-clefts (Onea, Chapter ); and negation (Tian & Breheny, Chapter ). Here we touch on the application of constraintbased approaches to a few additional domains.
.. Perspective-taking Physical co-presence, operationalized as information that interlocutors have visual access to, has served as a testing ground for asking when interlocutors can take into account differences in each other’s perspectives. Experimenters manipulate which information is common (shared) and privileged (available only to one interlocutor) (Keysar et al., ). Physical co-presence has been used to address questions about production and comprehension. Production studies under the rubric ‘audience design’ (Clark & Murphy, ) ask to what extent speakers tailor their utterances for their interlocutors. Comprehension studies, under the rubric of ‘perspective-taking’ (Keysar et al., ; Heller et al., ), ask to what extent listeners consider differences between their knowledge and that of the speaker in online processing. We focus here on comprehension, noting that the literature has followed the same trajectory as we have discussed for syntactic ambiguity and scalar implicature. There was always consensus that listeners can take the speaker’s perspective. The question was how heavily they weigh differences in perspective and whether there is an informationally privileged stage of egocentric-first processing. The rationale for an initial egocentric stage is that (a) taking into account an interlocutor’s perspective is resourcedemanding; (b) the listener’s own perspective is a good proxy for the speaker’s perspective; and (c) errors can be corrected when miscommunication occurs. Competing constraintbased accounts assume that multiple probabilistic constraints, which include beliefs about the reliability of physical co-presence, the amount of information provided by the specific context, and goodness of fit between a linguistic expression and potential referents, determine whether there will be strong and immediate or weaker and delayed effects of physical co-presence. Keysar et al. () provided striking evidence for initial egocentrism. A confederate speaker and a naïve listener sat on opposite sides of a box with cubbyholes: Some contained objects both participants could see (shared objects); others could only be seen by the listener (privileged objects). The speaker directed the listener to move objects. When a privileged object was a better fit for the referential expression than a shared object (e.g. ‘tape’ is the referring expression: a roll of sticky tape is privileged, and a cassette tape is shared), listeners typically looked first to the privileged object and often began to reach for it.
-
Subsequent research by Nadig & Sedivy () strongly qualified the interpretation of these results. When potential referents are equally good fits for the referring expression and the description is felicitous (Hanna et al., ; Heller et al., ), listeners neither look at nor reach for the privileged object. Hawkins & Goodman (), using a production task, demonstrated that the referring expressions used by Keysar et al. () are less informative than those generated by naïve speakers. Ironically, then, apparent failures of perspectivetaking might be attributable to sophisticated expectations about speaker behaviour—that is, to perspective-taking. An explicitly constraint-based approach to resolving the apparent conflict between the Keysar et al. () and Heller et al. () studies, couched in Bayesian terms (Heller et al., ), predicts the differences by assuming that listeners consider and apply appropriate weights to privileged and common ground. Listeners need to monitor both objects in privileged and shared ground for independent reasons: for example, a speaker asking an information-seeking question is most likely referring to an unseen object. Indeed, when a confederate asks a question, listeners rapidly direct their attention to cubbyholes with privileged objects (Brown-Schmidt et al., ; Brown-Schmidt, ).
.. Lexical precedents A question that is ripe for a constraint-based approach is whether there are immediate partner-specific effects on lexical precedents, as argued by Metzing & Brennan (). Consider a situation where an ambiguous object was previously labelled a vase or a funnel and then subsequently referred to by the same name, maintaining the precedent, or by a different name, breaking the precedent by either the same speaker or a different speaker. Based on a meta-analysis of results, Kronmüller & Barr () documented three different effects with different time-courses. They proposed that each arises from distinct processes: (a) an early automatic priming effect, which leads to a small, fast same-speaker advantage for maintaining a precedent; (b) large, speaker-independent costs when a precedent is broken; and (c) a late recovery stage where costs for broken precedents are reduced when the precedent established by the first speaker is broken by a new speaker. A constraint-based analysis would attribute effects to different stages or types of processing only a last resort after taking into account factors such as: () how confident the speaker was in using a name (confidence might be conveyed by different constructions and variations in prosody); () listeners’ probabilistic beliefs about whether or not a lexical precedent used by an interlocutor was a good fit for the object; and () whether the listener would expect a new speaker to use the same form as the previous speaker. For example, weaker effects of broken precedents should arise when the speaker expressed less confidence.
.. Learning: Adaptation, generalization, and speaker-specificity A growing body of research in language processing motivated by how listeners cope with speaker-variability has focused on questions of speaker-specificity, adaptation, and generalization. For example, in speech perception there is a general absence of simple, reliable
.
acoustic cues to phonemes and word boundaries. Moreover, there is variability within a talker and across talkers. Multiple factors contribute to variability, including variants of a language (e.g. American versus Australian English), smaller regional dialects (accents), socioeconomic class, and even different groups within the same high school (Eckert, ). From a constraint-based perspective, this leads to questions such as: () how do listeners adapt on the fly to specific speakers, modifying existing hypotheses based on prior knowledge? () How and when do listeners generalize to new speakers (see, for example, Kleinschmidt & Jaeger, )? Similar questions are beginning to be addressed in pragmatic processing. An eyetracking study by Grodner & Sedivy () suggested that when a speaker uses modification unreliably and listeners believe the speaker might be pragmatically challenged, they no longer anticipate that a prenominal scalar adjective is likely to signal contrast. This finding has recently been replicated and extended by Ryskin et al. (), who demonstrate similar effects from exposure to a range of utterances, without prior knowledge about the speaker’s pragmatic competence. Similarly, when a talker uses contrastive focus associated with the pitch accent L+H* unreliably with pre-nominal adjectives, listeners down-weight it as a cue when it occurs in a prosodic contour (tune) in a different construction (Kurumada et al., n.d.). Focusing on informativity, Pogue et al. () demonstrated that when listeners are exposed to two talkers, one of whom is under-informative, they judge new underinformative utterances as more likely to have been produced by that speaker. The same pattern did not hold for over-informative talkers. This is not surprising given that an utterance that technically gives more information that is strictly needed to identify a referent might be efficient because the additional information is useful, especially when there is uncertainty in the referential domain (Graf et al., ; Degen et al., ). In addition, Yildirim et al. () showed that listeners’ expectations about how quantifiers map onto specific quantities (e.g. judgements about how likely a speaker is to use ‘some’ to refer to a given quantity of green candies) shift with exposure to different distributions in the input. In future work, it will be crucial for the pragmatic processing community to examine the kind of variability that speakers naturally produce and, therefore, listeners encounter, using corpus studies and production studies to determine both the utterance choices speakers make in a given situation and how those choices might vary. Here again there is a lesson to be learned from constraint-based approaches to syntactic ambiguity, where architecturebased hypotheses were tested with stimuli for which the relevant constraints had not been identified or quantified. The importance of understanding what speakers typically say and the extent to which that influences listeners’ expectations is further highlighted by the trajectory of work on scalar implicature and perspective-taking.
. I
.................................................................................................................................. Across the domains of experimental pragmatics reviewed in sections . and ., there does not appear to be a case where a certain type of information is systematically preferred over another in processing. Instead, which type of information is processed when, and which
-
type of information is weighted more heavily in the resulting interpretation, varies as a function of the experimental context. This is the signature of a domain-general information processing mechanism that weighs and prioritizes information differentially. What is the principle according to which information is integrated (weighted in the resulting interpretation; prioritized during online processing)? Under the constraint-based approach, the probability with which an interpretation arises and the speed with which it is processed is a function of the contextual support for that interpretation. But what function? And how is contextual support computed? Because they are probabilistic, combine multiple sources of information, and typically appeal to architectural constraints as a last resort, constraint-based models share many of the same principles as rational (ideal observer) approaches to perception and cognition (Ernst & Banks, ; Alais & Burr, ; Knill, ). Over the past ten years, explicit computational models in the rational tradition have been developed within the emerging sub-field of probabilistic pragmatics (Franke & Jäger, ; Goodman & Frank, ). These formalize the Gricean programme by assuming that speakers try to produce utterances that strike a balance between being true and informative while minimizing utterance cost, implicitly encoding that speakers are cooperative. Interpretation is considered to be probabilistic inference. Listeners are hypothesized to combine information about their prior beliefs and the speaker’s likely utterances via Bayes’ Rule, while making minimal assumptions about resource limitations, thereby accommodating the fact that rationality is bounded. The speaker’s likelihood function provides a principled way to incorporate the factors discussed in sections . .. These models have been successful at explaining (variation in) a variety of pragmatic inferences, including scalar implicature (Russell, ; Goodman & Stuhlmüller, ; Degen et al., ); embedded implicatures (Potts et al., ; Bergen et al., ); hyperbole (Kao et al., ); interpretation of under-informative referring expressions (Frank & Goodman, ; Franke & Degen, ); and phenomena at the semantics-pragmatics interface such as the interpretation of gradable adjectives (Lassiter & Goodman, ) and vague quantifiers (Schöller & Franke, ), as well as the production and interpretation of pronouns (Rohde, Chapter in this volume). While rational computational models of pragmatics have achieved remarkable success in explaining the quantitative patterns in the outcome of the inference process at the sentence level, they have not yet been applied to online processing. We see this as a crucial step in developing constraint-based theories of pragmatic processing. The general question is: how is one to link computational model predictions to measures of processing effort? One a priori possibility is that the probability of an interpretation as predicted by a rational model directly predicts the processing effort associated with arriving at that interpretation. For instance, when the probability of a scalar implicature is high, it will be computed quickly. However, this simple idea cannot be right: in many of the Bott & Noveck () style ‘Some elephants are mammals’ experiments, the probability of the implicature appears to be high, as evidenced by large implicature response proportions, and yet the implicature is slow to process compared to the sentence’s literal interpretation. In addition, focusing only on the resulting implicature probability obscures how that probability arises: there are many circumstances in which implicatures are highly probable,
.
only some of which rely on cooperative speaker behaviour. For example, listeners may have uncertain prior beliefs about the world and then hear an utterance from a competent speaker that clearly addresses the QUD and that strongly biases towards the inference. In this case, one would expect the inference to arise quickly and seamlessly. By contrast, in other situations, listeners may have strong priors about what the world is like (and assume that these priors are in common ground, for example by assuming that all elephants are mammals and that this is a commonly known fact). Assuming that speakers are cooperative and trying to produce utterances that strike a balance between being true and informative, certain speaker utterances will be unexpected because they are either very unlikely to describe a true world or because they are highly under-informative, given the prior—‘Some elephants are mammals’ is an example of the latter. The reading time literature on syntactic processing suggests that reading times—a proxy for processing difficulty—are logarithmically related to surprisal (Levy, ; Smith & Levy, ): The more unexpected the utterance, the greater the processing effort incurred by the listener/reader. We believe that this approach can be straightforwardly extended to pragmatic processing, using richer contextual utterance probabilities as the basis for the computation of surprisal. Under the rational models, the surprisal of an utterance like ‘Some elephants are mammals’ will be high. If listeners are drawing on their model of the speaker in comprehension, then after Bayesian inference, the most likely explanation of the utterance is that the speaker intended the implicature. This would thus constitute a case of a high probability of the implicature being intended, while nevertheless being a difficult utterance to process because of its large surprisal value. Preliminary evidence for the promise of this ‘pragmatic surprisal’ view comes from Augurzky & Franke (), who showed using ERPs that listeners’ N amplitude scales with the contextual surprisal of utterances like ‘Some of the dots are black’. How does this relate to response times in a truth-value judgement task? While reading times and the N component tend to scale with utterance surprisal, truth-value judgements (and eye movements in the visual world) are likely to reflect a combination of both utterance surprisal and the resulting evidence for a particular interpretation. This requires a more complex linking function. The precise interplay of production and comprehension is an exciting avenue for future research, and one that is only made possible by building testable and incrementally refinable computational models that make quantitative predictions about the complex interplay between the various factors we have discussed in this chapter.
. Q , - ?
.................................................................................................................................. In the introduction, we laid out five developments that were crucial for the emergence of successful constraint-based models in syntactic processing. We use these as a guide in assessing the current state of constraint-based approaches in experimental pragmatics. As a reminder, the developments were (a) identifying and quantifying relevant constraints; (b) understanding the central role of context; (c) clearly specifying goal structures;
-
(d) developing explicit testable linking hypotheses; and (e) appreciating the richness of the signal. Along the way we will try to identify further promising questions for future work that we think will advance the field. Experimental pragmatics is in the early stages of identifying and quantifying the constraints involved in pragmatic processing. The initial focus on questions about how quickly pragmatic inferences are computed for a limited set of phenomena played a catalytic role in helping the field develop. But it also detracted both from the complexity of pragmatic inference and from efforts to understand the signal, that is, what speakers actually produce, which is a crucial part of understanding both how to prepare stimuli and how listeners process the necessarily artificial input provided to them by experimenters. One way of identifying and quantifying the constraints involved in pragmatic processing is to combine corpus studies with web-based experiments (e.g. Degen, ) to collect large numbers of judgements about, for instance, the role of each of the factors discussed in section . in the interpretation of naturally occurring language. Combining this distributional information with computational modelling and controlled psycholinguistic experiments will further allow for incrementally testing and refining explicit constraint-based theories of pragmatic processing that embrace the complexity of the phenomenon. While this might seem daunting, it is worth noting that studying vision in more complex natural tasks provided insights that ultimately simplified some problems (see Salverda et al., ). For instance, models of fixations in searching a scene on a screen account for less than per cent of the variance, suggesting that there are numerous unexplained factors contributing to fixations. In more complex goal-based tasks (e.g. making tea or making a sandwich), more than per cent of fixations are accounted for by the goal structure. Part of this process is acknowledging the central role of context, a development that is clearly underway. Awareness of clearly specified goal structures in experimental investigations of pragmatic phenomena is still somewhat lacking in certain areas of experimental pragmatics (e.g. scalar implicature processing), while in other areas the experiment-level QUD or goal of the experimental task forms the basis of paradigms (e.g. visual world eyetracking for investigating contrastive inferences from prenominal adjectives, where the goal, at least when coupled with a task (for discussion, see Salverda et al., ; and Salverda & Tanenhaus, ), involves referent identification and the QUD is therefore ‘Which of these images does the speaker intend me to select?’). Formulating explicit, testable linking hypotheses is often ignored in experimental studies in language processing, and it has not been a strong suit of experimental pragmatics. The emergence of computational models in probabilistic pragmatics has brought with it the need to clearly specify the link between model predictions and various dependent measures of utterance interpretation (Degen & Goodman, ), which is only exacerbated once one moves to online processing measures. Formulating and evaluating linking hypotheses will play an essential role in developing and testing scalable models of pragmatic processing. Finally, as touched upon earlier, experimental pragmatics quite generally would benefit from researchers appreciating the richness of the signal more than has hitherto been the case, via systematic investigations of naturally occurring language that are not subject to potential researcher bias in sampling relevant examples. Moreover, it will be increasingly important to take into account information provided by the speech signal, such as prosody, which provides important cues to a speaker’s intentions and the myriad other cues in the speech stream that are partially linked to information.
.
Making progress on all of these fronts will facilitate the building of explicit models of the rich interplay between bottom-up information (from the signal and context) and top-down information (conversational expectations and world knowledge). There are many interesting avenues to pursue in modelling both the outcome of inference processes as a function of context, and incremental belief update as utterances unfold over time. Another area in which this approach will have far-reaching consequences is in language acquisition and learning. Understanding the ways in which child-directed speech differs from adult-directed speech across different contexts will allow for extensions of pragmatic processing models to models of language acquisition. Finally, a rich under-explored area of pragmatics that merits much further investigation is social meaning. The focus of pragmatics is typically on information that speakers communicate about the world rather than about themselves or about the relation between speaker and listener. Yet the long history of sociolinguistics has documented the many ways in which speakers constantly communicate about these dimensions, whether voluntarily or not (Eckert, ). Recent initial attempts at bridging the gap between probabilistic pragmatics and variationist sociolinguistics are very promising: for instance, Burnett () demonstrated that the use of the in/ing variants (‘runnin’’ vs. ‘running’) can be explained by a game-theoretic model of speakers attempting to convey different personae (combinations of features like ‘friendly’ and ‘competent’). Earlier we treated the speaker’s epistemic state as one abstract factor that is likely to always matter for pragmatic interpretation. However, listeners’ estimates of speakers’ epistemic states are themselves subject to prior beliefs about the world. For instance, if a listener holds a prior belief that one gender is on average more competent than another, this would predict that listener should be less likely to derive scalar inferences and more likely to derive ignorance inferences from utterances produced by someone of the gender perceived to be less competent. In a similar vein, the phenomenon of testimonial injustice, whereby for instance the same utterance is interpreted as a command when produced by a man but as a request when produced by a woman (Fricker, ), highlights the importance of integrating prior beliefs about conversational goals based on perceived social category.
. C
.................................................................................................................................. In this chapter, we have surveyed the developments in constraint-based pragmatic processing and contrasted this approach with various types of informational privilege approaches in some example domains. Constraint-based approaches to pragmatic processing are still in their infancy. Nonetheless, in the domains where they have been applied, they emerge as the account that is most strongly supported by the data. Moreover, constraint-based approaches are well-suited for important future directions, both in their merging with probabilistic models of pragmatics and in the extension to social meaning.
......................................................................................................................
......................................................................................................................
. I
..................................................................................................................................
.. Overview T chapter reviews recent experimental research into questions about how language and other functions of the mind are integrated when humans communicate. It posits a Gricean system that serves this purpose and discuss how recent developmental and ethological research provides evidence for such a system’s existence. Subsequently it focuses on the much-studied phenomenon of scalar implicature. It first considers the phenomenon of scalar implicature in the broader context of pragmatic effects. A short review of theoretical debates as to the status of various sub-types of scalar phenomenon is followed by sections that discuss experimental research relevant to different interfaces in the Gricean system when it comes to scalars.
.. Pragmatics, communication, and a Gricean system Although it hardly began this way, the field of pragmatics is now increasingly seen as a discipline that is the concern of the psychological sciences. From a psychological perspective, one can consider the systems or functions that the mind realizes in terms of domains such as vision, language, memory, reasoning, and so forth. Pragmatics deals with communication as a domain that is somewhat separate from language. It typically views language as a function in the generative sense—as something which computes phonological and syntactic representations, as well as providing sets of rules for composing meanings of constituents. From this perspective one role of pragmatics is to characterize the systems for computing a mapping between the thoughts we express in situated linguistic utterances and the linguistic structures used in those utterances. A more general role is to explain human communication, be it linguistic or non-verbal. There is a growing consensus that human communication is very distinctive from that of other primates or other species in its use of social-cognitive abilities (Tomasello et al., ; Herrmann et al., ). Research on infants and primates reveals that humans are distinct
in assuming that communicative signals create shared information (Tomasello et al., ; Csibra, ; Moll & Meltzoff, ), and that the signaller’s aim is to provide a kind of good (Tomasello, ; Csibra & Gergely, ). Moreover, infants have been shown to display sensitivity to unseen mental states of communicators—their intentions, beliefs, and desires—in the earliest stages of communicative development (Southgate et al., ; Liebal et al., ). Infants also demonstrate expectations that other agents are rational in achieving their aims (Gergely & Csibra, ). It comes as no surprise, then, to find that the ideas of Grice and his contemporaries (Grice, , ; Lewis, ) are still current in the field of pragmatics. A ‘Gricean’ account of some apparently linguistic phenomenon typically attempts to explain attested meanings, in part, by appeal to a system for deriving meaning that a speaker intends her communicative gesture, or utterance, to have. In a typical Gricean explanation, the speaker’s meaning goes beyond, or departs from, the meaning which can be recovered from any information encoded in the stimulus via a lexicon and grammatical rules. Gricean explanations take for granted a commonly shared level of rationality plus some common expectations about what good the communicator aims to provide. Gricean accounts explain our intuitions about utterance meaning in terms of inferences about which thoughts a speaker intends to share. From a cognitive perspective, if we follow Grice’s proposals about how to account for certain aspects of what a speaker means, we can imagine an inferential system that interfaces with a range of cognitive functions to make decisions about speaker’s intended meaning. These functions include language, Theory of Mind, and long-term memory/conceptual structure. Given the developmental research, then, human infants appear to possess abilities that would constitute a Gricean system for inferring a speaker’s meaning. That is, they have a cognitive system that integrates information about the linguistic stimulus and the utterance situation, including the expectations and goals of the interlocutors, to compute the thoughts that the speaker intends to share. Moreover, the above research suggests that infants possess these abilities at a point where they barely manifest linguistic abilities beyond the one-word stage. Thus we should be encouraged that a Gricean system realizes a naturally occurring function that distinguishes human cognition from that of other species. Scalar implicature is one domain where linguists and philosophers have proposed Gricean accounts of apparently linguistic phenomena. In this chapter, ‘scalar implicature’ refers to a broad class that includes the sub-categories, ‘Straight Scalars’ (SS), ‘Ignorance Inferences’ (II), and ‘Embedded (Scalar) Enrichments’ (EE). These will be illustrated in section .., and contrasted with other pragmatic effects. A diverse range of broadly Gricean accounts of scalar phenomena have been offered over the years (Horn, ; Grice, ; Gazdar, ; Soames, ; Levinson, , a; Sperber & Wilson, /; Benz & Van Rooij, ; Schulz & Van Rooij, ; Franke, ; Russell, ; Frank & Goodman, ; among others). These Gricean accounts have not gone unchallenged (Chierchia, ; Fox, ; Chierchia et al., ; among others) due to some wellknown problems that arise. This debate will briefly be rehearsed in section ... The issue here has come to be about how language and Gricean systems interface. Experimental research in pragmatics has addressed, or is relevant to, this language/ pragmatics interface question when it comes to scalar implicatures. Are they a purely extragrammatical phenomenon? Are scales stored in a mental lexicon? Are there default scalar implicatures? Are some scalar phenomena better described solely within a linguistic
framework? Note that the status question applies independently to SSs, IIs, and EEs. For each sub-phenomenon, there are many nuanced shades in the theoretical positions adopted. In section ., we will review experimental research which can contribute to these debates. To the extent that scalar implicature phenomena are not purely linguistic, we can ask: how are the results of linguistic and extra-linguistic computations integrated in a Gricean inferential system? Experimental research has also produced results that bear on this question. This is reviewed in section ..
. S
..................................................................................................................................
.. Domain of inquiry Scalar implicature is a term that has been used to refer to a wide range of phenomena. As will be discussed briefly in section .., there is considerable debate about how best to account for these phenomena. Under most current approaches scalar phenomena do not get a unitary analysis. Moreover, different, opposing accounts have been offered for the same sub-category of scalar implicature. In order not to prejudge the nature of these phenomena, it is best give an initial sketch of the domain of the research to be discussed here by prototypical example. Let us begin with some examples of the most commonly discussed kind of case—what I shall call the SS. Consider () (), where what follows ‘~>’ would be a plausible implication in easily imaginable situations: ()
Some of the students got an A on the test. ~> Not all of the students got an A on the test.
()
A: Jane was planning to cut the grass and wash the car. How did she get on? B: She cut the grass. ~> Jane did not wash the car.
The SS implication has been characterized as the negation of some proposition that is related to the assertion in virtue of lexical association, as in (), or in virtue of contextual salience or relevance as in (), and perhaps also (). The negated element is almost universally referred to as an Alternative. For the implication under (), the alternative proposition would be All of the students got an A on the test. For (), it could be Jane cut the grass and washed the car, or simply Jane washed the car—depending on one’s theory of scalars (see Katzir, ). () is an example where there appears to be a lexical association involved in the relation between the asserted proposition and the alternative (in virtue of the fact that the alternative could be derived via a replacement of ‘some’ with ‘all’). Here we will refer to such cases as lexical scalars. More frequently perhaps, alternatives become available in virtue of context alone, as in (). Here, as elsewhere, these examples will be referred to as ad hoc scalars. The next sub-category of scalar phenomena is the II. Two examples are given below, both of which are discussed in Grice’s William James Lectures (see Grice, , ).
()
A is planning with B an itinerary for a holiday in France. Both know that A wants to see his friend C, if to do so would not involve too great a prolongation of his journey. A: Where does C live? B: Somewhere in the South of France. ~> The speaker does not know where in the South of France.
()
A: Where is Max planning to visit when he goes to Scotland? B: Edinburgh or Glasgow. ~> The speaker is not certain that Max is planning to visit Edinburgh. ~> The speaker is not certain that Max is planning to visit Glasgow.
In both cases, there is an implication that the speaker does not know about some contextually related propositions. The similarity between IIs and SSs can be made to appear stronger if we take the Gricean perspective on SS implicatures (see section ..). That is, we assume that a speaker who implies Not all of the students got an A does so in virtue of communicating her belief that not all of the students got an A. In other words, where the alternative A is negated in SS, it is in virtue of the speaker conveying that they believe not A. By contrast, the speaker conveys something weaker in II, which is that they do not have the belief that A. In both cases, arguably, an alternative (A) is involved. The final sub-category to be discussed here is EE. One way to characterize EEs of an assertion is to imagine that a sub-constituent of the assertion has been enriched by SS and the resulting enriched sub-constituent contributes to the overall proposition being expressed. For instance, a reading of (a) containing an EE could be glossed by imagining the constituent ‘hit some of the targets’ being given a reading hit some and not all of the targets. This is indicated in (b). ()
a. Exactly one player hit some of the targets. b. Exactly one player hit some but not all of the targets.
This example is taken from Potts et al. () which reports that participants in an experiment readily understood sentence (a) according to the gloss in (b). The significance of this sub-category of scalar phenomena will be discussed in section ... Here we can note that, unlike with SS or II, it is not possible in many cases of EE to describe the enrichment of the linguistically determined meaning in terms of the conjunction of the literal meaning of the sentence used and some other proposition. For instance in (), the literal meaning of (a) and the enriched meaning, (b), do not stand in an entailment relation with each other. This ends the survey of core phenomena typically discussed under the heading of scalar implicature. I round up this section with some comments on the location of these exemplars in the domain of pragmatics more generally. Earlier in this section I mentioned some prototypical examples of scalar phenomena. Over the years, linguists and philosophers have argued about other kinds of examples, as to whether or not they are cases of scalar phenomena. Below, I list three cases which will be relevant when we come to discuss experimental work on scalars: ()
Mary has two children. ~> Mary has no more than two children.
()
I will give you $ if you wash my car. ~> I will give you $ only if you wash my car.
()
You can have coffee or cake. ~> You can have coffee and you can have cake.
An SS approach to cases involving numerals (Horn, ; Gazdar, ; Levinson, ) proposes that ‘two children’, like ‘some children’, has an encoded meaning that could be glossed in purely lower-bounding terms—something like two or more children. Thus () would be true even if Mary has three or more children. The implication that she has no more than two is considered to be a SS, with the alternative being Mary has three (or more) children. This view has been contested, however (see Horn, ; Geurts, ; Breheny, ), and the issue remains somewhat open. In contrast to the numeral case, early Gricean accounts of so-called ‘conditional perfection’ argued against treating the implication indicated under () as SS (see Atlas & Levinson ), but treated it as another kind of implicature. However, others disagree and would treat this, or related implications as SS (Matsumoto ; Horn, ; Levinson, a; Von Fintel, ; Franke, ; Geurts, ). () is an illustration of a so-called ‘free-choice’ inference. Again, there is not universal agreement about how to explain the implication illustrated here. Some proposals favour an account based purely on the linguistic expressions used (Zimmermann, ; Geurts, ; Barker, ). However, many recent discussions propose that free-choice inferences can be explained by applications of the mechanism for SSs (Klinedinst, ; Fox, ; Geurts, ; Franke, ). Beyond the domain of scalars, a great number of phenomena have been classed as ‘implicature’ or have been analysed in broadly Gricean terms. These include Relevance implicatures (Grice, ), politeness implicatures (Brown & Levinson, ), figurative language such as metaphor, hyperbole, metonymy and irony (Grice, ; Sperber & Wilson, /), loose use or approximation (Sperber & Wilson, /), M- (or manner-) implicatures (Sperber & Wilson, /; Levinson, a) and other enrichments variously described as I-implicatures or R-implicatures (Levinson, , a; Horn, ; Carston, ). Also, diagonalization accounts of existential closure (Stalnaker, ), and certain accounts of presupposition (Stalnaker ; Simons, ) and anti-presupposition (Sauerland, ) have been cast in broadly Gricean terms. While space considerations preclude us looking at these other phenomena in detail, it is relevant to the discussion of experimental work on scalars to consider a broad class of pragmatic effects that contrast in interesting ways with scalars. This could be called the class of R-/I-/Relevance implicatures. Two examples are given below: ()
()
A comes upon B, who is looking under the bonnet of an obviously immobilized car. A: There’s a gas station around the corner. ~> The speaker believes it likely that the gas station is open. Mary tripped on a Persian carpet and she broke her wrist. ~> the wrist break resulted from the tripping over.
These implicatures contrast with scalars insofar as they result in the speaker communicating a stronger salient proposition (there’s an open garage around the corner; Mary tripped on a Persian rug and she broke her wrist as a result), rather than the negation of a stronger salient proposition. Also, arguably, they do not involve alternatives (Carston, ; but see Poppels & Levy, ). In addition to implicature and related phenomena where literal or linguistic meaning is either added to or changed, many have argued that a Gricean system plays a key role in decisions about what expressions used in an utterance literally mean (Wilson & Sperber, ; Neale, ; Kehler & Rohde, ). A typical illustration of this is disambiguation. Words can be ambiguous due to having multiple meanings and strings of words can be ambiguous due to the possibility of multiple underlying syntactic structures. When utterances are produced in context, they typically pose a decision problem as to the intended parse. To the extent that a Gricean system is employed in this decision process, then this also impacts on experimental research on scalars that seeks to detect the presence of Gricean vs. Linguistic computations, as we will see in section ... Although a formal corpus survey of the whole domain of linguistic pragmatics has yet to be undertaken, informally it seems that instances of scalar phenomena are very much in the minority when it comes to the domain of pragmatics. What has been more rigorously established (Degen, ) is that even for the prototypical lexical-scalar trigger, ‘some’, actual tokens found in the corpus were judged to give rise to a SS implicature (as illustrated in ()) less than half of the time. In addition, Van Tiel et al. () show that, if anything, other scalar terms would give rise to scalar implicature at lower rates than ‘some’. And yet, although it is perhaps a minority phenomenon, and not as prevalent as one might think even for lexical triggers, it has attracted more attention than any other in recent pragmatics research, particularly in experimental pragmatics.
.. Scalars as Gricean phenomena? Grice’s William James Lectures (Grice, , ) contain accounts of two examples of IIs. These are given in () () above. In his programmatic ‘Logic and Conversation’ Grice reserves a special category of conversational implicature for this phenomenon. This is Group B, where a speaker violates the maxim of informativeness on account of her desire not to violate the maxim of truthfulness. Group B implicatures result in an II—that the speaker does not provide the required information because to do so would be to assert something for which she lacks adequate evidence. These texts by Grice contain no clear discussion of SSs. However, subsequent work in the Gricean vein locate SSs within Grice’s Group A of pragmatic phenomena (Horn, ; Gazdar, ; Soames, ; Sauerland, ; Geurts, ). Here the speaker merely exploits the maxims of informativeness and truthfulness in order to get across the required information implicitly: the speaker is judged to have not explicitly expressed a more informative alternative to her assertion on the grounds not of ignorance but of believing that alternative to be false. Before we turn to Gricean approaches to EE, we can consider other, broadly Gricean, approaches to scalars. Recent Rationalist, Bayesian models of scalars in the game theory tradition (Benz & Van Rooij, ; Franke, ; Frank & Goodman, ) focus on the interest that speaker and hearer have in coordinating beliefs about the world. In Rational
Speech Act (RSA) approaches to scalars, as articulated in Frank & Goodman () and Bergen et al. (), informativeness comes into the derivation through a notion of specificity—an optimal speaker uses the most specific utterance she can, relative to what she believes about the world. Thus, if one makes assumptions about the alternatives we reason over (see Frank & Goodman, ), a speaker who utters () when she could have used ‘all’ is judged to be less likely to believe all than some and not all. Alternatively, Relevance-based approaches to scalars, some of which operate within the probabilistic/ Bayesian tradition (Sperber & Wilson, /; Merin, ; Russell, ) work with a comparative notion of relevance and propose that speaker and hearer reason about each other on the basis that the utterance is relevant. While the asserted utterance must be presented as at least relevant to some degree, alternatives may be judged as potentially more relevant, leading to either SS or II, depending on what can be assumed about the beliefs of the speaker. One observation made within relevance-based frameworks (see in particular, Sperber & Wilson, /; Russell, ; but also Magri, ) is that while an SS implicature may make the utterance more relevant, the explicitly asserted proposition has to be at least adequately relevant in the context. Russell () illustrates this point with the following kinds of example: ()
a. Oh crap! Some of the students passed. b. Oh crap! Only some of the students passed. c. Oh crap! Not all of the students passed.
In the context of an interjection like, ‘Oh crap!’, we can assume that (b,c) are felicitous due to the fact that not all is part of the explicit assertion, as it provides the relevant bad news. The infelicity of (a) suggests that it is not sufficient for the speaker to merely scalarimplicate not all to achieve a basic level of relevance in context. Evidence such as this suggests a difference in status of the some and not all implications between (a) and (b). This is a point that bears on experimental research, as discussed in section ... Turning to EE, as with SS, Grice makes no proposals in ‘Logic and Conversation’ for these. In fact, it has been a long-standing criticism of ‘Logic and Conversation’ and many subsequent Gricean proposals that they cannot account for many EE effects (Cohen, ; Wilson, ; Carston, ). There is however scope within the general Gricean programme to deal with local adjustments to meaning (Grice, ; Sperber & Wilson, /; Carston, ; Neale, ). In the Relevance Theory (RT) tradition, it is assumed that there is no presumption of literal truthfulness among language users, only of Relevance (in the technical RT sense). Translated into psychological terms, the proposal is that the lexically encoded association between expressions and meanings serves only as a starting point when it comes to inferring the proposition explicitly expressed by an utterance. The theoretical question of some interest then concerns the mechanisms that are responsible for modulating lexically encoded meaning. Recent proposals in the RSA framework (Potts et al., ; Bergen et al., ) also implement the idea that the computation of speaker meaning does not presume literal truthfulness. Rather, speakers and hearers take for granted that linguistic expressions may be mapped to novel meanings in context. Uncertainty about which mapping the speaker intends then becomes part of the bigger decision problem concerning under-specification of explicit meaning, given the linguistic form used in the utterance.
Notwithstanding the generality of EE phenomena and the fact that a formal description of the phenomenon is possible within a broadly Generative framework, recent nonGricean, ‘Grammatical’ approaches to scalars have cited EE as grounds for bringing both SS and EE phenomena back within a more squarely linguistic description. Grammatical Theory (GT) proposals are outlined in Fox (); Chierchia et al. (); and elsewhere. The basic idea is that SS can be represented in the syntactic structure for an utterance via a covert exhaustification operator, exh, that functions much like ‘only’. Thus (), repeated in (a) below, could have a syntactic representation (LF) as in (b), with exh[S], defined relative to a set, Alt, of alternative sentences in (c) ()
a. Some of the students got an A on the test. b. [exh [ Some of the students got an A on the test.]] c. ||exhAlt[S]|| = 1 iff ||S|| = 1 and for all A ∈ Alt, unless ||A|| ⊆ ||S||, then ||A|| = 0
Embedded scalars then become straightforward to accommodate, given that exh can appear at scope sites other than the root clause. For example, an LF for (a) could be as in (): ()
[Exactly one playeri [exhALT [ti hit some of the shots.]]]
Aside from EE phenomena, an important motivation for the GT comes from a problem for the general Gricean programme that has yet to be mentioned—the Symmetry Problem. This is a problem about the choice of alternatives for SS (and perhaps for EE). It is widely recognized that accounts based on Grice’s proposals cannot explain why speaker and hearer would reason about an alternative A, when its ‘symmetric’ alternative not A should be an equally viable alternative (see Fox, , and Katzir, , for details). It seems something other than informativeness and relevance to a purpose impact on the derivation. Grammatical theorists would argue that purely linguistic, or structural, factors are involved in the computation of alternatives (Fox & Katzir, ; Katzir, —but see Romoli, , Trinh & Haida, , Breheny et al., , for a discussion of well-known problems). If this conjecture is correct, it motivates a view of SS and EE as more squarely linguistic phenomena. It should be noted, however, that even the structural approach to alternatives presumes that Gricean systems need to interface with linguistic systems when it comes to selecting the contextually relevant alternatives from among the formally defined alternatives.
. E -G
..................................................................................................................................
.. Default implicatures: A testable proposal about the language-pragmatics interface Levinson (a) developed a hybrid Gricean account of scalar phenomena which includes the idea that at least lexical-scalar implicatures have a special default status, lying between
purely linguistic phenomena and purely Gricean pragmatic phenomena. Levinson conjectures that being default in this way has processing implications in that lexical-scalar items trigger the activation of the scalar implication and only a strong contextual bias against this implication would result in its suppression. To give an example, the implication indicated under () would be a default, and so we would expect comprehenders to activate that implication irrespective of context. Note that, according to all approaches, including Levinson’s, a sentence containing a lexical trigger like ‘some’ may or may not be understood to carry the SS implicature and, as mentioned, a corpus survey shows that it does so slightly less than half the time (Degen, ). Thus, where an utterance is understood to not carry the implication, the default account must assume some process of de-activation or suppression. Experimental research tested Levinson’s prediction that participants who include a scalar implicature in their response should do so either faster or at least not slower than participants who do not. In Bott & Noveck (), participants were given a verification task based on world knowledge. Critical items asked participants to verify items as in (a,b): ()
a. Some anchovies are fish. b. Some fish are anchovies.
A participant would judge (a) true when the SS implicature is not part of his interpretation of the item, but false when the scalar implicature, Not all anchovies are fish is part of the understanding. By contrast, (b) is true with or without the implicature. Bott & Noveck () report that with-implicature responders had more errors and took longer than no-implicature responders on critical items like (a). The results run contrary to Levinson’s predictions. They also showed that a group of participants given less time to respond were more likely to give a no-implicature response than a group given more time. Noveck & Posada () report similar results. Moving from sentence verification to reading-time studies, Breheny, Katsos, & Williams (; see also Katsos et al. ) demonstrate that participants take longer to read segments carrying implicature triggers (‘some of the consultants’) when reading texts where context biases the SS implicature compared to texts where they do not. This result was replicated in Bergen & Grodner () with items that controlled for a repeated-name penalty that might otherwise explain the previous reading-time results (but see PolitzerAhles & Fiorentino, ). Further independent disconfirmation of the default position comes from SpeedAccuracy Trade-off (SAT) studies reported in Bott et al. (). Using verification tasks similar to Bott & Noveck (), Bott and colleagues report that responses emerged faster for the no-implicature group than the with-implicature group. Finally, using a mousetracking paradigm, Tomlinson et al. () report that participants who respond ‘False’ to an item like (a) first move the mouse cursor in the direction of a ‘True’ response before correcting. By contrast, ‘True’ responders move the mouse directly to the ‘True’ side of the screen. Levinson’s default approach predicts that, if anything, the opposite pattern would be observed. Taken together, these results undermine the default position, as that was developed by Levinson. This being the case, do these results also support a broadly Gricean position, as against, say, the GT? We turn to this question in the next section.
.. Counting the cost of scalar implicatures One recurrent theme emerging from research on scalars is that a with-scalar response takes longer than a without-scalar response, although this is not a universally reported finding (see for instance Feeney et al., ; Politzer-Ahles & Fiorentino, ; and references cited in section .. below). Evidence for cost or delay for scalars has been reported in verification tasks (Bott & Noveck, ; Bott et al., ), reaction-time reading (Breheny et al., ; Bergen & Grodner, ) and visual-world eye-tracking (Huang & Snedeker, a). These results seem supportive of the contextualist idea that computing a SS implication involves enriching the encoded non-SS meaning in context. According to Gricean approaches, SSs and IIs involve reasoning about something that a speaker could have said but didn’t, and further coming up with a reason for the speaker not giving more information, in terms of what the speaker was in a position to know. In the literature, the first part of this process is discussed in terms of finding contextually relevant alternatives. The second part is referred to as an ‘epistemic step’ in the case of SS, although a similar process must take place in the case of IIs. It may seem that, by contrast, the Grammatical account of SS shortcircuits these stages by allowing that an operator be inserted into the syntactic structure for the sentence. However, matters may not be so straightforward. Recall that it is widely argued that decisions relating to facets of the linguistically determined proposition in context ultimately fall within the purview of Gricean pragmatics. In particular, a process of contextually determined domain restriction (or reference assignment) is involved in virtually all noun phrases uttered. The process of deciding on the set of alternatives for the exhaustivity operator is virtually identical to the process of choosing a domain for the operator, ‘only’. That is, both exh in (b) above and its explicit cousin ‘only’ quantify negatively over alternative propositions. Thus, the presence of an exhaustification operator in a syntactic structure entails a stage of selecting alternatives, more or less equivalent to the alternative-selection stage in the Gricean derivation. Moreover, even the decision to insert an operator or not to do so would depend on similar factors. Thus, a Gricean process of alternative selection may be required both for the derivation of a SS as a Gricean inference, and its derivation via a covert linguistic operator. This still leaves a potential difference between Gricean and Grammatical derivation of SSs since the Gricean derivation mandatorily includes a step which considers the speaker’s reasons for not asserting alternative propositions. Experimental research that manipulates whether the speaker would know that the alternative were true or not has shown that participants are sensitive to this step. In both Bergen & Grodner () and Breheny et al. (), effects of incorporation of SS implicatures disappear in conditions where the speaker is ignorant of the facts. This sensitivity to an ‘epistemic step’ is not strongly predicted by GT. However, these studies are not completely decisive since the GT does endorse the Gricean account of IIs (see Fox, , in particular). Thus, the results in both Bergen & Grodner () and Breheny et al. () could be accounted for in terms of the salient ignorance of the speaker in the experimental contexts. A Gricean system that decides on the inclusion or not of a covert operator would withhold inclusion in ignorance contexts, favouring instead an II. One straightforward means to test whether the costs of SS simply boil down to the costs associated with the insertion of a covert exh operator would be to compare cases when
participants incorporate SS with cases that involve sentences with an overt exhaustification operator, such as ‘only’. This comparison has been made several times in the experimental literature and the results are informative. In the SAT research reported in Bott et al. (), one experiment has participants responding pragmatically to items like in (a) above and includes items that are just like (a) but for the inclusion of ‘only’: ‘Only some anchovies are fish’. Bott and colleagues report that participants depart from random response sooner in the ‘only some’ condition than the pragmatic ‘some’ condition. Given that the function denoted by [only [some anchovies] and [exh [some anchovies]] would involve the same levels of complexity to compute and verify, it seems that there is some extra cost to processing the SS items compared to their explicit counterparts. This is not predicted by the GT. By contrast, Gricean approaches could argue that a richer representation of context is entailed by SS than by ‘only some’. That is to say, a Gricean scalar implicature involves representing a speaker’s reasons for not using the alternative and this is somehow the cause of the delay. A comparison between SS items like (a) and ‘only some’ items has also been made in a paradigm that manipulates working memory load. The paradigm was introduced to experimental pragmatics in De Neys & Schaeken (), who demonstrated that participants gave fewer SS-based responses to items like (a) when a parallel task made greater demands on working memory than in a low memory-load condition. Marty & Chemla () follow up on this work by comparing the effects of memory load in a ‘some’ condition with an ‘only some’ condition. Their results in the ‘some’ condition replicate De Neys & Schaeken (). They also report that while rates of with-SS responses drop in the high load case, this does not occur in the explicit ‘only some’ case. It is possible to interpret this result as favouring a Gricean account of SS over the GT. The idea would be that the GT implies that the process of representing and verifying the items in both the implicit and explicit conditions should come with comparable costs. Thus no difference would be predicted. By contrast, as mentioned, Gricean accounts tend to entail a richer representation of the utterance situation for SS items than ‘only some’ items. This would explain the greater effect of memory load in the with-SS condition. However, Marty & Chemla () caution against drawing this conclusion since the Grammatical view entails an extra layer of complexity for the with-SS condition, compared to the ‘only some’ case. This is due to the fact that, according to the GT, the sentence in (a) is inherently ambiguous between a parse that includes the exhaustivity operator, exh, and one which does not. Marty & Chemla argue that it is possible that the extra layer of decision-making involved in the disambiguation makes access to the implicature susceptible to extra memory load. This alternative account awaits confirmation but it is possible to make some observations based on results that we have already. First, ambiguity does not always imply extra cost. For example, Duffy et al. () show that reading times for equi-biased homonyms (i.e. ambiguous words whose senses are equally frequently employed) do not differ from unambiguous controls when the target stimulus is supported by context. By contrast, out of context, there is a cost. Returning to the corpus survey in Degen (), we can observe that around per cent of samples containing ‘some’ were judged to give rise to an implicature. If we assume scalar implicature is represented in linguistic structure, we seem to have a case of equi-biased ambiguity. The question then comes down to whether the items are supported by context or not. In lexical disambiguation research, an
ambiguous word appears only once in an experimental session. So context for that word would be associated closely with the trial in which it appears. In the memory-load studies, the same target expression appears numerous times in virtually the same context. Thus, if anything, we should consider these items to have some supporting context, especially for participants who tend to respond one way or the other across the bulk of trials (see Bott & Noveck, ; Feeney et al., ; Antoniou et al., ; among others). Thus, if we assume that the SS ambiguity is equi-biased but the experimental setting provides supporting context for the parse that a participant gives, we should not expect these ‘some’ trials to engender extra cost by virtue of any ambiguity inherent in the stimulus. We could, on the other hand, assume an equi-biased ambiguity but a neutral context in these working memory studies. If that were the case, it would be instructive to compare the effects of memory load on latencies of responses that include the implicature to those that do not. This comparison is made in De Neys & Schaeken (; see also Dieussaert et al., ). They report that ‘False’ responses (i.e. with-SS) are delayed under high load compared to low load. By contrast, ‘True’ responses (i.e. with no SS) are not delayed in the higher load condition. If it is an equi-biased ambiguity of ‘Some fish are anchovies’ that gives rise to a different result under load, then we should expect a similar difference regardless of whether participants choose the with-SS response or the without-SS response. Thus, the GT account is not confirmed by De Neys & Schaeken ()’s results. Nor is it confirmed by a comparison of reaction times for filler items, including those like (b) above. Again, we should expect to see a difference under load for a set of items that contain ambiguous ‘some’, but no difference is found. An alternative explanation for the memory-load comparison could appeal to the fact mentioned in section .., that even where a SS is derived, the asserted, lower-bounded meaning of the utterance should be relevant in its own right. In cognitive terms, we can assume that this means that where items like (a) give rise to an implicature, the information that some or all anchovies are fish should attract attention in its own right. The negative proposition that not all anchovies are fish, being implicated, should not be more salient than the assertion. This idea is supported not only by examples like those in (), taken from Russell (), but also by the mouse-tracking studies reported in Tomlinson et al. (), which show that participants who respond ‘False’ to (a) are drawn first to a ‘True’ response (consistent with the lower-bound meaning of the asserted sentence), then shift mouse direction. By contrast, the corresponding sentence with ‘only’ (‘Only some anchovies are fish’) does not prioritize the lower-bounded component in this way. In fact, there is a broad consensus that the semantic, asserted content is not all anchovies are fish, while there are arguments that the positive, lower-bounded part of the sentence is either a scalar implicature or a presupposition (Horn, ; Van Rooij & Schulz, ; Ippolito, ). If that is right, then the positive part of the ‘only’ item could not be more salient than the negative. How would an ‘explicit relevance’ account of these memory-load data go? Let us assume that what is explicit and (assumed to be) relevant must receive attention, in virtue of being activated by the linguistic stimulus. By contrast, potential scalar implicatures need not receive attention in virtue of the linguistic processing of the stimulus itself. Then it stands to reason that scalar implicatures are more likely to not be activated under increased memory load than asserted parts. For the ‘some’ items, this means that the negative, ‘not all’ component is liable to be not activated in high-load conditions, while for the ‘only some’ items, if anything the lower-bounding ‘some or all’ component is susceptible. In the latter case, not accessing the positive component would not affect rates of ‘False’ judgements on
critical items, since that judgment is based on the negative, not all component being activated (see also Tomlinson et al., ; Antoniou et al., ). By making these assumptions about the effect of a requirement that the explicit content is relevant, we can explain the pattern of results in this kind of memory-load study. This explanation of the working-memory effect and other effects (such as the mousetracking effect) is consistent with constraint-based accounts of processing scalar implicatures (Grodner et al., ; Degen & Tanenhaus, ; Degen & Tanenhaus, Chapter in this volume), which argue that even comprehension processes that require rich contexts need not be delayed or costly, if the relevant information is sufficiently activated or available. We return to a more detailed discussion of the processing question in section ..
.. Experimental research on potential scalars Up to this point, we have considered what experimental research tells us about the status of SS implicatures, whether they should be viewed as ‘Gricean’ or otherwise. As mentioned in section .., there are several phenomena that attract debate as to their status—whether they should be classified as scalar implicature in the first place or explained in some other way. In this section, I will discuss recent experimental research that bears on these debates. Let us begin with the case of numerically quantified noun phrases. The example in (), repeated below, indicates that sentences containing noun phrases like, ‘two children’, can (and often do) carry an upper-bounding, ‘no more than two’ implication: ()
Mary has two children. ~> Mary has no more than two children.
This could be explained as a SS implicature if the sentence containing, ‘two children’ encodes a lower-bounding meaning, two or more children and where the alternative for SS is formed by replacing the numeral, ‘’, with ‘’. However, a number of differences between SS items like ‘some’ and numerals have been observed (Horn, ; Breheny, ), leading to scepticism that the implication under () is in fact a scalar implicature. Many have argued that the implication under () is an entailment of an encoded meaning for the sentence (Geurts, ; Breheny, ; Kennedy, ; Buccola & Spector, ). This would be the case if the noun phrase is understood literally to mean something like exactly two children—although there are other analytical tools to account for this entailment (see Kennedy, ; Buccola & Spector, ). There are, however, many occasions when numerically quantified noun phrases have a lower-bounded-only reading. For example, in () below, the sense of ‘Mary must have two children’ is better glossed as ‘Mary must have two or more children’ rather than, ‘Mary must have exactly two children’: ()
To receive this benefit, Mary must have two children.
Different theories offer different approaches to these data. Some posit an ambiguity in the sentence between an exactly two and a two or more reading (see Geurts, ; Kennedy, ; Buccola & Spector, ). Breheny () takes a different route, arguing that a separate pragmatic mechanism operates to make available the two or more reading in this case. Both kinds of approach get some support from a recent study that looks at modified numerical expressions, as in ():
()
a. Less than three dots are red. b. Between two and five dots are red.
It happens that the mechanism for deriving the lower-bounded reading in () also derives an existential reading of (a,b) meaning that either of these sentences could be truthfully uttered of a situation where seven dots are red. Marty et al. () report that participants asked to verify (b) for displays involving seven red dots, respond ‘True’ in such a situation at a higher rate than their ‘False’ response to situations that make (b) false according to either lowerbound-only or doubly-bound readings (i.e. where only one dot or no dot is red). This data points to participants seeing ‘between m and n’ sentences as having a lower-bounding reading, as predicted by these non-scalar implicature accounts of numerical expressions (see Nouwen et al., Chapter in this volume, for greater discussion of modified numerals). More relevant data that numerically quantified noun phrases are not like other scalars comes from Marty et al. (), which employs the memory-load paradigm to compare the status of the readings of ‘two dots’ versus ‘some dots’. Recall that, under higher memory load, participants less frequently include the upper-bounding not all inference in verifying sentences with ‘some’. This means that more responses based on the lower-bound reading some or all are elicited under high memory load (De Neys & Schaeken, ; Marty & Chemla, ). Marty et al. () replicate this effect for target items, ‘Some dots are red’ when the visual display shows only red dots. However, for target items ‘Four dots are red’ with a screen displaying six red dots, the rate of lower-bound readings (‘True’ responses) decreases under load, while the rate of upper-bound, exactly responses increases. This experimental evidence supports the introspective evidence that the doubly-bounded exactly reading of numerals does not have the same underlying basis as the doubly-bounded not all reading for ‘some’. Furthermore, the fact that lower-bounded responses drop off under memory load supports the view that this reading is derived pragmatically (Breheny, ), although Marty and colleagues caution that the effect under load may be due to the lowerbound reading for numerals being more inferentially complex (see Buccola & Spector, , for details). On the other hand, verification of the lower-bound reading is less complex than for the doubly-bounded reading, since checking for four or more red dots only requires finding four red dots, not checking all dots, while checking for exactly four red dots requires checking all dots. So it is not clear that complexity is a factor here. Turning now to free-choice inferences, illustrated in (), it was mentioned in section .. that many researchers have argued these are derived using scalar implicature derivations. ()
You can have coffee or cake. ~> You can have coffee and you can have cake.
There is less experimental research on free choice than other types of pragmatic effect, but one notable result comes from Chemla & Bott (). This study presents participants with items such as () in contexts where it would be false if participants derive the free-choice inference. If they do not derive this inference, they would judge () true: ()
Elsie the engineer is allowed to save a kangaroo or a fork.
Chemla & Bott () report that, unlike the case with scalar implicatures, participants are not delayed in their ‘false’ judgements compared to their ‘true’ judgements. Moreover,
Chemla & Bott also included unambiguous control items (that have the same truth-value irrespective of whether the free-choice inference is included). They found an interaction between truth-value and condition such that ‘false’ judgements were delayed relative to ‘true’ for the controls, but not the critical judgements. The implications are that, as with numerals, free-choice inferences do not have the same underlying basis as other scalar implicatures. This could be because free-choice inferences follow from the conventional meaning of disjunction under certain operators. That is, they are not pragmatic after all. Alternatively, it could be that free-choice inferences are pragmatic but their derivation is not of the same nature as standard Gricean derivation for scalars. Chemla (a) explores one kind of derivation of free-choice effects that make them more like I/R-implicatures discussed in section ... By contrast, Van Tiel & Schaeken () explore the idea that free-choice inferences are derived as scalars but have a lower cost than typical SSs, like (), because their alternatives are easier to retrieve (being a sub-part of the asserted sentence). Van Tiel & Schaeken compared decision times on SS items involving ‘some’, free-choice inferences and also examples of conditional perfection (illustrated in () in section ..) and clefted sentences. They find that only ‘some’ items show a delayed response when participants include the respective inference. Van Tiel & Schaeken argue that if their three pragmatic inferences were in fact scalar implicatures, what sets them apart from scalars involving ‘some’ is that all three have easier-to-retrieve alternatives. However, an alternative explanation could be that none of these three pragmatic inferences are scalar implicatures. This would be consistent with the SAT evidence discussed earlier (see Bott et al., ) that the not all inference with ‘some’ emerges more slowly than the same inference with ‘only some’ even though the set of alternatives required to interpret both are identical. To sum up this section on the language-Gricean interface, experimental research has fed into the many debates as to the status of a wide variety of implications. Even relatively simple survey data and reaction-time studies, when designed in a way that is informed by theory, can play a key role in helping theorists to evaluate different accounts of different linguistic constructions.
. H G -
.................................................................................................................................. The focus of the experimental research reviewed in the last section has largely been on questions of status—for a given implication, should it be regarded as default, linguistic, or Gricean?; is it a scalar implicature or not? In this section, we focus on a different set of questions that take for granted the existence of Gricean pragmatic inferences, including scalar implicatures. These questions have to do with the architecture of a system that combines linguistic information with information about the speaker’s goals, beliefs, etc., about previous shared information, long-term encyclopaedic memory, and so forth. None of these questions are specific to scalar phenomena but several studies that are relevant to these architectural questions have used scalar phenomena as their subject matter.
.. Literal first or free-for-all? When it comes to proposals about how linguistic computations and Gricean inferences may be actually integrated in cognition, Grice himself has been widely seen as committing to a ‘literal first’ view in ‘Logic and Conversation’ (Grice, ), although Searle () is perhaps more definitive on this issue (see Gibbs, b; Cacciari & Glucksberg, ). In the domain of scalars, this could translate into the idea that the literal unenriched meaning is first accessed prior to an implicature being derived. However, it is by no means necessary that Grice’s account of implicature derivation commits us to the temporal priority of the literal in actual utterance processing. This is a point made in Recanati () and elsewhere (see Neale, ; Geurts, ; Breheny, ). It happens that there is much evidence against a literal-first view from research beyond the domain of scalars. In particular, in the domain of metaphor and figurative language processing, there is clear evidence that metaphoric meanings can emerge in the same time course as literal (McElree & Nordlie, ) and that they are generated automatically in language processing (Glucksberg et al., ).
... The integration of a Gricean System in online utterance processing One motivation to set ourselves against a literal-first hypothesis comes from the insight that many dimensions of under-specification are ultimately resolved by recourse to the speaker’s intentions. That is, when sentences contain lexical or structural ambiguity, or require referential domain restriction by context, as they typically do, one could assume that the ultimate arbiter of how to resolve this ambiguity is by inferring what sense the speaker intended (Wilson & Sperber, ). When this assumption is paired with the apparent fact that contextual information is integrated into language processing decisions in the same time course as linguistically derived information (Altmann & Steedman, ; Tanenhaus et al., ), then it would follow that a Gricean system for inferring speaker meaning should be intimately involved in all levels of language processing, not just the derivation of implicatures. Could this be correct? One set of studies looking at contrastive inferences in reference assignment point to a positive answer to this question. In Sedivy et al. () participants heard, ‘Click on the tall glass’ when a visual display included a tall glass and another tall object as well as two other objects. When one of those two objects was a shorter glass, participants looked more to the tall glass prior to the onset of the word ‘glass’, compared to a condition where both other objects were irrelevant distractors. This result shows that participants are exploiting an inference that people typically use modifying adjectives in referential phrases to serve a contrastive function. This inference is arguably Gricean in origin (see Sedivy, , for discussion). Grodner & Sedivy () explore the idea that, as a Gricean inference, the contrast effect ought to be diminished if the participant believes the speaker to have some problems that affect her ability to communicate efficiently. They contrast an ‘unreliable speaker’ condition with a control condition where the speaker is assumed to have no such problems and find, as a Gricean approach would predict, that the anticipation effect disappears in the unreliable speaker condition. Similarly, Hanna et al. () replicate Sedivy’s anticipation effect while participants hear the modifier ‘empty’ in a condition where both speaker and hearer share knowledge of the visual display that the participant sees. However, in a speaker-ignorant
condition, the display is changed without the speaker knowing. In that case, again, the anticipation effect goes away. These results and others (e.g. Davies & Katsos, ; Pogue et al., ; Rubio-Fernandez, ) demonstrate that comprehenders are able to exploit Gricean inferences during very early online processing of inherently underspecified or ambiguous referential expressions. Moreover, exploitation of Gricean inference in language processing is not simply default but influenced by information about the speaker’s abilities and also her mental state (see also Arnold et al., ; Heller et al., , ). This strongly suggests that the Gricean inferential system is well-integrated into the processing of linguistic stimuli. More generally, we can briefly consider research on the use of common ground or perspective-taking as further evidence that systems for inferring speakers’ intentions are integrated into decisions about definite reference. In one commonly employed perspectivetaking task, the participant-hearer and a speaker are jointly aware that the participant sees objects that the speaker does not, while they discuss objects that are commonly visible. As the speaker does not know what kind of objects these private objects may be, she cannot refer to them using phrases like ‘the small train’. Thus, participants ought to ignore these privately viewable objects when processing ‘the small train’ even if one of them is a good fit for that description. Although there is ample evidence that participants find it difficult to entirely ignore the private competitor object (Keysar at al., ; Hanna et al., ; see Brown-Schmidt & Hanna, ), there is also ample evidence that participants’ early anticipation of the referent is influenced by what they know about what could be the speaker’s intended referent (Hanna et al., ; Heller et al., ; Brown-Schmidt et al., ; Ferguson & Breheny, ; among others). That participants’ perspective-taking is not perfect is accounted for in terms of constraint-based models of utterance processing (MacDonald et al., ; Hanna & Brown-Schmidt, ; see also Degen & Tanenhaus, Chapter in this volume) according to which mental-state information acts as one of many competing constraints in automatic language processing.
... Literal first in scalar processing? Turning now to research on how scalar phenomena are processed online, we find a set of apparently mixed results. In the following discussion, one aim is to sort through studies that utilize time-sensitive methods with scalars to determine which might provide insights into how pragmatic scalar inferences are accessed and integrated online. To date, this broader, ‘How?’-question has been approached via a narrower question concerning the time course of access to and integration of scalar inferences. This research is at least relevant to the ‘literal first’ view of the relation between linguistic and Gricean inferences. Recall that section .. discussed studies that indicate some kind of delay or cost for scalars. This may lead researchers to conclude that, unlike other pragmatic or context effects, scalars are somehow special in incurring delays. But we should be cautious here. Many of the studies that give rise to these results are not designed to control for factors that may lead to a delayed response but are not relevant to the time course question. For instance, when ‘some’ is understood as some and not all it denotes a function that is in many ways more complex than when it is understood literally. Representing some and not all of the consultants had a meeting is more likely to involve representing both the consultants that had a meeting and those that did not; whereas the literal proposition (lacking the not all upper-bound) is less likely to do so. This richer representation of the
speaker meaning may contribute to a delayed response in reading time studies (e.g. Breheny et al., ) or verification tasks (Bott & Noveck, ). Thus, longer latencies in these studies do not clearly support a literal-first view. Verification studies that compare items with ‘some’ and ‘only some’ are potentially more informative on the time course questions since the extra complexity of the upper-bounding not all implication in the pragmatically enriched meaning is mirrored in the ‘only some’ condition. Bott et al. ()’s SAT study found a delay in the ‘some’ case compared to ‘only some’. As suggested in section .., this delay could be due to the richer representation for the pragmatic enrichment, involving the speaker’s reasons for choosing the weaker alternative. Still, even if the richer epistemic context for the pragmatic condition is responsible for this delay, we need to consider how it bears on our evaluation of the literal-first and other hypotheses about the time course of scalars. It could be that, in virtue of being more complex, a representation of the full meaning of the stimulus takes longer to emerge; or the verification decision task is delayed due to the extra memory demands implied by the more complex representation. Neither of these states of affairs imply that a representation of the literal part of the overall meaning emerges prior to the pragmatic enrichment. Thus, we should look beyond these results when considering the time course question. And so we turn now to studies using visual-world eye-tracking and Event-Related Potential (ERP) methods that are potentially more informative on the time course question. A key methodological contribution on the time course question was made by Huang & Snedeker (a), who look at embedded scalar enrichments (EE from section ..).1 These studies set out to compare the time course of access to and integration of a pragmatic interpretation of noun phrases containing ‘some’ with an interpretation of items containing ‘all’, and also with numerical determiners ‘two’ and ‘three’. At least the ‘all’ condition involves items that do not involve scalar implicature. Participants are asked to point to ‘the girl with some of the . . . ’ or ‘the girl with all of the . . . ’ in a visual display containing two girls. One girl has all of one set of three items (e.g. soccer balls) and one has some but not all of a different set of items—two of four socks. The residue of this partitioned set of socks is with one of two male characters also in the display. Participants preview the visual display prior to the instruction and thus have the opportunity to set up representations of the sets of objects prior to the verbal stimulus. This design therefore eliminates one of the extraneous costs associated with the complexity of the pragmatic interpretation vis-à-vis the semantic, since the relevant representations are available prior to the linguistic input. That is, prior to the instruction, participants not only represent the socks in possession of one of the girls but also the rest of the socks, in the possession of one of the boys. Another key aspect of this design for the ‘some’ items lies in the fact that if participants process the
According to the taxonomy in section .., when participants understand, ‘Click on the girl with some of the socks’ to mean click on the girl with some and not all of the socks, this cannot be a SS but it could be categorized as EE. As mentioned in that section, there are different theoretical views concerning the status of different scalar phenomena. However, it is interesting that neither the broadly Gricean camp (e.g. Geurts & Van Tiel, ; Bergen et al., ) nor the Grammatical camp see EE as involving the same kind of derivation as the standard Gricean derivation of SSs. Irrespective of the ways in which some of the scalar enrichments may be derived within a Gricean inferential framework, this literature on the time course of Gricean enrichments takes for granted that these are broadly Gricean, and the discussion proceeds from there. This is the stance adopted in this review discussion also. 1
instruction according to the literal meaning, the referential phrase is ambiguous up until the point of disambiguation in ‘socks’. Hence we would expect no bias to form until then. However, if the pragmatic inference can be exploited in the reference assignment process immediately when it becomes available (from the offset of ‘some’) then bias should form toward the target at the same rate as in the ‘all’ condition, which is unambiguous from the offset of the determiner. Huang & Snedeker report that participants’ visual bias to the target in the ‘some’ condition is significantly delayed relative to the ‘all’ condition. In particular, prior to the point of disambiguation of the noun, participants’ bias to the target in the ‘all’ condition is significantly above chance but in the ‘some’ condition it is not. In follow-up work, Huang & Snedeker () establish that participants can exploit the pragmatic inference prior to the later disambiguation point in the noun, but there is a considerable delay, compared to the ‘all’ condition. While Huang & Snedeker stop short of endorsing a full ‘literal first’ view, they argue that their results are consistent with the idea that the effects of pragmatic inferences must trigger off lexical decoding processes; and this would give some priority to linguistically driven inferences over pragmatic. In contrast to Huang & Snedeker (a, ), a number of similarly constructed visualworld studies show little or no delay between the ‘some’ and ‘all’ condition. Grodner et al. () utilize a very similar design to Huang & Snedeker, involving EE, and find no delay. Breheny et al. () utilize SSs and find no delay of ‘some’ items relative to ‘all’. The reasons for such a disparity in results have yet to be fully clarified. Two kinds of proposal have been made about this disparity, reviewed in Degen & Tanenhaus (; see also Grodner et al., ). One difference between the studies finding no delay and those that report a delay relates to the presence of number items. In Huang & Snedeker’s experiments, there are two ways of referring to each of the target items (using ‘some’/’all’ or ‘two’/’three’), while in Grodner et al. () and Breheny et al. () there is only one. It could be that, in the course of experimental sessions, participants come to pre-code targets on the visual display in terms of the quantifiers ‘all’ and ‘some’, bypassing processes that feed off both compositional semantic knowledge and Gricean inferences. This would explain why bias to targets does not differ between the ‘some’ and ‘all’ condition. An alternative line of explanation for the disparity (see Grodner et al., ) holds that the presence of items like ‘click on the girl with two socks’ undermines the basis for exploiting a Gricean inference in the items that involve ‘some’ instead of ‘two’. Although the reasons are not entirely clear why this would be the case, it could be down to the fact that ‘two’ is a better cue to the target, since number in the subitizable range are easier to identify visually. Degen & Tanenhaus () explore these two hypotheses in two visual-world studies. In the first, they replicated the results of Grodner et al. () in a study without number items but with various decoy trials that lower the possibility of pre-coding. A second study also contained number items, with the number in question being either in the subitizable range (as used in Huang & Snedeker’s studies) or larger, non-subitizable numbers. Degen & Tanenhaus () found that bias to target forms faster in ‘all’ trials than ‘some’ when the number of items was large but not when the number was small (in the subitizable range). It is interesting that Degen & Tanenhaus’s results show that big-set ‘all’ trials show faster bias formation than small-set ‘some’, replicating Huang & Snedeker’s findings. It seems clear that set sizes in these visual display plays a role in bias formation that has yet to be fully clarified (but see Chao & Breheny, under review).
Overall, these results are certainly problematic for the literal-first view. In Degen & Tanenhaus’s without-numbers study, there is less chance of pre-coding, yet Grodner et al.’s results are replicated. In the with-numbers study, comparing targets with small numbers, there is no difference between ‘some’ and ‘all’ conditions. So, the pragmatic condition is not necessarily slower than the non-pragmatic condition, even in the presence of number trials. On the other hand, an open question remains about what factors determine results when there is a difference. The expected delay when set size is in the subitizable range did not materialize. In fact, this is the condition where no delay was found.
... ERP studies on the time course of scalar implicature Like eye-tracking methods, ERP techniques have been widely employed in psycholinguistics due to the fact that fine-grained resolution can be achieved for the time course of effects (see Van Berkum, ). ERP studies looking at scalar phenomena have mostly adopted the verification procedure of Bott & Noveck (). Nieuwland et al. () used sentences like ‘Some people have lungs/pets . . . ’. In previous research on language processing, anomalous or unexpected stimuli tend to elicit an EEG response characterized by heightened negativity of wave forms centred around milliseconds (ms) from the onset of the stimulus (an N effect). Nieuwland and colleagues reasoned that if there were rapid access to the scalar implicature (not all people have lungs/pets) one might expect an N effect when participants encounter the object noun phrase ‘lungs’ compared to ‘pets’. They tested a group of participants and found no overall N effect. However, these participants also undertook an Autism Quotient survey (Baron-Cohen et al., ) which measures something like sociability across a range of domains, including communication. Nieuwlandet al. () report that, when participants are grouped by sociability according to this index, the high-sociability group do show an N effect. Hunt et al. () report an ERP study with a similar design except that the stimuli were evaluated relative to a visual context. They find that for an item like ‘The student cut some of the steaks in this story’, there was an expected large N effect when none of the steaks were cut (semantically False condition) compared to True controls. In a pragmatically False but semantically True condition (when all steaks were cut) there was an intermediate response. Similar effects are reported in Spychalska et al. (). These ERP studies are informative in showing that scalar inferences are integrated online. However, they do not speak directly to question about the early time course of scalar implicature integration because the ERP response measured is triggered by a word that appears some time after the onset of ‘some’. In fact, the delay is longer than normal due to the use of the one-word-at-a-time presentation procedure necessary for ERP. Moreover, N responses are known indicators of surprise or anomaly and do not necessarily indicate the time course of access to linguistic vs. pragmatic inferences. One study that attempts to probe ERP responses on the determiner word itself is reported in Politzer-Ahles & Fiorentino (). In a verification task, Politzer-Ahles & Fiorentino report that when ‘some’ occurs in an all context (i.e. when the sentence would be false under its some and not all reading), they found an effect similar to when ‘all’ appears in a falsifying visual context in an early ms time window. The authors argue that this effect is not like the standard N effect. Moreover, unlike the semantically anomalous ‘all’ items, anomalous ‘some’ items also gave rise to an effect in a later, ,ms, region. The authors suggest
this may reflect reanalysis of ‘some’, undoing the pragmatic enrichment. Thus there is some ERP evidence for rapid response to pragmatic ‘some’ in a comparable time course to ‘all’. Moreover, the response to inconsistent ‘some’ has a distinct signature, consistent with the distinct source of the relevant inference—linguistic vs. pragmatic. To sum up this discussion of time-sensitive research on processing scalar implicatures, the bulk of relevant evidence points toward a system in which inferences derivable from linguistic semantic knowledge and from Gricean pragmatic knowledge can become available in the same time course. There are studies, notably Huang & Snedeker (a, ), that run contrary to this conclusion, but subsequent visual-world research has revealed that factors such as set size impact on the measured response.
.. Beyond time course studies The time course studies reviewed in section .. are most relevant to whether language processing is literal first. They shed some light on the broader picture of how scalar implicatures come to be accessed and integrated in on-line processing. However, they leave a lot of questions open. As discussed in Breheny et al. (), if scalar implicature processing is more or less automatic, then results reviewed earlier raise interesting questions about how this could be. Typically, language processes are conceived of as selecting among choices pre-determined by the rules of the language and lexical information activated by the input stimuli. An ambiguous word or string of words typically leaves a relatively small number of parsing options to choose among, in terms of lexical meaning or syntactic structure. But when it comes to implicatures, particularly ad hoc implicatures (studied in Breheny et al., ), there is no clear sense that any word or string of words should come with a scalar implicature, an R-implicature, an M-implicature, etc. On the face of it, it is difficult to conceive of how language processing is set up to continuously decide if any of a set of pragmatic effects should be integrated as linguistic input is incrementally processed. When it comes to scalar implicatures, the decision must involve alternatives and, at least in the case of ad hoc scalars, these would not come pre-compiled with the lexical input. So, the question is, how would automatic language processes determine when a scalar implicature should be integrated and what it should be? The beginnings of an answer to this question comes from research into on-line computation of the source of relevance, or ‘Question Under Discussion’ (QUD). Taking a lead from suggestions in Sperber & Wilson (/), as well as Van Kuppevelt () and Roberts (, a), a series of recent studies provides some evidence that incremental processes anticipate not only the content of the utterance (its semantic and pragmatic interpretation) but also the source of relevance for the utterance, typically described in terms of QUDs (Tian et al., ; Clifton & Frazier, ; Clifton & Frazier, ; Tian et al., ; Kehler & Rohde, —see Breheny, , and Tian & Breheny, Chapter in this volume, for discussion). To illustrate the connection between scalars and QUD, consider an utterance of () in a situation where the indicated implication is communicated: ()
Albie ate the salad. ~> Albie did not eat the steak or dessert.
The relevant alternatives here (Albie ate the steak; Albie ate dessert) could be culled from the question which of the salad, steak or dessert did Albie eat?, since this determines a logical space in which Albie eats just steak, just salad, just dessert, steak and salad, salad and dessert, etc. (see Hirschberg, ). Imagine that anticipatory incremental processing is set up to compute likely QUDs given linguistic input and context. Then, depending on the prosody of the sentence (see Sperber & Wilson, /; Breheny, ) input up to the verb in () would make Albie ate what? a more likely relevant question. If, in addition, a small number of edible items are salient in the context (see Altmann & Kamide, ) then these become candidates for what is eaten, thereby constituting the alternatives for scalar inference. Recent work shows a positive impact of prosodic stress, or focus, on rates of scalar implicature (Chevallier et al., ; De Marneffe & Tonhauser, ; Cummins & Rohde, ). Focus constituents have long been thought to reflect the underlying relevant question/QUD in the typical case (Sperber & Wilson, /; Roberts, , a; Breheny, , —see Cummins & Rohde, , for discussion). So this offline data provides indirect support for the idea that QUDs play a role in determining which scalars may be derived from an utterance. To date, little work has been completed that examines on-line computation of QUDs and scalars together, but the prospects are promising that for at least one set of Gricean inferences, we will arrive at a better understanding of the factors that impact on their availability in incremental processing.
. C—B G
.................................................................................................................................. In the last section, I speculated that the question of how scalar implicatures are integrated as part of incremental processing might be answered via a theory of incremental anticipation of a specific feature of context—the source of relevance, or QUD. As mentioned, there is research beyond the domain of scalars suggesting that QUD anticipation may be part of incremental language processing and this may bear on more explicit aspects of the message such as reference assignment (Kehler & Rohde, ; Tian & Breheny, Chapter in this volume). In closing this review, I will discuss how evidence from disparate areas of research can be brought together to build a coherent picture of a human Gricean system. A most impressive difference between human infants, on the one hand, and chimpanzees or other great apes, on the other, lies not in their ability to represent goals or intentions, knowledge states and the like (Call & Tomasello, ), but in their response to communicative stimuli. One stark illustration of this difference comes with the so-called object choice task—where the participant is aware the target is in one of two locations and the experimenter communicatively points to the correct location. Behne et al. () report that -month-old infants interpret the signal correctly, while chimpanzees respond at chance, even though they notice the stimulus (Call & Tomasello, ) and they have the same goal to find the target. Although different accounts of these differences have been given (Tomasello et al., ; Gergely & Csibra, ) they have in common the idea that
humans expect a communicative signal to be relevant in a shared context. That communicative signals engender expectations of relevance has been demonstrated in pre-verbal infants from at least months of age (Yoon et al., ; Kovács et al., in press), while toddlers around year of age have been shown to rely on expectations of relevance in a shared context to interpret communicative gestures (Southgate et al., ; Liebal et al., ). Moving to later stages of language/communicative development, it has been shown that children given a sentence verification task need to be provided with a context of plausible denial (Crain et al., ). This is such that, if the target sentence is S then the utterance context ought to be such that whether S is true is a plausible QUD. Given recent adult processing research and also child developmental research, a reasonable hypothesis about a human Gricean system is that it develops to process communicative stimuli, including linguistic utterances, in a way that determines the relevance of that stimulus in a shared context. Just how we should define relevance, be it in terms of a formal structure involving questions and sub-questions (Roberts, , a; Ginzburg, ), or in terms of utterance utility (Sperber & Wilson, /; Van Rooij, ; Csibra & Gergely, ), or in terms of impact on the likelihood of a topic proposition (Carnap, ; Russell, ) remains to be determined. These analytical alternatives have descriptively similar coverage, including when it comes to scalars. Assuming a relevance-oriented hypothesis about the cognitive system underpinning utterance processes, the aim of future research would be to investigate what factors impact on the determination of the source of relevance for an utterance, on the content of the utterance itself and the interplay between these factors.
A Thanks to Chao Sun, Chris Cummins, and students at my Pragmatics and Cognition class for comments and feedback on earlier drafts of this chapter.
......................................................................................................................
() ......................................................................................................................
.
. I
.................................................................................................................................. L is often used to talk about events. While each event is in some sense a singular unique occurrence, we generally talk about them as happening in a certain way with particular consequences, carrying on or ending after some amount of time, and involving different participants who bear particular relationships to the event. These conceptually general and grammatically relevant properties of events are captured in theories of event structure. The goal of such theories is to explain the patterns in meaning that arise from the structural aspects of language. Event structures do this by decomposing an individual event into a set of sub-events which stand in particular relationships to one another. These component sub-events and their relationships are often encoded in independently articulated linguistic constituents that are scattered throughout a sentence. As a result, these components must be recovered and put back together for a comprehender to properly understand the event that a sentence denotes. This chapter examines how some of the grammatically determined interpretations arising from event structure unfold in real time, with a focus on sentence processing. We will examine studies from the psycholinguistic literature that have investigated different aspects of event structure using a variety of experimental techniques. But let’s first motivate the theory of event structure and decomposition before turning to consider how different component parts of event structure are put to use during real-time language processing.
. D
.................................................................................................................................. In many respects, the observation that language encodes certain properties of events stretches back to Aristotle, but more recently it has been Davidson’s () programme
()
that put events centre stage.1 That paper, along with Lakoff (), McCawley (), Dowty (), and Ross () set the stage for a wider appreciation of the treatment of events as being decomposed in language. Consider the relationship between the examples in (). The differences in meaning correspond with differences in the formal structure of each sentence. The presence of -en in (b) changes a non-verbal predication in (a) to a verbal predication and elicits a process meaning absent in (a) while retaining the end state of the steel being hard. The transitivity change from (b) to (c) introduces an agent that causes the process described, which is missing from (b), while retaining the process and end state meanings of (a) and (b). ()
a. The steel is hard. b. The steel hardened. c. John hardened the steel.
To capture these form-meaning correspondences, theories of event structure decompose the main event of the sentence into sub-events where predicates like DO, CAUSE,2 and BECOME act as relations over events. Roughly following Parsons (), the event analyses in () secure the intuitions that (c) entails (a,b) and that (b) entails (a). In each case the additional event predicates add processes, causes, and agents into the overall representation. ()
a. λe1 [ hard(e1, the steel) ] b. λe1 [ ∃e2 [ BECOME(e1, e2) & hard(e2, the steel) ] c. λe1 [ DO(e1, John) & ∃e2 [ CAUSE(e1, e2) & ∃e3 [ BECOME(e2, e3) & hard(e3, the steel) ] ] ]
Considerations of this type led researchers to generalize event structures into recurring event templates that host the verbal root of the event alongside its participants, including patients and agents, and describe the temporal unfolding of the event via the relationships between its sub-events (Dowty, ; Parsons, ; Pustejovsky, , b; Rappaport Hovav & Levin, ). The following three sections focus on these aspects of event structure and how they are employed during real-time language processing. Section . examines how event structure constrains the interpretation of verbal roots, with particular reference to expressions of manner or result, and explores how these constraints play out in 1
Events are not the only way researchers have tried to execute Davidson’s programme, though the relationship events encode may be indispensable. Bayer () argues that events themselves are a distinct type from possible worlds/situations. He notes that worlds (or slices of worlds) are unable to distinguish buying and selling, but these cases are linguistically distinct as shown with modification. (a) seems true given the context, whereas (b) is not even though the same transaction is taking place. ()
Mary’s religion has a tradition where wedding guests get a flower from a local vendor to symbolize their respect for the impending union. a. John bought a flower from a street vendor in honor of Mary’s wedding. b. A street vendor sold John a flower in honor of Mary’s wedding.
2 We will set aside questions concerning whether CAUSE is the right event relationship here (Parsons, ; for arguments against CAUSE, see Fodor, , Lombard, , Pietroski, , Williams, ).
.
language processing. Section . considers how verbal roots compose with event predicates and their arguments to establish an event’s temporal contours and whether psycholinguistic evidence has something to say about the components of that composition. Section . investigates how event participants that are implicit in different event structures are encoded and used during real-time language processing. Finally, section . offers a brief summary of the chapter with some observations on experimental study and strategy.
. T
.................................................................................................................................. At the core of events are questions about verbal roots and how they relate to event structure. Verbs contribute crucial information to event descriptions, including highly idiosyncratic information. For instance, the verb waltz provides fairly specific information about the motion of an action, whereas dance is more vague. From an event structure perspective, however, both encode the manner of motion in which an event is carried out, and thus differ from other verbs, like disintegrate and break which, regardless of their idiosyncratic differences, encode the result or change-of-state that occurs because of the event (Gentner ; Behrend ; Rappaport Hovav & Levin, ). This distinction between manner and result is grammatically relevant in a number of respects. For instance, manner verbs can occur with unspecified or non-subcategorized objects (a,b), whereas result verbs cannot (a,b). ()
a. Mary danced. b. Mary danced her feet sore.
()
a. *John broke. b. *John broke his leg bloody.
To account for facts like these, Rappaport Hovav & Levin () propose a limited inventory of event structures that verbal roots associate with. Manner verbs are those that modify the DO predicate, specifying the way an event’s activity occurs (a). Result verbs are those that are an argument of the BECOME predicate, specifying the resultant state of the event (b). Because they are associated with BECOME, result verbs require an argument to directly bear their result/change-of-state, thus blocking unspecified and nonsubcategorized objects. ()
a. Manner schema: b. Result schema:
λxλe1 [ DO(e1, x) & root(e1) ] λxλe1 [ ∃e2 [ BECOME(e1, e2) & root(e2, x) ]
To say that a verbal root is associated with one primitive predicate in an event structure is to say that there is an entailment of that meaning component. For example, while the
()
manner verb scrub encodes a particular manner of cleaning, it does not necessarily entail the result of cleaning. Thus in () it is acceptable to deny the implied result of scrubbing, but not acceptable to deny the entailed manner of cleaning. Similarly, the result verb clean encodes a particular resultant change-of-state, but does not provide the manner of cleaning. In (), it is acceptable to deny the implied manner of cleaning, but not the result of cleaning.3 ()
a. Mary scrubbed the bathtub, but it didn’t get any cleaner. b. Mary scrubbed the bathtub, #but not by scrubbing it.
()
a. Mary cleaned the bathtub, #but it didn’t get any cleaner. b. Mary cleaned the bathtub, but not by scrubbing it.
Similarly, the meaning component that is not lexicalized may be expressed outside of the verb itself, as shown in (), but it is odd or redundant to have the encoded verbal component also expressed outside of the verb, as in (). ()
a. Mary scrubbed the bathtub clean. b. Mary cleaned the bathtub by scrubbing it.
()
a. #Mary scrubbed the bathtub by scouring/scrubbing it. b. #Mary cleaned the bathtub sterile/clean.
Given that verbs can have manner and result meanings, we might expect to find verbal roots that lexically entail both a manner and a result. This third class of verbs, however, is thought to not exist.4 In a series of papers over the last twenty-five years, Levin and Rappaport Hovav have argued that the manner and result components of verbs are in complementary distribution (Levin & Rappaport Hovav, , , ; Rappaport
3 Such tests may seem trivial, but they can reveal rather surprising behaviour, for instance, with degree achievements as in (i).
(i) The physicists cooled the plasma (for hours), but it didn’t become cool. 4
Manner/result complementarity is not without controversies. Beavers & Koontz-Garboden () argue that manner of death verbs, which include behead, crucify, hang, drown, and electrocute, lexicalize both a manner (i.e. the way the death occurred) and a result (i.e. the death itself). This argument relies heavily on a claim about what constitutes a ‚‘lexicalized’ meaning. According to Dowty (), lexicalized components are meanings that are entailed regardless of the context. Husband () argues that to the extent both meaning components are encoded, the manner component of manner of death verbs is entailed while the result component is actually presupposed. As a presupposition, the result meaning projects out of negation, as shown in (i). (i) Socrates was not decapitated. a. He poisoned himself. b. #He didn’t die. This suggests that the manner and result components of manner of death verbs reside on different levels of meaning and that manner/result complementarity may capture something about how much of a verb’s meaning can be asserted.
.
Hovav & Levin, , ; see also Kiparsky, ). They propose that this complementarity reflects a constraint on how much meaning a verb root can lexicalize. ()
Lexicalization Constraint: A [verbal] root can only be associated with one primitive predicate in an event schema, as either an argument or a modifier (Rappaport Hovav & Levin, : ).
Building on this dichotomy of verbal meaning components, McKoon & Love () approached manner/result complementarity from a processing complexity viewpoint. As seen in (), result verbs represent a more complex schema than manner verbs. They hypothesized that such complexity would be revealed in a variety of experimental tasks, including lexical decision, speeded sentence acceptability, and a stops-making-sense reading task.5 In the lexical decision task, participants were presented with letter strings and indicated whether the letter string formed a word of English or not as quickly and accurately as they could. The logic of this task is that a more complex representation will take longer to access, predicting that response times to result verbs would be longer than to manner verbs when controlled for confounding factors like length and frequency. McKoon & Love reported that ‘yes’ responses to result verbs took an average of milliseconds (ms) longer than those to manner verbs, a finding consistent with the prediction that result verbs are more complex than manner verbs. In speeded sentence acceptability, participants were presented with a sentence and asked to judge whether it was acceptable as quickly and accurately as possible. Manner and result verbs were matched on all the characteristics of their lexical decision study and on the proportion of their transitive usage, and put in transitive sentence frames like (). Wholesentence reading times were the main measure.6 Sentences with result verbs took an average of ms longer to judge acceptable than those with manner verbs, again consistent with the complexity prediction. () a. The workmen banged the tiles. b. The workmen chipped the tiles.
(manner) (result)
These same sentences were included in a stops-making-sense judgement task where participants were presented each sentence word-by-word, pressing ‘yes’ to read the next word in the sentence until the sentence no longer made sense, at which point they responded ‘no’. While somewhat artificial in terms of the naturalness of reading, this task is thought to tap more deeply into incremental semantic processes. As in speeded sentence acceptability, result verb sentences elicited slower response times compared to
5
In previous work, McKoon & Ratcliff () compared change of location verbs (e.g. descend) with manner of motion verbs (e.g. drift) and found similar results to those reported here. This is predicted given that change of location is a type of result under the wider umbrella of change-of-state (Levin & Rappaport Hovav, ). 6 McKoon & Macfarland () argue that whole-sentence reading times are more sensitive than self-paced reading times to differences of this type when asking participants for acceptability judgements.
()
manner verb sentences. This difference emerged at both the verb itself (~ms) and on the last word of the sentence (~ms). Taken together, these studies support the prediction that the representation of result verbs is more complex than manner verbs. To the extent that these verbs were consistent with only manner or result meanings, the stops-making-sense reading task further supports the idea that this information is accessed immediately during real-time processing, as these complexity costs arose at the verb itself. The additional complexity of result meanings suggests that manner meanings might be the default interpretation during processing. This was examined in a learning study by Naigles & Terrazas (). They presented English speakers and Spanish speakers three training videos showing a motion event (e.g. a woman moving towards a tree while skipping) paired with a novel verb (English: kradding; Spanish: mecando) that appeared in one of three sentence frames (). After this training, participants were shown two more videos. One video preserved the manner but changed the result of the event (‘manner preserving’, e.g. a woman moving away from a tree while skipping), the other preserved the result but changed the manner of the event (‘result preserving’, e.g. a woman moving towards a tree while marching). Participants were asked to choose which of the two videos was consistent with their interpretation of the novel verb, indicating whether they had assigned the novel verb a manner or result interpretation. () a. She’s kradding. Ella está mecando. b. She’s kradding towards the tree. Ella está mecando hacia el árbol. c. She’s kradding the tree. Ella está mecando al árbol.
(neutral) (manner-biased) (result-biased)
Let’s focus first on the English-speaking participants. When presented with verbs in neutral or manner-biased sentence frames, English speakers showed a strong preference for manner preserving videos, suggesting that they had interpreted the novel verb as a manner verb, in line with the prediction that manner verbs would be easier to learn. However, when the verb was in the result-biased frame, English speakers showed no preference for manner or result preserving videos. These results suggest that more than just lexical complexity is involved in manner and result meaning; structural cues are also at play. Considering these structural effects, Mateu & Acedo-Matellán () approached manner/result complementarity as a constraint not on the lexicon but rather one that arises from general syntactic principles. They propose that a verbal root can either adjoin to v as in (a), or initially form a small clause followed by incorporation with v as in (b).7 The first derives a manner interpretation while the second derives a result interpretation, in line with (). From this analysis, manner/result complementarity arises as a simple syntactic fact: a verbal root cannot occupy both positions simultaneously in a single structure. The v’s in (14a) and (14b) may be of distinct flavours, giving rise to DO and BECOME predicates respectively (Folli & Harley, ). 7
.
() a. Manner construction: b. Result construction:
[vP [v √V v ] ] [vP v+√V [SC DP √V ] ]
This analysis suggests that syntactic structure should affect whether a verb is interpreted as manner or as result.8 For instance, verbs like climb have been proposed as counterexamples to manner/result complementarity because they exhibit both manner and result meanings. In (a), a manner akin to ‘clambering’ is attributed to the explorer, but in (b) such a manner is missing; instead, the result of the upward direction of the prices is attested. ()
a. The explorer climbed. [vP [DP the explorer ] [v √climb v ] ]
(manner)
b. The prices climbed. [vP v+√climb [SC [DP the prices ] √climb ] ]
(result)
Levin & Rappaport Hovav () propose that climb and its ilk are polysemous (see Schumacher and Rabagliati & Srinivasan (Chapters and , respectively, in this volume) for further discussion), but Mateu & Acedo-Matellán () observed that this apparent polysemy is structurally conditioned. They note that the presence of a directional Prepositional Phrase (PP) triggers a manner interpretation because this phrase occupies the result argument within the small clause, forcing the verb to adjoin to v.9 In (a), importantly, the ‘upward direction’ result of climb is unavailable and does not clash with the directional PP. Forcing a manner interpretation in (b) results in oddity, perhaps because it is conceptually unclear how prices can ‘clamber’. () a.
The explorer climbed down the mountain. [vP [v √climb v ] [SC [DP the explorer ] [PP down the cave ] ] ]
(manner)
b. ??The prices climbed down the market. Transitive constructions, too, condition the availability of manner and result interpretations. When the direct object is unaffected by the event, no change-of-state happens to it. Therefore, the only option is for the verbal root to adjoin with v. However, if the direct
Of course, if a known verbal root has no possible result meaning but finds itself in a result construction then unacceptability should arise; and similarly if it has no manner meaning but is in a manner construction. Such an account may be appropriate for the unacceptability of *Mary danced her legs and of examples in (5) involving break where both verbs carry only a manner or result meaning, respectively. 9 Given that verbs like climb are ambiguous between these manner and result interpretations, there is a straightforward question about how this ambiguity is resolved during real-time processing. One might expect an initial preference for manner interpretations over result interpretations, either because the manner schema is simpler than the result template (McKoon & Love, ) or because the manner structure is simpler than the result structure (following minimal attachment; Frazier, ). To our knowledge, a study testing this ambiguity has not yet been conducted. 8
()
object is affected by the event, the verbal root must begin its life as the result argument within the small clause and then incorporate with v. Climb in these cases behaves like a manner verb because no change-of-state occurs to, for instance, the mountain in (), as compared with break, a result verb (). ()
a.
()
a. ??The explorer broke the mountain. b. The prices broke the market. (result) [vP [DP the prices ] [v’ [v v+√break ] [SC [DP the market ] √break ] ] ]
The explorer climbed the mountain. [vP [DP the explorer ] [v’ [v √climb v ] [DP the mountain ] ] ] b. ??The prices climbed the market.
(manner)
Given this analysis, it is clear why the sentence frames used in Naigles & Terrazas () had their effect. In English, the path-denoting PP in the manner-biased frames in (b) occupied the result argument within the small clause, thus requiring the verb to adjoin to v and be interpreted as manner. For the transitives in the path-biased frames, however, manner or result depended on whether participants considered the action to have a change-of-state, leading to a mixed result. Turning to the Spanish-speaking participants, they showed a strong preference for manner preserving videos when the novel verb was in the neutral frame, much like English speakers. However, unlike English speakers, Spanish speakers showed a strong preference for result preserving videos in the result-biasing frames, and no preference for the mannerbiasing frames. This difference suggests that cross-linguistic factors modulated the structural manipulations that drove participant preferences. As a Romance language, Spanish is typologically distinct from English in that it typically conflates result rather than manner with its motion verbs (Talmy, , ). Romance languages in general do not allow manner verbs to occur with bounded path PPs, requiring a paraphrastic translation of an English manner verb construction (; from Levin & Rappaport Hovav, ). ()
a. An old woman limped in from the back. b. Une vieille femme arriva en boitant de l’arrière-boutique. an old woman arrived in limping from the.back-store
[English] [French]
Levin & Rappaport Hovav () suggested that languages differ as to which linguistic unit manner/result complementarity holds over. For manner-conflating languages like English, this unit is the verb. Result-conflating languages like Romance and Greek use the Verb Phrase (VP). Mateu & Acedo-Matellán () related this proposal to Mateu’s () proposal that resultative elements such as directional prepositions obligatorily incorporate in result-conflating languages. This restricts the adjunction of a verbal root with v, blocking manner verbs in those languages when a resultative element is present. Result-conflating languages such as Catalan thus allow (b) where the directional component is incorporated into the verb, but not (c) where the resultative/directional component (e.g. fora, ‘out’) is stranded.
.
() a. En Joe escalà. det Joe climbed [vP [DP en Joe ] [v √climb v ] ] b. En Joe sortí del túnel escalant. det Joe went+out of.the tunnel climbing [vP [v v+√foraP ] [SC=PP [DP en Joe ] [P’ foraP del túnel ] ] ] c. *En Joe escalà fora del túnel. det Joe climbed out of.the tunnel
(non-directional interpretation only)
Again, this structural analysis explains the pattern of results reported in Naigles & Terrazas (). Spanish speakers’ strong preference for result preserving videos in the result-biasing frames derives from the obligatory incorporation of an abstract directional element (likely triggered by the oblique a preposition adjoined to the definite article in al árbol, see (c)). This incorporation blocks the possibility of adjoining the verb to v, deriving the result interpretation. The lack of preference in the manner-biasing frame also follows since either manner or result verbs can appear with PPs denoting unbounded paths. Finally, both English and Spanish showed a preference for manner interpretations in neutral contexts. Feist () suggests a pragmatic account. She proposes that, although neutral contexts allow both manner and result verbs, the lack of an overt object reduces the possibility of the verb itself encoding a result because object and result are conceptually connected as the object is the goal of directed motion. Invoking Grice’s Maxim of Quantity, she argues that without overt mention of the object, overt mention of the result is infelicitous. Alternatively, in each of Naigles & Terrazas’s scenarios, the subject is clearly acting as an Agent. From an event structure perspective, Agents are arguments of DO and begin their syntactic lives in the specifier of v. Having no need for a small clause, a manner interpretation emerges as the preferred interpretative choice. Because events are linguistically decomposed, elements like verbs and their surrounding functional structure have to be carefully considered when determining what their individual contribution to the event is. Verbal roots can occupy one of two structural positions in event structure, taking on manner or result meanings. The complexity of these meanings is thus structurally determined, triggering costs during real-time sentence processing and guiding verbal meaning, including the acquisition of new verbal concepts.
. T
.................................................................................................................................. Properties of event structure extend beyond how verbs encode properties like manner and change-of-state. Since Vendler (), aspectual properties that describe the temporal contours of events have been relevant to the grammar and processing of event structure, including dynamicity, durativity, and telicity. Telicity has by far received the most attention
()
in the literature.10 At first blush, telicity relates to whether an event has a natural temporal endpoint or culmination.11 Parsons () introduced a CULMINATE predicate to the theory of event structure to capture this meaning. Kratzer () offers () as a denotation, where R takes the relation between an individual and an event and asserts that the event culminates given this individual.12 ()
⟦ [telic] ⟧ =def λRλxλe [ R(e,x) & CULMINATE(e,x) ]
Theories of telicity have proposed how verbs and their arguments combine together in a VP to determine whether an event description reflects an unbounded (atelic) or bounded (telic) aspectual interpretation, often revealed by the VP’s compatibility with terminative modifiers like in ten minutes that measure the temporal endpoint of the event.13 Verbs themselves, for instance, can contribute directly to telicity. Win is compatible with a terminative modifier but swim is not. ()
a. John won in ten minutes. b. John swam *in ten minutes.
Aspectual interpretations can also be detected by durative modifiers which act as natural temporal measures of unbounded events. In (b), the swimming event lasts for at least ten minutes. Durative modifiers are also compatible with bounded events, but only if such events can be coerced into an unbound temporal frame, a phenomenon called aspectual coercion. Under a durative modifier, win is coerced to require multiple events of winning.14 ()
a. John won for ten minutes. b. John swam for ten minutes.
(multiple events) (one event)
10
On the real-time processing of dynamicity and durativity, see Gennari & Poeppel (); Yap et al. (); Bott (); and Coll-Florit & Gennari (). 11 Building on Jackendoff (), Gawron () argues that spatial aspect shares properties with temporal aspect, suggesting that the notion of endpoint/culmination encoded by telicity is wider than what is being discussed here. See Grigoroglou & Papafragou (Chapter in this volume) for further discussion on spatial uses of language. 12 Kratzer’s (2005) final denotation for CULMINATE in (i) spells out this condition in terms of subparts of the object (under the appropriate measure) with sub-parts of the event. Taking eat the apple as an example, to culminate is to have a sub-event of eating a sub-part of the apple for all sub-parts of the apple: (i) ⟦ [telic] ⟧ =def λRλxλe [ R(e,x) & ∃f [ measure( ƒ ) & ∀x’ [ x’ ƒ (x) ! ∃e’ [ e’ e & R(x’)(e’) ] ] ] ] 13
Results themselves can be distinguished from the temporal boundedness of event telicity (Levin & Rappaport Hovav, ). Some result verbs require telic interpretations (arrive, die, find), but others do not (cool, descend, fall). And non-result verbs can occur in telic predicates (read a book). 14 Bott () reports processing costs for the addition of a process to an achievement, suggesting that durativity can also be aspectually coerced. (i)
a. The climber reached the summit three hours ago. b. The climber reached the summit in three hours.
(no process) (process)
.
This requirement of coercing a bounded event to receive an unbounded interpretation has been the focus of research interested in how aspectual interpretations unfold in real time. Early work by Piñango et al. () investigated the verbal contributions to telicity by focusing on the processing of durative modifiers (e.g. until, for a long time, for hours) when they were preceded by unbounded verbs like glide or bounded verbs like hop15 using a cross-modal lexical decision task. In this task, participants were asked to listen to a sentence like () while watching a blank screen in front of them. At some point during the sentence, a letter string appeared and participants indicated whether the letter string formed a word of English or not. The logic of this task is that making a lexical decision will be more difficult if the participant is currently devoting more processing resources to understanding the sentence. Assuming that the initial mismatch plus repair by aspectual coercion between a bounded event and a durative modifier is a more costly process than a match (and thus no repair) between an unbounded event and a durative modifier, they predicted more difficulty with the lexical decision task when aspectual coercion was required. Sentences like () were presented auditorily with letter string probes presented ms after the offset of the durative modifier. Lexical decisions were an average of ms slower when aspectual coercion was required than when it was not. These results were replicated and extended in Piñango et al. () which again found slower lexical decision times during aspectual coercion ms after the offset of the durative modifier, but not when the letter string probes were presented immediately after the offset of the durative modifier, suggesting that aspectual coercion may be a late repair process triggered by an initial meaning mismatch. ()
The insect glided/hopped effortlessly until it reached the far end of the garden that was hidden in the shade.
Brennan & Pylkkänen () reversed the order of the temporal adverbial and verb found in () to focus on the aspectual contribution of bounded verbs to aspectual interpretation. They manipulated whether the modifier preceding the bounded verb was durative (throughout the day) or non-durative (after twenty minutes), as shown in (). Participants performed a word-by-word self-paced reading task, and made a sensicality judgement at the end of the sentence. They found a ms slowdown on the reading times of bounded verbs that were preceded by durative modifiers compared to non-durative modifiers, consistent with the idea that aspectual coercion is costly. ()
a. Throughout the day the student sneezed in the back of the . . . b. After twenty minutes the student sneezed in the back of the . . .
Following this, Brennan & Pylkkänen investigated the underlying neural sources of this processing cost with a similar design using magnetoencephalography (MEG), a noninvasive neuroimaging technique that measures magnetic fields arising from neural activity. The same sentences were presented to participants word-by-word for a set amount of time 15 Events like hop and sneeze are typically classified as semelfactives, momentary or punctive events (Moens & Steedman, ; Smith, ). This can be compared to achievements like win or reach the summit which appear to denote the boundaries of events (Piñon, ).
()
(ms, with ms blank screen between words) while neuromagnetic fields were recorded over the whole head. They found that the cost of aspectual coercion on the bounded verb was reflected in greater neuromagnetic activity at two different neural sources. The first of these gave rise to a peak in activity occurring ms after the onset of the verb in the right frontal and anterior temporal lobes, and appeared similar to the N, an electrophysiological component thought to reflect lexical semantic access and integration (Lau, Phillips, & Poeppel, ; Kutas & Federmeier, ). The second source, giving rise to activity ms after the onset of the verb in the anterior midline region, was argued to be the same source that is responsible for repair of the meaning mismatch in complement coercion.16 Brennan & Pylkkänen argued that these two neural sources reflected two component processes: first, the composition of an anomalous meaning, and second, the aspectual coercion repair process. This is consistent with Piñango et al. ()’s result of aspectual coercion as a late repair process triggered by an initial meaning mismatch. Paczynski et al. () also investigated the neural sources of aspectual coercion, extending Brennan & Pylkkänen (), by using both bounded (pounce, ) and unbounded (prowl, ) verbs in an event-related potential (ERP) study using electroencephalography (EEG). This technique is similar to MEG, but measures electrical instead of neuromagnetic activity. Bounded verbs elicited an anteriorly distributed sustained negativity beginning ms after verb onset when preceded by durative modifiers (b) compared to nondurative modifiers (a). This sustained negativity ERP was distinct from the N response, but similar to ERPs elicited by complement coercion (Baggio et al., ; but cf. Kuperberg et al., ). Unbounded verbs did not elicit any differential ERP response to durative/non-durative modifiers, which was as expected since no coercion process was necessary. ()
a. After several minutes, the cat pounced on the rubber mouse. b. For several minutes the cat pounced on the rubber mouse.
()
a. After several minutes, the cat prowled about the backyard. b. For several minutes the cat prowled about the backyard.
Given that aspectual coercion is often interpreted as iterating a single bounded event, Paczynski et al. () also included frequentive modifiers as in () which also gives rise to a multiple event interpretation, but through quantification instead of coercion. ()
Several times, the cat pounced/prowled . . .
Bounded verbs preceded by durative modifiers elicited a similar frontal sustained negativity (beginning ms after verb onset) when compared to frequentive modifiers. Paczynski et al. () argued that this similar ERP effect suggests that the cost of aspectual coercion is Complement coercion (e.g. John began the book) is another well-studied case of meaning mismatch repair. In complement coercion, the mismatch between an event-selecting verb (begin) and entitydenoting complement (the book) is repaired by coercing an event from the entity-denoting complement. Like aspectual coercion, this process is costly (Traxler et al., ) and elicits greater neuromagnetic activity in the anterior midline region (Pylkkänen & McElree, ). 16
.
not driven by event iteration but instead by the insertion of an aspectual operator that repairs the meaning mismatch. This operator then generates a multiple event interpretation not reflected by differential ERPs. Taken together, these studies suggest that bounded verbs contribute immediately to the unfolding aspectual interpretation.17 Mismatches between bounded verbs and durative modifiers trigger a repair process that gives rise to an iterative interpretation of the event. Event structure is, however, more than just the meaning of the verb in the sentence. Verkuyl () established that telicity actually ranges over larger structural units. In particular, the countability of the internal argument of a VP plays an important role in determining the telicity of an event, suggesting that telicity is actually calculated over the VP. For a verb like eat, a quantized (i.e. count) internal argument licenses a bounded interpretation while a homogeneous (i.e. non-count) internal argument licenses an unbounded interpretation. ()
a. John ate the apple in ten minutes. b. John ate apples *in ten minutes.
The processing cost found for aspectual coercion of events that are bounded by their verb in the aforementioned studies has also been found for events that are bounded by their internal argument. In a stops-making-sense reading task, Todorova et al. () investigated this compositional component of aspectual interpretation by manipulating whether the internal argument was an indefinite singular (i.e. quantized; b) or a bare plural (i.e. homogeneous; a). If participants rapidly composed the quantity of the internal argument with the verb to calculate aspectual interpretation, we would expect to find the processing costs of aspectual coercion on durative modifiers. Todorova et al. found longer reading times on durative modifiers that followed VPs with quantized internal arguments compared to those with homogeneous ones, evidence that aspectual interpretation takes into account both verbal and non-verbal sources of information during real-time language processing. ()
a. Even though | Howard sent | large checks | to his daughter | for many years, | she refused to accept his money. b. Even though | Howard sent | a large check | to his daughter | for many years, | she refused to accept his money.
Interestingly, the properties of internal arguments do not always appear to contribute to aspectual interpretation. Some verbs appear to block the contribution of their internal argument. VPs with bounded verbs like find seem to always receive a bounded
17 The literature has not been univocal in reporting costs of aspectual coercion. Pickering et al. () attempted to replicate and extend Piñango et al. () and Todorova et al. () using eye-tracking, but found no differences in eye movement measures. Townsend (), however, reported costs for aspectual coercion in first-pass reading times. Husband & Stockall () suggested that this difference is due to the use of frequentive instead of durative modifiers in Pickering et al. (), since frequentive modifiers do not trigger aspectual coercion (following Rothstein’s () event quantification analysis).
()
interpretation regardless of the countability of their internal argument. Similarly, those with unbounded verbs like push appear to always carry an unbounded interpretation. ()
a. John found (the) books in ten minutes. b. John pushed (the) carts *in ten minutes.
The evidence from (a) and (b) suggest a complex relationship between verbs and their arguments in forming aspectual interpretations, with some verbs being inherently specified for telicity and others not. In general, research in linguistics (Mittwoch, ; Rothstein, ; Rappaport Hovav, ) and in the psycholinguistics reviewed earlier has supported the idea of inherently bounded verbs.18 The status of inherently unbounded verbs has, however, proved to be more contentious, with some researchers supporting a class of inherently unbounded verbs (Dowty, ; Krifka, a, ; Verkuyl, ; Tenny, ; Kennedy & Levin, ) and others arguing against an inherently unbounded class of verbs (Schein, ; Borer, b). Stockall & Husband () took an experimental approach to this question. They varied the three possible verb classes (inherently bounded, inherently unbounded, unspecified) with a quantized or homogeneous plural direct object in two word-by-word self-paced reading studies. In the first study pairing inherently bounded verbs (lost) with unspecified verbs (read) as in (), they found a slowdown in reading times on the first word after the direct object noun when an unspecified verb was followed by a bare plural compared to a definite plural. This slowdown was not present when the verb was inherently bounded. These findings suggested that participants combined unspecified verbs with the homogeneous property of bare plurals to arrive at an unbounded interpretation, slowing down processing perhaps because unbounded interpretations are more semantically complex than bounded ones (Husband & Stockall, ). The other three conditions did not differ in reading times since all three yielded bounded interpretations. In particular, inherently bounded verbs ignored the homogeneous property of bare plurals, behaving like the other bounded interpretations. ()
The expert physicist lost/read (the) files on the formation of black holes.
In the second study pairing inherently unbounded verbs (roam) with unspecified verbs (inspect) as in (), they found a slowdown in reading times on the first word after the direct object noun when either unspecified or unbounded verbs were followed by a bare plural. The same slowdown was not, however, present for unbounded verbs followed by a definite plural. Instead, reading times were similar to those of unspecified verbs. Given that unbounded and unspecified verbs patterned together, Stockall & Husband () argued against a distinction between inherently unbounded verbs and unspecified verbs, suggesting that they, in fact, form a single class of unspecified verbs since both cases showed sensitivity to the countability of their internal argument during real-time processing.
18 Verkuyl () stands out as one case of an argument specifically against inherently bounded verbs, suggesting that those that appear to be inherently bounded are actually those that world knowledge specifies as having punctual/momentary durations.
.
()
The local horticulturalist roamed/inspected (the) gardens in the neighborhood.
Such a finding is in line with Borer (b), who also argued against a class of inherently unbounded verbs. She noted that the classically cited cases for inherently unbounded verbs like push act like bounded events in sentences where world knowledge permits a bounded interpretation. As shown in (), pushing a cart, for instance, has no natural endpoint in the real world and behaves like an unbounded event with respect to durative modifiers, whereas, pushing a button does have a bounded sense in the real world and thus behaves like a bounded event with respect to durative modifiers. She argued that without real-world support for a bounded interpretation, only an unbounded interpretation is possible. () a. Beth pushed a cart for ten minutes. b. Beth pushed a button for ten minutes.
(one event) (multiple events)
Properties of events such as telicity further demonstrate how the relationship between verbs and their surrounding elements must be pulled together to properly interpret the event. Telicity is sensitive to temporal adverbs that measure the duration and endpoints of events, coercing the underlying event when possible to achieve a coherent aspectual interpretation. Arriving at an aspectual interpretation is not without cost, and is detectable in a variety of different experimental techniques that shed light on its time course. Such techniques can also give us a handle on the aspectual properties of verbs. Through their similar processing behaviour in terms of their relationship to their internal argument, inherently unbounded verbs and unspecified verbs appear to form a natural class, albeit one where the ultimate interpretation of the event is still beholden to our knowledge of the real world.
. P ()
.................................................................................................................................. The arguments that enter into events do more than establish its temporal contours; they also find interpretations of their own within event structure.19 Any given event description has a multitude of associated and entailed components. Events are identified in general by their time and location, they often but not always involve some objects or persons, and they can be highly idiosyncratic from verb to verb; for example, every event of dribbling in basketball involves a player, a ball, and at least one bounce. Natural language predicates, however, encode only some of these components as grammatically privileged.20 These 19
Beyond what is discussed here, another important aspect of the interpretation of the arguments of an event is that they may be interpreted collectively or distributively, which has further consequences for the real-time processing. See in particular Syrett (Chapter in this volume) for the effects of distributive items such as each on the interpretation of event structures. 20 Williams () proposes that privileged event participants form a representation he calls a sketch. We might think of a sketch as those participants of an event that are given by default in cognition. Such a representation appears to underlie behaviour in studies on event perception. Wellwood et al. () used
()
privileged components are known as event participants. One common theoretical device used to capture different event participants is the notion of thematic roles, an intuitive set of roles that identify the relationship of an event participant to the event (Fillmore, ). For example, Agents perform the action of the event, Patients undergo this action, and Instruments are used to assist with the action. These thematic roles are encoded to some extent, in poorly understood ways, by different verbal concepts. A verb like jab, for instance, seems to encode an Agent, Patient, and Instrument while other roles such as Location and Time are not explicitly encoded though they may still be expressed. Similar verbs like attack appear to encode fewer roles, only Agent and Patient. ()
[The pygmies]AGENT jabbed/attacked [the lion]PATIENT [with a spear]INSTRUMENT [in the field]LOCATION [yesterday]TIME.
Researchers since Jackendoff () and Dowty () have observed that thematic roles are intimately bound up in the representation of event structure, suggesting that they may derive from event structure itself. The notion of Agent or Patient, for instance, might be more elegantly captured by appealing to the roles they play in event structure, that is as the causers of a sub-event in (a) or the undergoer of a change-of-state in (b) (Levin & Rappaport Hovav, ; Van Valin, ; Schein, ).21 This implies that the elements that realize event structure are also implicated in the way that event participants themselves are linguistically encoded. ()
a. λxλe[ Agent(e,x) ] =def λxλe2[ ∃e1 [ DO(e1,x) & CAUSE(e1,e2) ] ] b. λxλe[ Patient(e,x) ] =def λxλe1[ ∃e2 [ BECOME(e1,e2) & root(e2,x) ] ]
a similarity judgement task in which participants viewed two videos depicting different events described in (i,ii) and rated them on how similar they were to one another. The first video showed three participants, the second either the two core participants or three participants where the third was merely associated to the event. Wellwood et al. reported that giving vs. hugging (i), where giving requires three arguments while hugging only requires two, showed significant differences in their response times. The other conditions which could be described with either a three-participant (jimmy, steal, bean) or twoparticipant (open, pick-up, hit) verbal concept also showed differences in either response times (iia,b) or similarity ratings (iib,c), suggesting that participants that are engaged in the event are reflexively highlighted in a sketch-like representation, even if they are not frequently expressed with an overt constituent. (i) Anne gave a teddy bear to Beth vs. Anne hugged Beth (while holding a teddy bear). (ii) a. Anne jimmied the box (with a screwdriver) vs. Anne opened the box (while holding a screwdriver) b. Anne stole the box (from Beth) vs. Anne picked up the box (while Beth stood by). c. Anne beaned Beth (with a ball) vs. Anne hit Beth (while holding a ball). 21 Instruments are also linked to event structure. Pylkkänen () notes that Instruments (as a type of High Applicative in her theory of argument structure) are related to CAUSE. Instrumentals are, for instance, prevented from occurring with unaccusatives like (ic) which lack a CAUSE in their representation.
(i) a. John broke a window with a stone. b. A window was broken with a stone. c. *A window broke with a stone.
.
Agent participants have been the focus of much of the experimental work on event structure. Many events are initiated and controlled by Agents, but Agents are not always overtly expressed in a sentence. For instance, every selling event has a seller, but sentences describing a selling event need not explicitly say who that seller was. Passives are one linguistic device used to demote an Agent. In their full form, like (a), the Agent is still overtly expressed in a by-clause, but passives can omit this by-clause, leaving the Agent unexpressed, as in (b). These short passives appear to be superficially similar to an intransitive clause like (c) which also lack an overt Agent. ()
The collector sold the vase immediately.
()
a. The vase was sold by the collector immediately. b. The vase was sold immediately. c. The vase sold immediately.
(full passive) (short passive) (intransitive)
Although it is tempting to treat short passives and intransitives in the same way, there are formal differences between them that correlate with the encoding of an Agent. To probe for the presence of an Agent, we can take advantage of rationale clauses, which are only acceptable if an Agent is present, whether explicit or implicit.22 Rationale clauses distinguish short passives from intransitives, suggesting that only short passives encode Agents implicitly. ()
a. The vase was sold immediately to raise money for charity. b. #The vase sold immediately to raise money for charity.
These different formal cues are rapidly used during real-time processing when encoding Agents. Following initial research in Mauner et al. () that used this property of rationale clauses to detect the presence of unexpressed Agents, Mauner & Koenig () investigated the processing of implicit agents using a stops-making-sense judgement task. In their second experiment, they presented participants with either short passive or intransitive clauses followed by a rationale clause as in (). In both conditions, an Agent of some kind was implicated. The main clause (including the adverb) was presented in its entirety, followed by the rationale clause read word-by-word. () a. The car’s engine was started abruptly to test the adjustments that the mechanic had made to the fuel pump. (short passive) b. The car’s engine had started abruptly to test the adjustments that the mechanic had made to the fuel pump. (intransitive)
22
McCourt et al. () argue that anaphora in rationale clauses may be pragmatically mediated, with its reference being sensitive to a notion of responsibility, a concept that is often related to that of an Agent. For the following studies, the particular licensing conditions of rationale clauses are not directly relevant, so we will set these issues aside.
()
Mauner & Koenig reported disruption in reading for the intransitive compared to the short passive condition. Participants made significantly more ‘no’ judgements starting on the verb of the rationale clause (% vs. %) which rapidly accumulated over each subsequent word. By the final word of the rational clause, intransitives had elicited a cumulative . per cent ‘no’ responses compared to per cent for short passives. This disruption was also reflected in ‘yes’ response times. Although some participants had not yet indicated that the sentence had stopped making sense, ‘yes’ response times were significantly slower at the initial to of the rationale clause for intransitives compared to short passives, a trend that continued throughout the clause. In their third experiment, Mauner & Koenig reversed the order of rationale clause and main clause as in (). Participants again performed a stops-making-sense task, reading these sentences word-by-word. ()
To prevent the champ from killing his rival, a. the fight was ended in the fifth. b. the fight had ended in the fifth.
(short passive) (intransitive)
Reading disruption was again found in the intransitive condition compared to the short passive, with significantly more ‘no’ judgements elicited at the main verb (.% vs. %). Reading times of ‘yes’ responses also slowed somewhat, eliciting an ms difference at the postverbal position and ms difference on the final word. Mauner & Koenig reasoned that readers rapidly encoded an implicit agent in short passives, enabling them to license the rationale clause early in processing. This effect appeared immediately on the verb in their third experiment, suggesting that participants made immediate use of the structural cues for an implicit Agent. Given that the events encoded by their verbs implicated an Agent participant, their findings suggest that linguistically relevant event structures are responsible for these effects, and not some extra-linguistic conceptual or real-world knowledge. Some initial evidence that Patients are also represented as event participants by some verbs was reported in Carlson & Tanenhaus (). They took advantage of the requirement that definites need a discourse antecedent for felicitous use or must otherwise be accommodated into the discourse. Assuming that accommodation is a costly process (Haviland & Clark, ), they hypothesized that a context that provides a suitable unfilled event participant role would allow for a more rapid accommodation of a definite than one that merely set up a likely scenario for the definite. They asked participants to read one of two context sentences which either introduced an implicit Patient (a) or did not (a’) and then judge whether a following sentence that began with an unfamiliar definite made sense in the given context. () a. Bill hurried to unload his car. a0 . Bill hurried to catch his plane. b. The suitcases were very heavy.
(implicit Patient) (no implicit Patient)
Judgements for the target definite sentences were ms faster in implicit Patient contexts compared to no implicit Patient contexts. Participants also rejected target definite sentences less often in implicit Patient contexts, with per cent judged to make sense, while
.
only per cent were judged to make sense in no implicit Patient contexts. This result suggested that implicit event participants can guide definiteness accommodation, allowing for a more rapid integration of discourse that maintains coherence between sentences. Sluicing, an ellipsis construction that involves an overt wh-phrase, offers another way to probe for the presence of event participants. Frazier & Clifton () compared encoded event participants to those that are left as simply entailed by the verb. An event of writing, for instance, encodes whatever is written as a Patient, but also entails something used to write with as an Instrument. Frazier & Clifton () manipulated the wh-phrase that introduced the elided material of the sluice to target either the Patient () or an unencoded role (including Instrument, Goal, Location, or other oblique; ). The antecedent was either overtly mentioned with a some indefinite (a, a) or left unexpressed (b, b). ()
a. Kathy wrote | something | but nobody | seems | to remember what. b. Kathy wrote | but nobody | seems | to remember what.
()
a. Kathy wrote with something | but nobody | seems | to remember with what. b. Kathy wrote | but nobody | seems | to remember with what.
They found that overt antecedents were read faster than implicit antecedents by a lengthcorrected average of ms. Additionally, encoded patient sluices were faster than unencoded argument sluices, but this difference was not statistically reliable (~ length corrected ms; F = ., F = .). These results were replicated by Dickey & Bunger (), who reported a statistically reliable difference in Patients compared to unencoded arguments (~ms).23 Together, these studies suggest that the processing of sluicing is sensitive to event participant status, even in a technique like self-paced reading which does not require participants to make sensicality decisions at each word. Finally, the integration of arguments within a sentence is guided by their event participant status. Koenig et al. () reported a study that examined whether Instrument participants were more easily integrated into sentences with verbs that encode Instruments as event participants. Events of beheading require an Instrument to be used whereas events of killing are more general (and may require no Instrument at all). Koenig et al. () hypothesized that events requiring Instruments like (a, a) would more quickly integrate an Instrument into a gap than those that do not (b, b). ()
a. Which sword | did the rebels | behead | the traitor king with | during the rebellion? b. Which sword | did the rebels | kill | the traitor king with | during the rebellion?
()
a. With which sword | did the rebels | behead | the traitor king | during the rebellion? b. With which sword | did the rebels | kill | the traitor king | during the rebellion?
23
Dickey & Bunger () also compared sluiced sentences to ellipsis like (i) and found similar effects of event participant status on reading times in the final region (~ms). (i)
a. Kathy wrote (with) something | but nobody | seems | to remember (with) what she wrote. b. Kathy wrote | but nobody | seems | to remember (with) what she wrote.
()
For both wh-NP () with-wh-NP () sentences, they found that reading times on the direct object (+with) were faster with instrument-encoding verbs compared to no-instrumentencoding verbs (NP filler: ~ms; PP filler: ~ms), suggesting that verbs that encode Instruments as event participants guide the rapid integration of displaced Instrument arguments during real-time sentence processing.24 Arguments and their interpretations are tied up with event structure. Even when unexpressed, the cues to event structure, either formal or via verbal concepts, can lead to an implicit encoding of event participants. These implicit participants are accessed and used in real-time sentence processing to guide interpretative processes that put an event together incrementally.
. S
.................................................................................................................................. In this chapter, we have looked at the consequences that different aspects of event structure have for real-time sentence processing. Because events are decomposed and articulated in constituents across a sentence, cues to their disparate components must be recognized and put back together to construct a complete and coherent representation of the event under discussion. This is made all the more complex during real-time processing as these components arrive one after another in quick succession, and yet studies show that speakers are highly sensitive to the right cues, able to rapidly extract them from individual units, and use them to guide interpretation in a highly incremental fashion. For verbal roots, event structure constrains their interpretation in terms of expressing manner or result. The complexity of result meanings involves additional processing compared to a manner meaning, but is structurally conditioned, sensitive to formal cues and crosslinguistic differences. Verbal roots also participate in establishing the temporal contours of an event structure in accordance with their inherent aspectual properties, if they have any, and properties of their arguments. This initial aspectual interpretation may go on to enter into further aspectual elaboration, at times requiring a resolution of mismatched aspectual meanings by costly coercion processes reflected in behavioural and neural measures. Finally, verbal roots and the formal structure surrounding them license the interpretation of different participants in event structure. Idiosyncratic encodings by verbal roots combine with formal cues to mark the presence or absence of implicit event participants, including Agents, Patients, and Instruments, regardless of whether they are implicated by the event or not. Encoding of these implicit event participants is triggered 24 Of course, other extra-linguistic sources of information may be driving Koenig et al.’s () effects. Stowe et al. () reported faster self-paced reading times for sentences with plausible wh-NPs compared to implausible wh-NPs at the verb (~ms).
(i) a. The teacher wondered which book the students read during their class. b. The teacher wondered which song the students read during their class. Reading times, therefore, can be influenced by factors other than those that are linguistic in nature. It could be that expectations of or co-occurrences between the wh-NP and the verb sped processing through prediction or priming. As with simple acceptability judgements, careful manipulation and control of the stimuli are necessary to understand the source of any processing cost.
.
immediately by these verbal and structural cues, and such encoded participant roles go on to license rationale clauses and speed the processing of unfamiliar or displaced arguments. Throughout our discussion of various experiments we have noted differences between more passive implicit tasks, like self-paced reading and electrophysiology, and those like stops-making-sense reading tasks which require more active decision-making on the part of experimental participants but prove to be rather robust techniques for detecting incremental interpretative processes. Whether the differences between these techniques affects the underlying conclusions of any one study is always up for debate; however, convergent evidence from multiple techniques resulting in relatable findings puts us on a surer footing. As a matter of general experimental strategy, initial processing cost results in self-paced reading and other behavioural tasks warrant further investigation with more fine-grained but experimentally complex techniques such as eye-tracking and electrophysiology, which may tap into the time course of sentence processing in such a way that behavioural techniques find difficult to discern. Given our overview, further experimental research on different aspects of event structure holds much promise. Linguistic theories have laid a solid foundation for understanding how event structure is decomposed and represented in language. Building upon this, experimental research has endeavoured to bring to light the details of the consequences of this decomposition during real-time sentence processing. Going forward, these techniques offer potentially new and interesting ways to investigate how these structures are put together in real time, and perhaps even help to decide between competing theoretical perspectives by investigating shared processing behaviours. Such approaches hold promise for establishing a deeper connection being grammatically determined aspects of sentence meaning and the interpretative processes that guide its use online.
A Many thanks to Lisa Levinson and those who attended the LSA Summer Institute course ‘Experimental Approached to Verb Meaning’ for their thoughtful discussion on many of the issues raised in this chapter. We also thank the editors for their helpful feedback on previous drafts.
......................................................................................................................
, , ......................................................................................................................
. C
.................................................................................................................................. A central issue in the study of meaning in natural language is to capture what the different aspects contributing to the ‘overall conveyed meaning’ of an utterance are. This requires both establishing diagnostics for identifying separate ingredients of meaning and characterizing their properties, including their source and status. As such, this enterprise is right at the nexus of semantics and pragmatics: many crucial issues relate to the question of whether a given piece of meaning is encoded conventionally or has its source in reasoning about the use of specific expressions, potentially based at least in part on more domaingeneral resources not specific to linguistic knowledge. The traditional starting point involves distinctions between conventionally encoded truth-conditional meaning, conversational implicatures, conventional implicatures, and presuppositions, following seminal work by Strawson () and Grice (), among others. But many current debates are concerned with the potential need to refine these distinctions or redraw the lines between aspects of meaning. While this has given rise to various important and novel theoretical approaches, the crucial development for present purposes is that, in recent years, researchers have increasingly turned to experimental methods from psycholinguistics to shed light on theoretical questions in this area, both to obtain solid data on subtle phenomena and theoretical predictions that are hard to assess through introspection and to understand how abstract characterizations of linguistic knowledge relate to real-time cognitive processes in language comprehension. The current chapter surveys what has been achieved in this regard in the study of presuppositions in particular, both in terms of methodological developments and the theoretical implications of experimental results.
Presuppositions convey information that is typically assumed to already be taken for granted by the discourse participants. Furthermore, they are characteristically unaffected by a variety of linguistic embedding environments, such as negation, conditionals, and questions. Given this direct relation to the discourse context as well as the interaction with their intra-sentential linguistic environment, presuppositions provide a particularly useful perspective on the interplay of linguistic and domain-general processes in language comprehension. The chapter is structured as follows: the remainder of this section introduces some basic background and provides a sketch of key issues in the current literature. These concern debates about the semantic vs. pragmatic nature of presuppositions and their interaction with context, as well as the potential need for distinguishing sub-classes of presupposition triggers. Next, we turn to experimental approaches to investigating presuppositions experimentally, beginning with triggers in simple, non-embedded linguistic contexts and their relation to the discourse context. Section . turns to issues relating to ‘projection’ phenomena for triggers in embedded contexts. In section ., we review recent developments concerning the relation between presuppositions and more elaborated aspects of the discourse context. Section . concludes and provides an outlook on future directions for the field.
.. Basic phenomena The modern notion of presupposition traces back to Frege (), who considered the existence of an entity with the property of a noun phrase within a definite description a ‘Voraussetzung’, that is, a pre-condition or presupposition for the sentence containing the expression to be meaningful. While definites gained further importance through the famous debate between Russell () and Strawson (; see Reimer & Bezuidenhout, , for a collection of recent contributions on the topic), the literature on the topic soon recognized that the key phenomena were paralleled by a host of other expressions (Karttunen, , ; Stalnaker, , ), for example: factive verbs (e.g. know), aspectual verbs (e.g. stop and continue), iteratives (e.g. again and too), and clefts. In particular, all of these expressions contribute something to the overall meaning that is backgrounded and, at least in many cases, taken for granted by the discourse participants already. Consider the following examples. ()
a. Sue invited John again. b. Mary has stopped throwing parties. c. It was James who sent Anna the invitation.
The main point (commonly also referred to as ‘at-issue’ or ‘proffered’ content) of (a) would seem to be that Sue invited John, whereas the information introduced by again, that she had done so previously, is backgrounded, or presupposed. Similarly, the main point of (b) is that Mary doesn’t throw parties at present, whereas the notion that she used to do so in the past is presupposed. Finally, the cleft construction typically is associated with an existential presupposition to the effect that the relative clause property holds of some individual; the main point, of course, is that James is the culprit in the present case (see Onea, Chapter in this volume, for more details).
, ,
The backgrounded nature of presuppositions is reflected in various other empirical facets. First, there is a common perception, going back to Frege’s and Strawson’s discussions of definite descriptions, that sentences whose presupposition is not met are neither true nor false, that is, they can’t straightforwardly be assigned a truth-value in intuitive terms. To illustrate, ask yourself whether (b) is true or false if considered in a context where it is known that Mary has never thrown a single party. Another related property of presuppositions is that they cannot generally be used to introduce new information. For example, it would be decidedly odd (and certainly seem highly presumptuous) if I casually uttered () in conversation; similarly, () does not make for a good way of breaking the news of one’s engagement to one’s parents. ()
President Obama called me again last night.
() I’m going to Hawaii with my fiancé next week. A common reaction to such pronouncement might well start with ‘Hey, wait a minute, I didn’t know that . . . ’ (see Von Fintel, , for discussion of the ‘Hey, wait a minute test’, inspired by Shanon, ), reflecting the sense that crucial and important information has been introduced in an inappropriate, underhanded way. But while it often seems impossible, or infelicitous, to use presupposed material to introduce new information, there are plenty of cases where this seems perfectly natural. Following Lewis (), these are thought to be cases of ‘accommodation’, which is typically seen as a type of repair mechanism (see Von Fintel, , for detailed recent discussion). For example, I can easily utter () to someone that I just met, and who doesn’t knows anything about my home life: () I’m sorry I’m late, I had to take my cat to the vet. The notion that I have a cat is standardly assumed to be introduced as a presupposition here, by virtue of the possessive description my cat; and yet, despite being completely new information, there is no sense of anything going wrong by treating this information as ‘taken for granted’. But note that various details, such as the fact that having cats as pets is common and that I should be an authority on what pets I have, play an important role (change cat to elephant, and this becomes much harder to swallow). Various intricate issues arise in trying to flesh out the factors at play. For recent discussions, see Beaver (), Simons (), Beaver & Zeevat (), and Von Fintel (). The fact that presuppositions are not part of the main point of the utterance also shines through when considering simple denial responses: responding No (, that’s not true!) to any of the examples ()–() will generally challenge the truth of the main point (Obama having called me last night, going to Hawaii, having gone to the vet), but typically not question the presupposed information—doing so requires more elaborate replies (Wait, that can’t be true—you don’t even have a cat!). The point about presuppositions not being the target of denials has a much more general correlate in sentences where presupposition triggers appear in the scope of embedding operators, which is commonly taken to be another (if not THE!) hallmark feature of presuppositions. In a wide range of such embedding environments which prevent the main at-issue content from contributing to the conveyed content at the matrix level,
presupposition-based inferences remain intact; that is, they appear to not be affected by the embedding—a phenomenon commonly referred to as ‘presupposition projection’ (Langendoen & Savin, ; Karttunen, , ). The following selection from the ‘family of sentences’ from Chierchia & McConnell-Ginet () illustrate this (see their discussion for the full set of environments): ()
a. Sue didn’t invite John again. b. Did Mary stop throwing parties? c. If it was James who sent Anna the invitation, then he should sit next to her.
Negation, questions, and conditionals are environments where the at-issue content of the clauses in question no longer is conveyed by the utterance as a whole: we would not conclude from these sentences that Sue invited John this time around, that Mary doesn’t throw parties at present, or that James sent Anna the invitation. These pieces of information are precisely what is negated, questioned, and considered as a candidate state of affairs respectively. But the presuppositions (that Sue invited John before, that Mary used to throw parties, that someone invited Anna) remain untouched as part of the overall meaning; that is, they are seemingly unaffected by the embedding altogether. However, things are more complex than that, as there are other cases where the perceived global presence of the relevant inferences depends on other content within the sentence: ()
a. If Sue is hosting the party, then it was James who sent Anna the invitation. b. If Anna received an invitation by mail, then it was James who sent her the invitation. (He’s the only one who knows her physical address!)
While the presupposition that someone invited Anna is clearly present for the sentence as a whole in (a), this is not so for (b), presumably because of the content of the antecedent clause, which is the only part changed from (a), and which clearly bears on the issue of whether or not someone sent her an invitation. The problem of explaining the interaction of embedded presuppositions with their linguistic context has constituted one of the core challenges in the theoretical literature, and we will briefly review some of the main approaches below.
.. Semantics vs. pragmatics and explanatory challenges To capture the empirical phenomena concerning presuppositions that were sketched earlier, a classical truth-conditional approach to linguistic meaning (as laid out, for instance, in Heim & Kratzer, ) has to be enriched to allow for the separation of different components of meaning in one way or another. This can be done either by handling presuppositions at a pragmatic level, or by amending the semantics. Some of the earliest approaches (based on Kleene, ) chose the latter route, by supplementing accounts based on classical logic with a third truth-value (standardly labelled ‘#’) to represent the status of sentences whose presupposition is not met as being neither true nor false (for recent proposals in this tradition, see Beaver & Krahmer, ; Fox, ; George, b; as well as supervaluation-based variants, going back to Van Fraassen ).1 1
Parallel formal options also have been explored in the study of Vagueness (Alxatib & Sauerland, Chapter in this volume); see Zehr () for a unified approach to the two phenomena.
, ,
An alternative pragmatic approach was first offered by Stalnaker (, ), which is based on the notion of the Common Ground that represents beliefs mutually held by the discourse participants for the purposes of conversation. The point of assertions is to add information to the Common Ground, while the role of presuppositions is to indicate that the presupposed information already is in the Common Ground. If it is not, the context update fails (which mirrors the third truth-value of semantic approaches). While this notion of presupposition is fundamentally pragmatic—speakers of utterances, rather than sentences, presuppose—Stalnaker leaves open the possibility that certain expressions could nonetheless be conventionally associated with presuppositional uses (see also the notion of ‘Stalnaker’s Bridge’ in Von Fintel, , as well as further discussions in section ... on the issue of how presuppositions come to be). Influential work by Heim (a, ) initiated the enterprise of dynamic semantics, which provides an integrated perspective on intra-sentential semantics and the discourse context, effectively incorporating Stalnaker’s approach of modelling the contribution of (utterances of) sentences in terms of their impact on the discourse context. On this view, the semantics of sentences is characterized as their context change potential (which can be straightforwardly related back to more standard truth-conditional characterizations). Presuppositions can then be seen as conventionally encoded constraints on the types of contexts for which a given expression allows updates. A similar approach was also developed within Discourse Representation Theory (DRT; Kamp, a; Van der Sandt, ), which provides a representational framework for modelling the discourse context. Presuppositions here are seen as a type of anaphora that directly links back to information already present in the existing Discourse Representation Structure (DRS). While, for some time, dynamic approaches dominated the discussion and were assumed to be successful in particular because of their successful account of (at least the majority of) projection phenomena, recent years have seen a revival of various alternative approaches. The chief motivation have been challenges to the explanatory adequacy of standard dynamic accounts. The remainder of this section turns both to these issues and a brief sketch of the alternative paths pursued in the literature to meet the newly raised challenges. The two fundamental questions raised in this debate are a) Where do presuppositions come from? (the triggering problem); and b) Why do presuppositions exhibit their characteristic projection behaviour?
... The triggering problem On purely semantic accounts of presuppositions, they form part of the conventionally encoded meaning of the expression in question. Such accounts are therefore not particularly concerned with the question of where presuppositions come from. Typically, they content themselves with being able to diagnose a given piece of meaning as presuppositional in virtue of its projection behaviour, and then posit it as part of the lexically encoded meaning. Accounting for the projection behaviour is then seen as the key challenge. But a case can be made that simply positing presuppositions as part of lexical entries falls short of the explanatory adequacy we should strive for. How come, one might ask for example, that we find correlates of so many presupposition triggers across languages without any variation in what is presupposed and what is asserted? Take stop and start, which seem to have straightforward correlates across many, if not all, languages: they all presuppose the preceding state or activity and assert what is happening now. This hardly can be a
coincidence, but simply stipulating the presupposition of stop, start, and their correlates in their lexical entries fails to explain this apparent uniformity. Various pragmatic approaches share the central goal of meeting this challenge in explanatory adequacy, at least for a sub-class of triggers, starting with remarks by Stalnaker (; for a more extensive review of these accounts, see section of Abrusán, , which the present discussion follows loosely). The general idea is that triggers such as stop and know respectively come with distinct entailments (e.g. that x used to V and that x V’s now; that x believes p and that p is true), and that, in Stalnaker’s words, a speaker ‘would be leaving unclear whether his main point was to make a claim about’ one or the other of these entailments. The challenge then is to provide a systematic account of how one of these entailments winds up as backgrounded or presupposed, without alluding to any conventional encoding of such notions. Sperber & Wilson () attempt to do so in terms of focus; more recently, Simons () proposes that the truth of a proposition q winds up as presupposed when an utterance raises the question Whether p and p asymmetrically entails q. Abusch (, ) bases her account on the notion that presupposition triggers are associated with alternatives (see Skordos & Barner, Chapter in this volume), and that presuppositionality results from reasoning over these alternatives (for a similar approach, based on different assumptions about alternatives, see Romoli, ). Simons et al. () put forth a more general novel approach to ‘projective meaning’ (including conventional implicatures), which relates presuppositionality directly to atissueness in terms of the relation of content to the Question Under Discussion (QUD; Roberts, ). Finally, the most recent approach has been laid out by Abrusán (, ), who frames the foreground-background distinction in terms of general principles of attention. All of these proposals face substantial challenges (again, see section of of Abrusán, , for a brief review) or are still in active development, so the jury is still out on what line of attack proves to be most successful, if any. Also note that, in principle, semantic approaches positing conventional encoding of presuppositions could be reconciled with certain versions of these approaches by grounding the conventional encoding as the result of pragmatic pressures leaving an impact over time, that is, seeing conventionalized presuppositions as fossilized versions of ever-present pragmatic forces (there is little work pursuing a historical perspective on presuppositions, but see Beck & Gergel, , for a recent example). For present purposes, one particularly important aspect of this debate consists of the fact that many of the pragmatic approaches liken the process by which a presupposition (or the presuppositional nature of a given proposition) is derived to other pragmatic inferences, in particular conversational implicatures. These, in turn, have been the focus of the first major wave of research in experimental pragmatics (see Breheny, Chapter in this volume), which opens up a variety of avenues for exploring potential predictions of presupposition theories experimentally. These will be discussed in the following sections.
... The projection problem As already noted, explaining presupposition projection patterns has long been taken to be the main challenge for a theory of presuppositions. Just like in the case of the triggering problem, there are proposals on both sides of the semantics-pragmatics divide. Once more, Stalnaker’s seminal work provided the starting point for grounding projection phenomena in the dynamics of linguistic interaction. To illustrate, take another variation on a pair of complex sentences where the presupposition in the second clause acts differently:
, , ()
a. Sue is hosting the party and it was James who sent Anna the invitation. b. Anna received an invitation by mail and it was James who sent her the invitation.
Stalnaker’s proposal boils down to the idea that when assessing the second conjunct here, the relevant context is that established by adding the first conjunct to the Common Ground. This makes all the difference given the initial conjuncts: in (a), Sue hosting in no way addresses the issue of whether someone sent Anna an invitation. Therefore, this presupposed proposition has to be met by the global context, as reflected in the intuitive sense that the sentence as a whole is associated with the presupposition. In contrast, the first conjunct in (b) entails that someone sent Anna an invitation, and thus it is guaranteed, no matter what the global context is, that the presupposition of the second conjunct is met. Stalnaker sketches the general direction of how this approach can be extended to other embedding environments. But this was made much more concrete by dynamic semantic approaches, which take the same general perspective on context change but incorporate it directly into the semantics. This is straightforward for conjunction, but also extends to other operators. For example, for conditionals (assuming a material implication analysis for simplicity), context update involves removing those worlds from the initial context where the antecedent is true and the consequent is false. Technically speaking, updating a context c with a conditional If p then q proceeds as follows:2 () c + (If p, then q) = c
((c + p)
((c + p) + q))
Crucially, the consequent q is interpreted relative to the original context updated with the antecedent (c + p), rather than just the original context c alone. This makes the account of the fact that (a) imposes no constraints on c entirely parallel to the conjunction case earlier: c + p is a subset of p (and thus entails the presupposition of q, that someone invited Anna), and it is thus ensured that update with q cannot fail.3 While dynamic semantic approaches enjoyed the status of being the standard and were essentially seen as a successful solution to the projection problem, they face an explanatory challenge that was first raised by Mats Rooth (in a letter to Irene Heim) and Soames (; also discussed by Heim, ), and gained prominence through recent discussion by Schlenker (a, ). In a nutshell, there is no inherent theoretical reason why the various operators should map onto context updates that reflect the linear order in the sentences. For example, nothing prevents an alternative formulation of context change potentials, say for conjunction, that reverses that order: ()
a. Usual context change potential for conjunction: c + p AND q = (c + p) + q b. Alternative context change potential for conjunction: c + p AND q = (c + p) + p
2 ‘+’ represents the operation of context update, which in the simplest case amounts to set-theoretic intersection of propositions construed as sets of possible worlds. Similarly, ‘ ’ stands for the settheoretic difference. 3 DRT offers an alternative dynamic approach that is largely parallel in empirical coverage, though with some differences which also relate to experimental results discussed in section ...
While the usual version successfully captures the projection phenomena in (a) and (b) above, the theoretically possible alternative version in (b) would make different predictions, namely that projection should operate on the reverse of the linear order.4 Clearly, lexical entries corresponding to these alternatives are not attested in any language, but, the explanatory challenge goes, dynamic semantics has no explanation to offer for this. One highly influential response to this problem has been developed by Philippe Schlenker, in various versions (Schlenker, a, , b; for an introductory overview, see Schlenker, ). The basic move boils down to combining a classical, non-dynamic semantics with Stalnaker’s pragmatic approach to projection, while maintaining the predictions of a Heim-style dynamic semantics. In doing so, the explanatory challenge is resolved through an appeal to the role of the incremental unfolding of spoken language in comprehension. At the same time, the full generality of the system in terms of coverage of operators (as well as quantifiers) that Heim’s system achieved is maintained. While Schlenker’s account is based on a specific, formalized notion of Local Contexts, resting on an idea of efficient interpretation where only possibilities that are crucial are considered along the way, there are various other similar approaches couched in multivalent semantic systems (Kleene, ; Peters, ; Beaver & Krahmer, ; Fox, , b; George, a,b). Trivalent approaches provide interpretations for connectives via non-classical truth-tables. The Strong Kleene version of such a truth table (Kleene, ) posits that a complex sentence gets the value ‘#’ just in case one of its atomic sentences has that value and the truth-values of the other atomic statements do not suffice to determine the truth-value of the entire sentence based on standard logic (for a recent account, see Beaver & Krahmer, , and Beaver & Geurts for an overview). This captures projection phenomena, as can be illustrated for (a) and (b): the latter will always receives a classical truth-value, because (assuming a material implication analysis) if the antecedent is false, the entire conditional will necessarily be true, regardless of whether the presupposition is met or not. If the antecedent is true, on the other hand, then the presupposition is also true, since the antecedent entails it. Conversely, for (a), there is no such relationship between the antecedent and the consequent, and thus it’s possible for the entire sentence to receive the value ‘#’ if the presupposition of the consequent is not met. The main novel feature of these recently developed accounts is that they offer a modular perspective that can separate out the semantic machinery underlying projection from pragmatic principles crucially driven by the left-to-right properties of actual comprehension. In particular, these can be thought of as a semantic component that is independent of linear order on the one hand, and a processing-driven preference for calculating presuppositions on the fly with order-based, incremental effects on the other. This allows for a principled explanation of the asymmetries stipulated by dynamic semantics, while leaving open the possibility that symmetric interpretations are available as a dispreferred
4
i.
There are other potential alternatives, e.g. c + p AND q = c + (p + q)
which would also make different predictions in terms of whether the context c and the first conjunct can jointly conspire to ensure the presupposition of the second conjunct.
, ,
alternative. It also opens up interesting questions about presupposition processing, as will be discussed in section .. The final recent approach to mention here is the QUD-based approach represented by Simons et al. (, and related work). As noted in section ..., it relates presuppositionality directly to at-issueness, which in turn is explicated in terms of relevance to the QUD. Projection is also explained in these very terms, by positing that it is precisely the content that is not at-issue that projects. Key features of this account are a) that the projective nature of a given piece of content can interact directly with the discourse context; b) that there is no need for semantic machinery that deals with projection; and c) that the triggering problem and the projection problem are grounded in the same property, not-at-issueness. This perspective on presuppositions also opens up interesting questions for experimental investigation, though first steps towards their exploration have only been taken recently. In addition to projection from sentential operators, intricate issues also arise when presuppositions appear in the scope of quantificational expressions, as in (): ()
Most semanticists headed to the bar again after the talk was over.
Descriptively speaking, we can differentiate three types of theories: a) universal theories, which predict that regardless of quantifier, the presupposition has to hold of all the individuals in the domain quantified over, that is, that all semanticists are presupposed to have been to the bar before (Heim, ; Schlenker, b, a; Chemla & Schlenker, ; Mayr & Sauerland, ; also see George, b, and Fox, a, for slightly weakened variants); b) existential theories, which uniformly predict that this presupposition merely needs to hold of some individuals in the domain (Van der Sandt ; Beaver ; Geurts ); c) mixed theories of various sorts, for example, ones that assume that either reading is generally available or that the quantificational force at play in the presupposition varies across different quantifiers (Chierchia, , ; George, b; Chemla, a; Romoli, , ; Sudo, ; Mandelkern, ; Mayr & Sauerland, ). While we cannot go into the details of how these predictions are derived formally on the various accounts for reasons of space, the empirical question of which readings are actually available and how they are derived constitutes another important area for experimental work, to which we turn in section ....
.. Distinguishing types of presupposition triggers An additional theoretical development, which at least in part has gone hand-in-hand with attempts to explain the triggering problem, concerns proposals to distinguish different types of presupposition triggers. In particular, various of the proposals for analysing ‘soft triggers’ in terms of conversational reasoning alluded to earlier (e.g. Simons, ; Abusch, , ; Romoli, ) draw a distinction in terms of the source of the presuppositions of various triggers, as their pragmatic account only applies to a sub-class of triggers, namely the ‘soft’ ones. The ‘hard’ triggers, such as too and again, are usually assumed to involve lexically encoded, conventional presuppositions. But various other proposals for potential distinctions between triggers exist in the literature as well, which focus more on how different triggers relate to the context. Kripke
(, originally published in ) noted early on that some triggers are much more susceptible to accommodation than others, which he tries to capture in terms of whether or not they are anaphoric, whether or not they require an antecedent in the discourse context (in a manner quite parallel to pronouns; see Rohde, Chapter in this volume). Similarly, Zeevat () distinguished ‘lexical’ from ‘resolution’ triggers, where the former involve a presupposition that’s directly related to the main asserted content (e.g. know) and the latter are anaphoric (e.g. too). Glanzberg () argues for yet another related distinction between ‘weak’ and ‘strong’ presupposition triggers, based on observations about whether they are associated with obligatory or optional accommodation, the main idea being that in the latter case, context update with the asserted content is possible even when the presupposition is not met, for example, in the case of sentences containing the triggers again or too, whose presupposition is highly independent of the main asserted content. In another line of work, Sudo (; also see Sudo & Spathas, ) and Klinedinst () take another approach by differentiating triggers in terms of how the presupposition relates to the entailed content: the presupposed content of some triggers, they suggest, such as stop, has a dual presence in the lexical representation, so that they are part of both the presupposed and the entailed content (for a related idea within an analysis of NPIs in definite descriptions, see Gajewski & Hsieh, ; Gajewski, ). Finally, Tonhauser et al. () provide a classification of ‘projective content’ that is both broader in scope and more fine-grained than a traditional view. In particular, among expressions that are classically considered to be presupposition triggers, they differentiate based on whether a given trigger has ‘obligatory local effects’ (i.e. contributes to the compositional interpretation of its clause, even when embedded; this is parallel to Zeevat’s notion of ‘lexical’ triggers as well as to Sudo’s () and Klinedinst’s () notion of presuppositions being entailed). Furthermore, they divide triggers based on whether they come with a strong contextual requirement, which roughly corresponds to whether or not (or how easily) their presupposition can be accommodated. Combining these two distinctions, they propose a system of four types of projecting content (where content with neither a local effect nor a strong contextual effect roughly corresponds to the notion of conventional implicature in the sense of Potts, ). As will be seen in section ., these discussions about differentiations between triggers have played an important role in much of the experimental work, as the various proposals inherently come with empirical criteria for distinguishing types of triggers, which in turn can straightforwardly be investigated more systematically with experimental tools.
. P
.................................................................................................................................. In this section, we turn to experimental work on presuppositions and its relation to the theoretical issues discussed in section .. We begin by considering issues arising for presuppositions in unembedded environments, and turn to questions specific to embedded environments in the following section. From an experimentalist’s perspective, there are a number of questions to ask about the role of presupposed content in language
, ,
comprehension,5 each of which in turn can be addressed using a variety of methodological approaches. The following discussion is organized by these questions, in particular the following.6 Is a given putative presupposition indeed at play in experimental settings? (section ..) What reflexes does the arguably special status of presuppositions have in behavioural measures? (section ..) What happens when a trigger is used in a context where its presupposition is not already established? (section ..) What is the time-course of presupposed content becoming active in online comprehension (both in absolute terms and relative to other aspects of meaning)? (section ..) In the course of discussing these questions, the core methodological approaches used for studying presuppositions to date will also be introduced. Furthermore, the crucial theoretical issue of whether there are different sub-classes of triggers will pop up throughout (a discussion of this issue in its own right will be postponed until section .., as the most crucial contrasts relate to how different triggers behave in embedded environments).
.. Detecting presuppositions experimentally The perhaps most fundamental empirical question about a given expression that is assumed to be associated with a presupposition is whether the presupposition is indeed present. This question has at least two parts: first, is the content in question associated with the expression? And, second, does it have the properties we expect presuppositions to have? We begin by considering two studies exemplifying methodological attempts to address the first question. Much of what follows in later sections will naturally bear on this as well, but also inform the second question. In addition to empirically confirming theoretical assumptions about the impact of triggers on meaning, basic methodological questions are also at stake: given that the prevalent theoretical views see an inherent relation of presuppositions to the discourse context, it cannot simply be assumed that presuppositions will arise in their usual way in abstract experimental settings, which often involve the presentation of isolated sentences by an unknown speaker (or writer). Thus, it is methodologically crucial to ensure that presuppositions play their usual role in such settings. One approach to testing for the presence of a given piece of content in a sentence is to provide multiple paraphrases and ask participants to select the one that best corresponds to their interpretation of the sentence. Schwarz () used precisely such a task to investigate syntactically ambiguous German sentences, whose structural ambiguity interacted directly with the presupposition of the trigger auch (‘also’). The ambiguity was created based on syncretism for N(ominative) and A(ccusative) case marking in German, as in the following: () Die Frau, die das Mädchen sah, hatte auch der Mann The woman/ who/ the girl/ saw had also the man gesehen. seen ‘The woman that (saw the girl/ the girl saw) had also been seen by the man.’ 5
In principle, there are equally interesting and important questions about production, but to my knowledge, these have not been approached experimentally. 6 For a presentation covering mostly the same ground but organized in terms of methodological approaches, see [Schwarz (a)].
Subjects had to select one among various paraphrases, which differed in whether the woman was said to have seen the girl or vice versa. Based on syntactic processing preferences, the former is independently preferred. However, the other interpretation ensures that the presupposition introduced by also (that someone else had seen the woman) is satisfied. The results indicate that subjects indeed take this into consideration, as the paraphrase in line with the presupposition was chosen more frequently than in a presupposition-less control condition. This provides evidence that even in abstract experimental settings with sentences out of context, interpretation choices seem to be affected by a desire for interpretations that fully incorporate their presupposition, which may override independent processing factors biased in the opposite direction. Other variants of tasks aiming to directly identify whether a presupposition is part of the interpretation of a sentence target inferences in other ways, either by asking questions pertaining to the presupposed content (e.g. Domaneschi et al., ; Tiemann, ; Tiemann et al., ) or by directly asking participants whether they would draw the relevant inference when presented with the sentence (e.g. Chemla & Schlenker, ). These will be discussed in sections .. and ..., respectively, as they bear on additional, more complex issues. In addition to the paraphrase selection study, Schwarz () also presented reading time evidence supporting the notion that the presupposition is active in comprehension within the experimental setting. This consisted of self-paced reading results for both German and English on disambiguated versions of the sentences, where the presupposition is either met or not supported within the presented sentence. Reading times increased significantly on the region containing also in the latter case. Along the same lines, Tiemann et al. () report self-paced reading results for various triggers as well, with parallel slowdowns either on the region containing the trigger or the one following it (see Tiemann, , for additional results on again). Another reading time study, by Clifton (), further enriches the methodological perspective and reinforces the point that it should not be taken for granted that presupposition-based effects always arise in experimental setups. He looks at definite descriptions, focusing on the effect of the uniqueness presupposition. A context sentence is used to establish whether there is one or multiple of the relevant items (e.g. In the kitchen . . . vs. In the appliance store . . . ), and the following target sentence contained either a definite or an indefinite description ({The / A} stove). Despite clear intuitions about the variation in the felicity of the materials, Clifton found no slowdown effect in an initial self-paced reading study, in contrast to the previously mentioned studies. However, a follow-up with memory load in form of a simple arithmetic task added in between reading the sentence and answering a comprehension question found a clear effect in the region following the definite, with longer reading times in the multiple-item ( non-uniqueness) condition. This highlights an important methodological aspect, namely that subjects may not fully engage in linguistic processing in an experimental setting when the task at hand does not require it. A different, and in some ways more straightforward, approach to testing for the presence of presuppositions in the interpretation of a sentence is to use tasks where the presence or absence of the potential presuppositional inference in question will be revealed indirectly
, ,
through subjects’ behaviour in a non-linguistic task. One methodologically quite ingenious implementation along these lines can be found in the acquisition study by Dudley et al. (), which investigates the factive presupposition of attitude predicates such as know. A crucial question for them is to what extent children are sensitive to this aspect of meaning, specifically in contrast to non-factive verbs such as think. Dudley et al. () address this question through a guessing game, where an experimenter hid a toy in one of several boxes. Before the child guessed which box contained the toy, a puppet would whisper in the experimenter’s ear. The experimenter would then relay what the puppet said by saying Lambchop (doesn’t) {know / think} that it’s in the red box. If children are aware of the factivity of know, their response behaviour for the two verbs should differ. Dudley et al.’s () results show that at least some of the -year-olds in their study have an exquisite understanding of the factive component of know (on par with expected adult behaviour). This extends not only to the truth of the complement of know, but also to its core presuppositional property of projecting out of the scope of negation. Studies such as the ones exemplified here constitute a methodological proof of concept that presuppositions can be fruitfully investigated experimentally, and thus open up a wide range of possibilities for investigating further aspects of presuppositions. The next section turns to the status of presuppositions in comparison to asserted content.
.. Reflexes of the status of presuppositions A crucial aspect of distinctions drawn between presupposed content and other aspects of meaning, chiefly the conventionally encoded, truth-conditional content that forms the core of what’s asserted, is that they differ in status. This can affect a variety of behavioural measures, from truth-value judgements to response times and available continuations or responses to presuppositional sentences. Given the historical role of definites in the study of presuppositions, let us begin with the standard claim that non-referring definites lead to a sense of ‘squeamishness’ (Strawson, ), that is, reluctance to judge them to be either true or false. Abrusán & Szendrői () utilized a truth-value judgement task in which they provided a third option labelled as ‘can’t say’. Their materials included a number of variations in terms of the role of the definite in the sentence, based on notions such as topicality and verifiability, which have been argued in the literature to affect the strength and/or presence of squeamishness (Reinhart, ; Lasersohn, ; Von Fintel, ), as well as negated versions. Somewhat surprisingly, affirmative sentences were found to be judged as ‘false’ quite consistently by subjects, with little use of the ‘can’t say’ option, suggesting that either there is no relevant difference in status attached to the existential implication, or that this task is not sensitive to it. However, the negated versions of their sentences displayed significant variation in the distribution of judgements, which suggests that the existence condition of definite descriptions does indeed have a status distinct from basic entailments, whose impact on judgements can furthermore be modulated by a variety of pragmatic effects. More recently, Zehr () reports a similar study presenting sentences with the presupposition trigger stop paired with visual contexts. In this study, the third choice is introduced as ‘neither’, which subjects choose about per cent of the time, suggesting that squeamishness can indeed be captured for affirmative sentences in truthvalue judgement tasks with additional response options.
Since truth-value judgement tasks do not always provide straightforward evidence for such a distinction, other studies have taken the approach of looking at the time-course of truth-value judgements to assess whether false judgements based on false asserted vs. false presupposed content might be differentiated in terms of their time-course. First, Kim () investigated the presupposition of only. A sentence like Only the girls have books commonly is taken to presuppose that the girls have books. Kim’s experiments present such sentences in visual contexts that either did or did not conform to this presupposition (i.e. showed the girls as having books or not). The truth of the asserted content (whether or not non-girls had books) was also varied across conditions. Subjects took longer in their responses when they were based on an unmet presupposition than when they were based on false asserted content. Kim interprets this result as a reflex of the backgrounded nature of presuppositions, which impacts the verification procedure employed in the task: presuppositions are literally taken for granted, and not initially verified, in contrast to asserted content, which leads to longer reaction times in falsifying them. Schwarz (b) takes a similar approach to the study of the existence implication of definite descriptions, again by asking subjects to provide truth-value judgements on sentences relative to visual contexts which either falsify the asserted or the presupposed content. Indefinites serve as a control, where essentially the same information is asserted in both conditions. The results yield a significant interaction between type of determiner and the contextual information affecting which part of the sentence is falsified. This suggests that the status of the existence implication is indeed distinct from that of the main asserted content, and thus supports presuppositional analyses of definites in the tradition of Frege () and Strawson (), in contrast to accounts in the tradition of Russell (), which see it as a mere entailment. Another set of tasks that has proven useful in investigating presuppositions involves comparisons between different versions of continuations of a presuppositional sentence. Onea & Beaver () and Destruel et al. () used this to investigate the exhaustive inference of focus and clefts (Onea, Chapter in this volume; see also Velleman et al., , for other triggers). Example (), from Destruel et al. (), illustrates a forced choice version of the task, where subjects had to indicate which of several continuations, including the ones below, best matched the context: ()
It was a necklace that Phillip bought his sister. a. Yes, but Phillip also bought his sister a bracelet. b. No, Phillip also bought his sister a bracelet.
For clefts, as compared to exclusive statements with only, subjects more frequently selected continuations like the one in (a). The authors explain the difference between clefts and exclusives in terms of the status of the exhaustive inference, which is at-issue in the latter but not in the former, and thus is more or less likely to be targeted by Yes and No. Other studies use an acceptability rating version of this approach. Cummins et al. () and Amaral & Cummins () investigate various triggers in English and Spanish and test the acceptability of Yes, although . . . and No, because continuations: ()
Q: Did Brian lose his wallet again? A: Yes, although he never lost it before. A’: No, because he never lost it before.
, ,
Across all triggers, both responses are degraded relative to controls, suggesting that contradicting the presupposition comes with a cost. But interestingly, the triggers in their results seem to be grouped into two classes, which the authors relate to the distinction between lexical and resolution triggers (Zeevat, ). Recall that the latter are anaphoric and directly relate back to entities (or events) in the context (e.g. again and too), while the former involve cases where the presupposition is a requirement that comes with the asserted component of the trigger (e.g. regret, stop, still, continue). For lexical triggers, the authors find systematically higher acceptability ratings for continuations parallel to () A’, whereas there is no difference in acceptability between the continuations for the resolution triggers. This is in line with Zeevat’s distinction, as for these (and only for these), ‘the responses in condition [A] appear self-contradictory, if we assume that the presupposition is a logical prerequisite for the at-issue content of the trigger’ (Amaral & Cummins, : ).
.. ‘Novel’ presuppositions: To ignore, accommodate, or cancel? As noted in section .., presupposition triggers can be used in many circumstances where their presupposition is not already established, despite the common view that they impose felicity requirements on the context. Such cases are usually assumed to involve accommodation, a repair process in which hearers quietly accept the presupposition, leading them to adjust the Common Ground to entail the presupposition in order to be able to interpret the sentence relative to this adjusted context (Lewis, ; Von Fintel, ). But other characterizations of such cases are in principle available. It might be that presuppositions are sometimes ignored altogether, or cast aside once it’s clear the context doesn’t support them (see, in particular, the proposal by Glanzberg, , to which we return later in this section). This section reviews experimental work contributing to our understanding of what happens when hearers encounter such ‘novel’ presuppositions. In light of the view of accommodation as a repair mechanism, a fair share of work in this area has tried to determine whether there is any measurable cognitive cost involved. This is particularly pressing as accommodation is indeed quite common, at least for a certain range of triggers (see Spenader, , for extensive corpus data illustrating this). Methods of assessing the extent to which accommodation is available, and whether it is costly, include reading time measures and acceptability ratings in contexts that do not directly support the presupposition. Some early psycholinguistic studies investigated related issues based on definite descriptions, especially cases of bridging (Clark, ). For example, Haviland & Clark () compared contexts that required a bridging inference (e.g. mentioning picnic supplies) with ones where some entity (e.g. beer) was mentioned explicitly, and found longer reading times on a subsequent presentation of a definite (e.g. the beer). O’Brien et al. () showed, however, that prior mention of a referent is not necessary if the context is sufficiently specific: the definite the knife was read more slowly when the antecedent was more general (a weapon) than when it directly matched the noun phrase (a knife), but only when the context involved a general verb such as assault, and not when it involved stab, which is more closely associated with knives. In an acceptability rating study, Carlson & Tanenhaus () find that a sentence like
The suitcases were heavy is judged to make sense more frequently following the sentence Bill hurried to catch his plane than the sentence Bill hurried to unload his car, presumably because the presence of suitcases is more salient in the former case. A more recent set of studies illustrates how fleeting accommodation-based effects for definites can be. Frazier () looked at the impact of plausibility on accommodation, for example by varying the noun phrase occurring in a definite, using a passive context sentence (My order was taken) followed by a plausible or implausible definite or indefinite (a/the waiter/busboy), and only found effects of plausibility, not of definiteness, in a reading time study. Another line of work has used neurolinguistic measures. For example Burkhardt (; among others; also see Van Berkum, Brown, Hagoort, & Zwitserlood, , and subsequent work) used ERP studies to identify neural correlates of bridging by looking at definites such as the conductor in contexts with an explicit antecedent, a bridging antecedent (a concert), and no antecedent. Intriguingly, the bridged cases display neural hallmarks of both new and old information in the form of a reduced N effect, followed by a P effect. In the more recent literature, one rather comprehensive set of studies involving acceptability measures comes from Tiemann et al. (). These authors look at contexts with varying degrees of support for a variety of German presupposition triggers, including possessives, factives, iteratives, and aspectual verbs. For all these triggers, the use of presupposition triggers in contexts that do not explicitly support the presupposition is rated as less acceptable than both non-presuppositional controls and variations with contexts that support the presupposition. At the same time, they are rated consistently as more acceptable than variants where the context is directly inconsistent with the presupposition (also see Schwarz & Tiemann, ). This suggests that while there is some cost associated with accommodation, the associated decrease in acceptability is only a moderate one. While Tiemann et al. () find the same overall pattern for the various triggers in their study, the strength of the accommodation effect appears to be somewhat varied. This comes as no surprise, as it is commonly claimed that triggers differ substantially in their ease of accommodation, although it is by no means clear how to account for this theoretically (Beaver & Zeevat, ). As noted in section .., Kripke () claimed that triggers like too resist accommodation altogether. However, Tiemann et al.’s () intermediate acceptability findings for auch, ‘too’ (as well as for wieder, ‘again’) indicate that accommodation is better than baseline controls with completely implausible adverbs or contexts that directly conflict with the presupposition. This is in line with observations by Von Fintel () and Chemla & Schlenker () that accommodating too is possible when in line with plausibility in context. Singh et al. () directly addressed this issue experimentally, using a stops-making-sense task, where subjects see a sentence unfold word by word as they press one button and are instructed to abort the trial with another button if it no longer makes sense to them. They compare the triggers the and too to presupposition-less controls in both plausible and implausible contexts, as in the following sentences: ()
Context: Bill went to {a club / the circus} on Friday night. Target: {A / the} bouncer argued with him there for a while.
()
Context: John will go to {the pool / the mall} this morning. Target: Peter will go swimming {tomorrow / too} after he gets back from school.
, ,
Rather strikingly, they find that in plausible contexts ( . . . a club and . . . the pool, respectively), the presence of the presupposition trigger has no impact on the stops-making-sense task, and subjects overwhelmingly accept the sentences for both triggers. In contrast, the presence of the trigger has a strong effect in implausible contexts, suggesting that accommodation is not viable. But in the plausible context, accommodation seems to be just as readily available for the and too. Reading time effects suggest that it may nonetheless be slightly harder in the case of too, but it is clearly not ruled out in plausible contexts. In sum, acceptability rating tasks help to shed light on the relation of presupposition triggers to context. The results to date support the traditional notion that presuppositions in general impose contraints on felicitous contexts of utterance, but also suggest some need for distinctions between different types of triggers. While acceptability rating and reading time measures help to assess the costs involved in encountering triggers in contexts that do not support their presupposition, they do not directly speak to the question of whether the presupposed content indeed does wind up getting added to the context. Several recent studies have addressed this point head-on using inference-based tasks, and argue their results to show that at least for certain triggers, the presupposition actually gets ignored, based on the notion that accommodation as a costly process is avoided whenever possible. Domaneschi et al. () use comprehension questions for texts containing triggers to assess the presence of presupposition-based inferences. They auditorily presented short stories that contained a variety of presupposition triggers. The key measure came from a True/False comprehension question, which related directly to the presuppositions, none of which were explicitly supported in the story. (They also included a variation in cognitive load to assess processing efforts involved). The theoretical motivation was based on the distinction between triggers introduced by Glanzberg (), where triggers are assumed to differ in whether the presupposed information is obligatorily processed in non-supporting contexts (as with strong triggers such as factive verbs) or not (as with weak triggers, such as iteratives like again). Their results suggest that these two types of triggers indeed differ in terms of how present the information introduced by them is when answering questions about the previously heard text, in that accuracy is overall much lower for weak triggers. While these findings are in line with the distinction between strong and weak triggers, alternative explanations should also be explored, for example in terms of the level of backgroundedness or immediate relevance for the current topic of discussion when the trigger is encountered. Another related inference-based task is incorporated into the study of again by Tiemann () (also see Tiemann et al., ), with results that seem to mirror those by Domaneschi et al. () for the iterative prefix re-. Sentences such as Linda received a pink lamp again are presented in contexts where Linda either had received a pink lamp previously or not. For one-third of the items, a comprehension question assessed the extent to which subjects accommodate Linda receiving a pink lamp on a previous occasion when the immediate context did not support this. In particular, subjects had to answer whether Linda had received one or at least two pink lamps in total. Somewhat surprisingly, the presupposition of again hardly affected subjects’ answer choices at all in the non-supporting context: that is, they overwhelmingly chose ‘one’ as the answer. The authors interpret this as suggesting that accommodation is a last-resort mechanism that is to be avoided if at all possible. Alternatively, one might explain subjects’ behaviour in terms of narrowly interpreting the question with regards to the immediate context, but
even so, it is very interesting that the presupposition does not seem to counter such a restrictive interpretation at all. To investigate the proposals from the previous two studies directly, Bacovcin et al. () use a picture matching task with variations in whether the presupposed content is explicitly supported, left uncertain (through the use of ‘?’ in parts of the picture), or explicitly not met. Comparing the triggers continue and again, they find clear differences in confidence ratings and reaction times between conditions allowing for accommodation and ones explicitly inconsistent with the presupposition. Furthermore, hardly any differences arise between these triggers, and no cost of accommodation is detected. The authors argue these results to speak against accounts based on the idea that presuppositions of triggers like again can be ignored entirely, or alternatively that accommodation is avoided for them whenever possible. Their results provide clear evidence that the presuppositions of both again and continue are considered even in an experimental setting where they do not play any crucial role for the task at hand. They thus suggest that the presupposed content is fully considered for all triggers across contexts, and furthermore that accommodation even takes place when not strictly necessary.
.. The time-course of presupposition interpretation While some of the response time and self-paced reading studies discussed in the preceding three subsections help to shed some first light on the time-course of presupposition interpretation in online processing, they only do so at a fairly coarse-grained level. A more fine-grained temporal perspective not only serves to increase the general understanding of the cognitive processes involved in interpreting presuppositions, but also helps to assess theoretical comparisons between presuppositions and implicatures. Various authors (including Bott & Noveck, ; Huang & Snedeker, ) have argued that implicatures are delayed in online processing, and if certain presupposition triggers in fact are a type of implicature (among others Abusch, ; Romoli, ), we may expect similar effects here.7 In recent years, more fine-grained methods for investigating online processing have been used to study presupposition as well. First, some of the self-paced reading studies in section .., mainly have been extended to eye-tracking during reading. Most relevantly, Schwarz & Tiemann (; also see Schwarz & Tiemann, ) investigate German sentences with again in contexts that either are or are not consistent with its presupposition. They find slowdowns in the earliest fixation measures, including first fixation duration, in reading times on the verb that immediately follows again. These effects provide temporally finegrained evidence that presuppositions are integrated with the discourse context more or less immediately (at least in unembedded contexts; see section .. for embedded cases). Along the same lines, Clifton () also reports eye-tracking effects parallel to the selfpaced reading data discussed in section .. in first-pass time measures. In addition to these reading studies, several recent eye-tracking studies have used the visual world paradigm (Tanenhaus et al., ) to investigate presupposition processing. 7
But note that there is an ongoing debate on whether implicatures are indeed delayed (Grodner et al., ; Breheny et al., ; Degen & Tanenhaus, ).
, ,
These involve visual stimuli with a number of alternative candidates for reference, paired with auditory linguistic stimuli. Participants’ eye movements are monitored as the linguistic input unfolds, and the general design is set up so that looking preferences can be interpreted as indicating the availability of the interpretation of interest at a given point in time. Chambers & San Juan (, ) investigate another and return with this method, and find rapid shifts of fixations based on the respective presuppositions. More recently, Romoli et al. () look at English also in comparison with the asserted part of only. They find shifts in eye movements based on the presupposition of also as early as ms after its onset, indicating that the presupposition is utilized in determining the referent before further disambiguating information is introduced. Schwarz (b) contrasts the same two expressions, and observes a shift in fixations as early as ms after the onset of also, suggesting that the presupposition introduced by also is immediately available and utilized in identifying the referent. A second experiment looks at the interpretation of stressed also, which associated with the subject of the sentence, again in comparison to the asserted exclusivity of only. While also again gave rise to an essentially immediate shift in fixations towards the target (starting at ms after the onset of also), the exclusive inference introduced by only did not give rise to a parallel shift until ms after its onset. Extending this approach, Schwarz () compares a hard and a soft trigger, again and stop, to assess whether the potentially pragmatic nature of the latter might lead to differences in processing speed, given related findings for implicatures (Huang & Snedeker, ). However, both triggers very much parallel the time-course observed in the other studies on also, with immediate shifts in fixations based on the presupposed information. The results from these studies, together with the reading results taken in conjunction with other results on reading, thus do not provide any support for the notion that at least some presuppositions are pragmatically computed in a costly manner associated with processing delays. This may be most naturally compatible with accounts that assume all presupposed content to be encoded conventionally. But it is also possible that we are looking at rapid pragmatic effects, so the results do not per se settle the question about the source of presupposed content.8 (Parallel issues arise in the experimental study of scalar implicatures; see Breheny, Chapter in this volume.) Nonetheless, they provide the most direct and time-sensitive evidence yet that presupposed information is available and utilized as soon as the presupposition trigger is introduced.
.. Interim summary Experimental work on presuppositions in unembedded contexts has established basic methodological approaches for detecting presuppositions, although it is also clear that their impact can be fleeting when subjects do not fully engage with the linguistic stimuli. The backgrounded status of presuppositions is reflected in the fact that they do not seem to receive much attention in initial phases of sentence verification, given the response-time 8 Another possibility is that presuppositions in affirmative contexts are also part of what is entailed (see section ... for discussion), which could be another explanation of the fast availability of presupposed content.
delays for judgements based on false presupposed information. While no knock-down arguments for or against any theoretical perspective can be made based on the time-course of presupposition interpretation, it is important to note the apparent contrast between presuppositions and implicatures in this regard, given theoretical approaches that essentially reduce certain presuppositions to implicatures. When presuppositions are not supported in the context, this generally invokes accommodation. While triggers do seem to differ in terms of ease of accommodation, the common claim in the theoretical literature that some triggers entirely resist accommodation does not seem to be borne out. Future work is needed to gain a better understanding of what factors facilitate accommodation and to account for them in theoretical terms.
. T
.................................................................................................................................. Since projection is generally taken to be a hallmark feature of presuppositions, it is only natural for experimental work to address projection-related phenomena as well. A first challenge is to identify precisely what the nature of the projected content is. Theories differ in their predictions in this regard, both with respect to embedding under quantifiers and sentential operators, and the empirical situation is in many ways too subtle to settle the facts without more comprehensive experimental approaches. As noted in section ..., projection does not always take place, and the availability and status of the resulting local interpretations has been subjected to experimental study as well. Finally, much of the discussion about potential differences between triggers is based on claims about differences in projection behaviour, and various experimental attempts have been made to assess these claims.
.. Identifying what projects ... Projection from the scope of quantifiers As noted in section ..., an important question to settle is whether presuppositions in the scope of quantificational expressions project existentially or universally (or whether this varies by trigger). Chemla (b) provides a first experimental attempt to address this by looking at sentences such as No student knows he’s lucky. He presents these sentences, and variations with a range of other quantifiers, to subjects and asks them to judge whether (or, in a second experiment, how strongly) the sentence suggests that all of the students are lucky. Judgements for the quantifier no pattern together with those for every and indicate a universal presupposition, whereas universal inferences for numerical quantifiers (e.g. more/ less than ) are less strongly supported. Chemla proposes to capture the results in terms of Similarity Theory (Chemla, a), which allows the quantificational force at play in the presupposition to vary across quantifiers. Using eye-tracking during reading, Tiemann () reports a German study that manipulates context sentences for quantificational target sentences precisely with respect to whether the relevant presupposition is met universally or not. Her results find slowdowns
, ,
in reading time for jede (‘every’) in non-universal contexts, but not for ein (‘one’). Parallel to Chemla’s findings, this suggests that the nature of the projected presupposition depends on the quantifier. Geurts & Van Tiel () investigate the effects of presuppositions on domain restriction. Pairing simple geometrical figures with sentences in a truth-value judgement task, they look at quantified sentences such as Each of these circles has the same color as the square that it is connected to. Rather strikingly, they find that even a picture where only two out of the seven circles presented are connected to the square next to them (and have the same colour) yields a substantial amount of ‘true’ judgements—up to per cent of the time based on the visual display. The authors analyse this in DRT and propose that such judgements are based on intermediate accommodation, but the effect could also be attributed to a form of domain restriction. Another finding, which seems to be in direct contrast with Chemla (b), is that acceptance of sentences with none are at ceiling level throughout, suggesting an existential, rather than a universal presupposition. Most recently, Zehr et al. () argue based on a picture matching task using the trigger win that, descriptively speaking, none can exhibit both existential and universal readings (in addition to a local accommodation reading). They then lay out various options that different theories have for deriving alternative (weakened or strengthened) readings through other mechanisms. In a follow-up acquisition study, Zehr, Bill, et al. () find that child data exhibits a contrast between conditions revealing a universal projection reading, but no evidence for existential readings is found. They argue the emerging picture to favour a view where projection for none is universal, with additional mechanisms like domain restriction responsible for deriving a weakened, existential reading (which may be developmentally later, given that it is more complex).
... Projection from sentential operators Turning to embedding under sentential connectives, there is substantial disagreement in the theoretical literature on whether a presupposition trigger in the consequent of a conditional (such as the possessive definite in ) gives rise to a conditional presupposition (a, e.g. on classical dynamic semantic accounts) or a non-conditional one (b, e.g. on DRT accounts): ()
If Al goes surfing, he’ll wear his wet-suit. a. If he goes surfing, he has a wet-suit. b. He has a wet-suit.
Both interpretations seem to be attested, but theories differ in terms of which one they see as basic. Romoli et al. () provide a first experimental exploration of this topic using a covered box picture matching task (Huang et al., ), and argue their results to favour accounts that predict a conditional presupposition as the basic one. Their results also support the notion that whether or not the presupposition intuitively can be seen as dependent on the content of the antecedent affects judgements. Another line of experimental work on projection investigates the role of incrementality, based on the idea from Philippe Schlenker’s (a,b, ) work that presupposition projection can be broken down into a symmetric semantic component and a processing-
based incremental component. Crucially, order-based incremental effects (as in b) are attributed to left-to-right processing, but are in principle violable. This opens up interesting questions about presupposition processing. Chemla & Schlenker () home in on this issue and test presupposition triggers in conditionals, disjunctions, and unless-sentences in configurations where the presupposition trigger appears either in the linearly first or second clause. In an inference judgement task, they find that subjects endorse conditional inferences more strongly than non-conditional ones, regardless of where the presupposition trigger is introduced. They interpret this as support for a symmetric theory of presupposition satisfaction, where material introduced later on in the sentence in principle can provide support for an earlier presupposition. Schwarz (c) varies this paradigm by looking at conditionals in a covered box picture selection task and varying the position of the if-clause. The results here are more mixed, in that if-clause initial conditions suggest a fairly strong role of incrementality, while the if-clause final conditions are more in line with symmetric predictions. Hirsch & Hackl () investigate the effects of incrementality in disjunctions. These pose a potential challenge to a general processing-based effect of linear order, as they seem entirely symmetric, for example, in the following famous example due to Barbara Partee: ()
Either the bathroom is in a funny place, or there is no bathroom.
Unlike in conjunctions, a trigger in the first disjunct does not generally project globally, that is, () appears to be equivalent to a variant with the disjuncts reversed. However, Hirsch & Hackl () argue that we are actually dealing with a case of local accommodation (discussed in more detail in the next section), as a global presupposition interpretation would be inconsistent with the non-presuppositional disjunct, and each disjunct is independently required to be a live possibility in the global context. Rather than predicting an overall asymmetry in projection parallel to conjunction, an incremental account of projection then will merely predict a processing effect due to a garden path effect, as the global reading is considered first and then abandoned. The authors assess this by having subjects select the more natural of two sentences, one parallel to (), the other with an additional presupposition trigger that is consistent with a global interpretation of the other trigger. The results are in line with the authors’ predictions in that a stronger preference emerges for the version consistent with a global presupposition when the trigger appears in the second conjunct. In contrast, control sentences that at no point suggest a global presupposition interpretation display no effect of order. Initial online evidence in line with this analysis based on a global, non-conditional presupposition is provided by Hirsch et al. (), who use visual world eye-tracking and argue that global interpretations are temporarily considered even when they cannot be ultimately maintained.
.. Local readings While it’s commonly known that there are exceptions to projection, it’s important to note that just because a presupposition does not project doesn’t mean that it disappears entirely. Indeed, many accounts assume that what happens in (at least some of) such cases is that it gets locally accommodated (as first proposed within context change semantics by Heim, , though parallel effects can be derived in other frameworks, e.g. through Beaver &
, ,
Krahmer ()’s A-operator in trivalent ones). This offers an explanation for the fact that sentences such as The king of France is not bald—because there is no king of France!, (already discussed by Russell, ), are quite acceptable, despite the incompatibility that would be expected based on a global existence presupposition of the. The consensus in the theoretical literature has been that local interpretations are dispreferred (beginning with Heim, ), though this is only based on authors’ intuitions. More recently, Chemla & Bott () offered the first experimental evidence to support that assessment, using a truthvalue judgement task with sentences such as () and looking at reaction time measures. ()
Zoologists don’t realize that elephants are reptiles.
The factive verb realize presupposes the truth of its complement clause, and on its global interpretation, this presupposition prevails even in the context of negation (see Tian & Breheny, Chapter in this volume, for a more general discussion of negation). However, a local interpretation would have that inference negated. In the latter case, the sentence should be judged true, whereas on the former, it should be judged false. Both types of responses are given by subjects throughout the experiment, but the ‘true’ responses take significantly longer than ‘false’ responses. Chemla & Bott () interpret this as evidence for traditional, semantic accounts that take local accommodation to be a last resort repair strategy. In contrast, the results are argued to be incompatible with pragmatic accounts, such as Schlenker (a), which assume that the local reading corresponds to a literal semantic reading, while the global reading requires additional pragmatic inferencing. Romoli & Schwarz () utilize a different task to investigate the speed of local interpretations of the presupposition introduced by stop under negation, namely a ‘Covered Box’ version of a picture selection task (Huang et al., ). Subjects have to select a match for a given sentence among various pictures, one of which is ‘hidden’. The basic idea is that if the presuppositional inference of interest plays a role in subjects’ interpretation, then they should choose the covered box in cases where no overtly shown image is compatible with the inference. Their experiment compares cases where the overt picture supports the presupposition with ones where it doesn’t. Acceptance rates were much lower for target pictures corresponding to the local interpretation. Furthermore, response times for target choices were slower for local target acceptances than for global ones, in line with Chemla & Bott (). Extending this approach to other populations, Bill et al. () and Kennedy et al. () use the same task, though without measuring response times, for testing the interpretation of presuppositions under negation in children and Broca’s aphasics respectively. Both groups turn out to be much more likely than healthy adults to adopt a global presupposition interpretation, which the authors argue to support the notion that local interpretations are derived, rather than basic, parallel to the findings in Chemla & Bott ().
.. Differences in projection between triggers ... Soft vs. hard triggers and local accommodation Much of the discussion in the literature concerned with identifying differences between (classes of) presupposition triggers is based on the observation that some triggers seem to
project more persistently than others. Abusch (, ) considers examples like the following, for example: ()
I dont know if Paul participated in the race . . . a. but, if he won, he must be very proud. b. ?? but, if Mary participated too, they probably had a drink together just after.
The presupposition of win, that Paul participated in the race, does not seem to be globally present, as it would be inconsistent with the context sentence. However, the global presupposition of too, that someone else (salient in the context) participated in the race seems to give rise to a certain amount of oddness, suggesting a global reading is required, or at least more strongly preferred. Jayez et al. () investigate this contrast experimentally, by looking at presupposition triggers in the antecedent of conditionals. Looking at French aussi (‘too’), regretter (‘regret’), and clefts, they present evidence that the distinction is not entirely robust, in that ‘hard’ triggers, too, allow local readings. Differences between triggers furthermore seem to interact with other contextual factors. The authors argue their results to be consistent with a three-way distinction between presupposition triggers, in line with Jayez (). In a related vein of ongoing work using the Covered Box task, several studies have found evidence for the availability of local accommodation for Abusch’s ‘hard’ triggers. For example, Bacovcin & Schwarz () report evidence for local interpretations of again in negated conjunctions (where the local contribution can be uniquely identified while making an alternative interpretation in terms of metalinguistic negation highly unlikely). Furthermore, Zehr, Hirsch, et al. () report on the interpretation of again in both positive (Either . . . or . . . ) and negative (Neither . . . nor . . . ) disjunctions, and find evidence for the existence of local interpretations, which furthermore can be primed by exposure to blocks of trials where such an interpretation is the only one compatible with any given picture. Other work aims for a more comprehensive overview of ‘projection strength’ across a larger number of triggers. Smith & Hall () investigate projection strength of various presupposition triggers, as well as of conventional implicatures, in a host of ‘family of sentences’ environments (Chierchia & McConnell-Ginet, ). They use a ‘surprisal’ judgement, where subjects have to assess how surprised they would be to learn that the presupposed proposition holds after hearing a sentence containing the trigger. Their findings are uniform for conventional implicatures and presuppositions (which they argue to speak in favour of a unified treatment of projection, as in Tonhauser et al., ), but also suggest that projected content has a somewhat weaker presence than nonprojected (i.e. asserted or unembedded presupposed) content. They also find some variation between triggers, though it does not line up neatly with theoretical differentiations proposed in the literature. Ongoing work by Judith Tonhauser and colleagues, reported in Tonhauser (), takes this comparative approach further and finds that projectivity is correlated with measures of at-issueness, in line with the pragmatic approach to projection spearheaded by Simons et al. (). They furthermore find that triggers that Tonhauser et al. () argue to give rise to obligatory local effects (also see discussion in the next section) are relatively less likely to project than other types of projective content that don’t (necessarily) contribute locally.
, ,
The work by Cummins et al. (; and also Amaral & Cummins ) discussed in section .. also relates to the issue of differences between triggers with regards to projection. In particular, the relative goodness of No, because . . . replies, where the presupposition is directly contradicted in the because-clause, for some triggers (such as regret, stop, still) can be analysed as reflecting those triggers’s amenability to local accommodation. While they base their discussion on the distinction between lexical vs. resolution triggers (Zeevat, ), the data largely align with Abusch’s ‘hard’ vs. ‘soft’ distinction as well (the case of regret being a notable exception). Another line of work already discussed, starting with Romoli & Schwarz (), also directly bears on the discussion of possible theoretical distinctions between triggers. While their initial findings of differences between soft triggers and implicatures called into question whether there was solid evidence for a distinction between the two, the subsequent work on different populations by Bill et al. () and Kennedy et al. () provide evidence against an analysis of soft triggers as implicatures (in particular in the version of Romoli, ), as they provide a double dissociation with distinct patterns across healthy adults, Broca’s aphasics, and children. In sum, it’s clear that descriptively speaking, there is variation in the extent to which experimental tasks yield projection-based interpretations across triggers. However, based on the evidence to date, the overall picture of such variation seems to be more gradient than might be expected on categorical distinctions between triggers. It is thus not clear that any such distinctions receive direct support from the empirical work so far, though their proponents can of course call onto other factors to account for the gradient data. The key challenge that remains for any account of projection is to explain why triggers exhibit such heterogeneous behaviour, and whether that should be captured theoretically via a difference in projection mechanisms or through independent factors.
... Local readings reconsidered: Entailment vs. local accommodation While the terms ‘soft’ vs. ‘hard’ have become fairly standard descriptive labels for distinguishing types of presupposition triggers, there are a variety of different theoretical accounts of this distinction that are separate from the line advanced by Abusch (, ) when she first introduced these. One common thread in many accounts is that triggers differ in whether or not the entailed and presupposed content directly relate to one another. For example, Zeevat’s () lexical triggers are characterized as prerequisites for the asserted content: ‘the application of a concept is only an option if certain conditions are already met’ (Zeevat, : ). Tonhauser et al. () distinguish a sub-class of triggers that give rise to ‘obligatory local effects’. Both papers cite examples involving triggers in the scope of attitude verbs as indicative of whether a trigger’s presupposition is part of its clause’s contribution to standard compositional interpretation. In a similar vein, Glanzberg () considers a trigger ‘strong’ if context update is entirely impossible unless the presupposition is already met or added to the context. All of these share the notion that with certain triggers, you don’t get to work with entailed content to the exclusion of the presupposed content. One fairly direct way of dealing with this theoretically is to assume that presupposed and entailed content do not have to be mutually exclusive. Sudo ()
fleshes out a proposal where triggers differ precisely in whether or not the presupposition is also part of the entailed content, and Klinedinst () argues that such a distinction can be used to explain the ‘soft’ vs. ‘hard’ distinction. While Klinedinst’s discussion of this approach suggests that there may be no direct way of finding independent evidence in favour of this entailment-based distinction, Sudo proposes that the scope of non-monotonic quantifiers, such as exactly one, provides a diagnostic. In effect, what such contexts will reveal is whether or not the presupposition winds up being considered when taking a count based on the predicated properties. ()
Context: Some students asked the professor lots of questions on day of the class. a. Exactly one student {continued/stopped} asking the professor lots of questions [on DAY 2]F. b. Exactly one student (also) asked the professor lots of questions (again) on day 2.
The contrast between triggers like continue and also, according to Sudo, arises when considering a scenario where another student, who had not asked many questions on day , does ask many questions on day (while only one of the day -questioners does so on day ). With continue, this does not seem to affect the truth of the exactly statement, as there only is one student who asked lots of questions both on day and day (parallel observations hold for stop with appropriate changes in the scenario). In contrast, the argument goes, with also (as well as again), a new student asking many questions on day makes the sentence false, because the entailed content evaluated in the scope of the quantifier merely is about question askers on day . While the theoretical prediction seems clear and provides a useful angle for distinguishing triggers and testing the entailment-contrast hypothesis, the intuitive picture may not be quite as straightforward. Zehr & Schwarz (to appear) present a first attempt to test the analysis of Sudo () experimentally. They use a covered box picture matching task with critical target pictures following the logic of the scenarios described earlier in this section, comparing stop and also in exactly one-sentences. They find a clear contrast between triggers, with also yielding comparable result-patterns in the critical condition and false controls, while stop yields significantly more acceptance of the visible picture, supporting the notion that there is a relevant difference between triggers. However, intriguingly, the results also reveal a difference between the critical stop items and true controls, which is unexpected if the presupposition is also part of the entailed content evaluated in the scope of exactly one. The authors argue that, in principle, this could be explained as a task effect of one sort or another, to maintain the entailment-contrast explanation of the difference between triggers. Alternatively, one could revert to a two-dimensional theory where no trigger’s presupposition is directly part of what is entailed, though it can become so through local accommodation. What remains to be explained then is why the triggers differ in the availability of this process, as this assumption would be needed to explain the contrast between triggers. While these options will need to be teased apart in further work, a blocked version of the experiment in Zehr & Schwarz (to appear) where stop items are seen before also items clearly suggests a role of local accommodation for the latter trigger, as an increased rate of critical target acceptances shows up here, albeit with a slowdown in reaction times.
, ,
There is yet one further aspect of the interpretation of triggers under quantifiers that the experiments reported in Zehr & Schwarz (to appear) bear on. Charlow (; as well as Sudo ) claimed that ‘hard’ and ‘soft’ triggers differ in their projection patterns across quantifiers, with the former uniformly giving rise to universal presuppositions, while the latter vary between universal and existential ones across quantifiers. The true controls of Zehr & Schwarz (to appear) for both triggers only satisfied the presupposition existentially, however, and yielded ceiling-level acceptance rates for both triggers, suggesting that at least in the context of exactly one, both types of triggers only involve an existential presupposition. While the various reported findings here advance our understanding of the empirical picture concerning triggers in the scope of quantificational expressions substantially, the intricacies at both the theoretical and empirical level require much more work to achieve a more satisfying level of clarity of what the facts are and what analyses best fit them. Thus, this is a fertile area for further research, which will also be crucial for our understanding of the interpretation of presuppositions in embedded contexts more generally.9
.. Summary on embedded triggers The behaviour of presupposition triggers in embedded contexts is an intricate topic both in terms of theoretical and empirical perspectives. The empirical work to date suggests that projection from quantifiers is not uniform across all quantifiers, though there is little support for the notion that different triggers give rise to different projection patterns. Experimental results on projection from the scope of sentential operators provides some support for conditional presuppositions projecting from conditionals, as well as for symmetric interpretations paired with incremental effects that may be attributed to left-to-right processing. Evidence from disjunction, on the other hand, suggests that there is at least a strong preference for global, non-conditional interpretations, which seem to be available early on in processing, which opens up the possibility that other interpretations are best accounted for by local accommodation. Local readings, in turn, have been quite consistently shown to be less available than projecting interpretations, though again the extent to which this is true varies across triggers. As with global accommodation, it does not seem to be the case that there are triggers which resist local accommodation altogether, however, which also raises the question to what extent the same mechanisms are involved in local and global accommodation. A different perspective on local contributions of triggers opens up if we consider the possibility that some triggers also contribute their presupposition to the entailed content. Initial experimental results support the notion that triggers differ with regard to their contribution to the local content, though it is not yet clear whether that has to be captured in terms of presuppositions being entailed. Given the complexity of the issues involved, it is unsurprising that the empirical efforts so far have raised at least as many questions as they have begun to answer, but there clearly is a great need to push this field of study further in order for experimental results to have an even greater impact on theoretical questions. 9
The interpretation of presupposition triggers in the scope of attitude verbs constitutes another important area where experimental work is called for.
. P
.................................................................................................................................. While presuppositions are standardly seen to crucially relate to the discourse context, much of the work discussed so far does not incorporate more intricate notions pertaining to the structure of the discourse. But some recent lines of work move towards seeing presuppositions in the light of richer characterizations of discourse structure. First, we’ll discuss a different approach to investigating projection phenomena, which compares, among other things, intra- vs. cross-sentential resolution of presuppositions. Second, we’ll briefly survey experimental avenues for investigating the relation between presuppositions, projection, and QUDs. Finally, some initial experimental work on the phenomenon of ‘obligatory’ presupposition triggers is considered.
.. Resolving presuppositions in the discourse context Two studies relating to projection are concerned with the resolution of presuppositions in context, either intra-sententially or in the discourse context, and its time-course in processing. First, in two reading time studies using eye-tracking, Schwarz & Tiemann () find embedding of presupposition triggers to modulate processing effects. In the first study (already mentioned in section ..) immediate eye movement effects on the critical word are found when the context was inconsistent with the presupposition, but only when the trigger (German wieder, ‘again’) was outside of the scope of negation. No effects of context emerged when it was embedded under negation, and follow-up studies suggest that this is not due to a general availability of local interpretations. In a second study, presuppositional support for wieder in the consequent of conditionals is introduced in varying locations, namely in the antecedent or in a context sentence. Schwarz & Tiemann () interpret the results from this study as suggesting that the hierarchical distance in terms of the projection search path assumed by DRT directly affects reading times on the critical region. Such an effect is less straightforward to derive on non-representational accounts (such as dynamic semantics). Kim (), using the visual world paradigm, takes a different angle and investigates the effects of discourse structure on the selection of an antecedent for also. This is done by presenting multi-sentence discourses, which provide various possible antecedents for also in the final target sentence. In two initial comprehension studies, Kim asked subjects to choose one of several descriptions of what the sentence with also conveyed, which reflects how they resolve its presupposition in the discourse. While there was a general preference for linearly local antecedents in the comprehension studies (where also was understood relative to the immediately preceding sentence), a structurally (but not linearly) local interpretation also became available when the discourse structure was manipulated. In a visual world eye-tracking experiment, Kim also found a preference for structurally local interpretations. The eye movement results for the condition that involves a structurally local antecedent furthermore add to the evidence from the previous two studies, showing that the presupposition of also is available immediately in online processing.
, ,
.. Factives, prosody, and discourse In section ..., we mentioned recent theoretical developments, spearheaded by Simons et al. (), to explain the projective behaviour of certain triggers in terms of at-issueness, which in turn can be analysed in relation to the QUD. In particular, content is at-issue precisely if it helps to answer the QUD. In more recent theoretical work, Simons et al. (), this approach is fleshed out further for the case of factive verbs. An important aspect of this approach, which arguably sets it apart from more traditional approaches, is that whether or not a presupposition projects in a given discourse context will be affected by what the QUD is. Observations to that effect have been in the literature at least since Beaver (, with drafts circulated as early as ), who described effects of prosody (which on the relevant view is indicative of QUDs) on projection, but not until very recently have their been attempts to investigate this prediction experimentally, with a first published report in Tonhauser () (see Kim and Tonhauser, Chapters and respectively, in this volume). By way of background, one of the central contrasts in this line of work is represented by the following: () a. If the T.A. discovers that your work is [plagiarized]F, I will be [forced to notify the Dean]F b. If the T.A. [discovers]F that your work is plagiarized, I will be [forced to notify the Dean]F. (Beaver, : ) () a. Perhaps he discovered that she’s a [widow]F b. Perhaps he [discovered]F that she’s a widow.
(Tonhauser, )
With focal stress on plagiarized in () (or elsewhere within the complement of discover), the factive presupposition does not seem to arise: that is, this does not suggest that the addressee is guilty (though a weaker presupposition that the T.A. will discover something may be at play). In contrast, with stress on the factive discover itself, () does seem to take it as already settled that the addressed student is guilty of plagiarism. Tonhauser () uses carefully implemented prosodic manipulations of sentences like () to experimentally test this variation in projectivity (for an earlier approach testing general effects of prosody on the availability of different interpretations based on various aspects of meaning, including presuppositions, see Cummins & Rohde, ). She finds that when asked how confident they are in the embedded clause being true, participants indeed give significantly higher ratings when the focal accent is on discover, compared to either on (the equivalents of) she or widow, though confidence ratings for the projected content are relatively high across the board, at around out of . (An additional experiment revealed that the nature of the pitch accents in the complement clause could make a difference as well.) As Tonhauser lays out in detail, the effect of prosody on projectivity can be captured by the QUD-based approach, as only the versions with stress on the factive have a current QUD that entails the embedded clause, predicting higher confidence ratings in this condition. She also argues that traditional approaches that have to appeal to a notion of local accommodation to get non-projecting readings do not straightforwardly capture the
effects of prosody, as they only invoke this process as a last-resort strategy to avoid serious problems such as uninterpretability. It remains to be seen to what extent it may be possible to tie in QUD-based views of discourse with a more traditional perspective on presuppositions, which can call on local accommodation based on inferences about the discourse context suggested by prosody. Whichever direction theoretical advances take, the investigation of interactions between the behaviour of presuppositions and prosody is certainly an area ripe for more extensive experimental exploration.
.. Obligatory triggers One final area with some recent initial experimental approaches is that of so-called obligatory presupposition triggers and corresponding inferences (sometimes called antipresuppositions or implicated presuppositions). It has long been noted that it often seems to be the case that when a presupposition can be used, it must be used to render the utterance felicitous, as illustrated in the following examples: ()
{#A/✓ The} sun is shining.
()
John {#thinks/✓ knows} that Paris is in France.
() John came to the store. Bill did {#/✓ too}.
(Bade, : )
Initial theoretical discussions of the definite article go back to Heim (), who proposes a principle, Maximize Presupposition, that requires speakers to make their contribution presuppose as much as possible (another related line of work investigates the impact of informativeness on referential expressions more generally; see Davies and Arnold, Chapter in this volume). This approach has been further developed and refined by various subsequent authors (e.g. Percus, ; Chemla, ; Singh, ; Schlenker, ; Sauerland, ). An alternative proposal building on insights by Krifka (b) and Sæb () was developed by Bade (). The central idea is that leaving out the trigger would give rise to exhaustivity inferences (which she analyses as obligatory implicatures) that are incompatible with the context. More specifically, Bade assumes that sentences are interpreted as exhaustively answering the implicit QUD. While the reader has to be referred to Bade’s work for further details, Bade () crucially homes in on two predictions on which the two accounts come apart: first, Maximize Presupposition predicts that the relevant inference remains present even when embedded under operators such as negation, whereas Bade’s Obligatory Implicature approach does not predict them to arise in that context. Second, only the Obligatory Implicature approach predicts that the presence of the inference could be modulated by varying the QUD. Bade () reports experimental investigations of additives, iteratives, and definite descriptions to assess these predictions. Her main findings, based on acceptability rating studies, are that the first two do not give rise to the relevant inference under negation but are subject to modulation by contextual manipulations affecting the QUD, suggesting that the relevant inferences in these cases are best accounted for by the Obligatory Implicature approach. (Interestingly, Bade also argues know to pattern with additives and iteratives in these regards, though she does not present experimental data on
, ,
this.) In contrast, definite descriptions pattern essentially the opposite way, suggesting that they are cases where the relevant inferences are attributable to Maximize Presupposition. From this brief discussion, it should already be clear that the phenomenon of obligatory triggers and corresponding inferences provides rich grounds for more extensive experimental exploration, in particular in light of the potential interplay of different principles and different aspects of meaning, including exhaustivity, implicatures, and presuppositions. Bade’s finding that there does not seem to be a uniform picture with regard to how triggers pattern relative to crucial theoretical predictions furthermore contributes to our understanding of the variations in the behaviour of presupposition triggers, which need to be integrated into the broader theoretical debate about distinguishing types of triggers.
. C
.................................................................................................................................. Recent years have seen rapid growth of experimental work on phenomena related to presuppositions, and quite a bit of progress has been made already, in refining methodologies, clarifying the empirical landscape, and consequently informing theoretical debates. Even in fairly artificial experimental contexts, the contribution of presuppositions is largely robust and automatic. The interplay of presuppositions with both their intra-sentential and larger context requires careful control over all aspects of experimental stimuli. Many results have lent further support to the notion that (classes of) triggers differ from one another in various ways, but these difference are neither absolute or categorical, nor do they straightforwardly support any current conceptual approach to differentiating triggers. While all aspects of the study of presupposition will benefit from further experimental work, the behaviour of embedded triggers and the relation of triggers to more intricate aspects of discourse and discourse structure seem like an especially important area that deserves further scrutiny.
A The content of this chapter draws extensively on previous overview papers, in particular Schwarz (a, a), though in reorganized form. The summaries of experimental work in sections ., ., and .. in particular have largely been adapted from sections and of Schwarz (a). The overview provided in this chapter has greatly benefited from useful discussion with numerous colleagues and collaborators over a number of years; special thanks are due to close collaborators on related projects, in particular Jacopo Romoli and Jérémy Zehr, as well as attendees of my lab meeting group at the University of Pennsylvania. Chris Cummins provided helpful editorial feedback. Part of the work on this chapter was supported by NSF grant BCS-.
......................................................................................................................
......................................................................................................................
. I
.................................................................................................................................. T representation of space is a fundamental cognitive ability and all human languages can and do represent space. There is good evidence that spatial language is organized using a set of basic principles that include a shared—potentially universal—set of non-linguistic spatial distinctions (Miller & Johnson-Laird, ; Mandler, ; Landau & Jackendoff, ; Bowerman, ). Nevertheless, careful examination of the means different languages use to encode space has revealed considerable cross-linguistic divergence (Choi & Bowerman, ; Levinson, , ; Levinson & Wilkins, ). There is currently a wealth of experimental evidence on how both shared and language-specific factors conspire to shape the nature of spatial language and the way spatial terms are acquired and processed. In this chapter, we provide a selective review of this large literature focusing on three main subdivisions of the spatial domain: location (i.e. the static position of an object in space); motion (i.e. the dynamic displacement of an object in space); and Frames of Reference (FoR; i.e. abstract spatial-coordinate axes imposed on spatial configurations). Towards the end of the chapter we consider the possibility that spatial language itself could affect the non-linguistic representations of spatial categories.
. L
.................................................................................................................................. Languages analyse the location of an object in terms of three elements: the object to be located (figure), the reference object (or ground), and the relationship between the two (e.g. containment, as in English in, or support, as in English on; Talmy, ; Landau & Jackendoff, ). It is widely recognized that, in several respects, the linguistic encoding of location reflects a set of shared, pre-existing conceptual notions that constrain both the nature and the acquisition of spatial vocabulary across languages. Several sources of evidence support this position. First, there are many similarities in the way the cross-linguistic encoding of
location is organized (Miller & Johnson-Laird, ; Landau & Jackendoff, ). For instance, there are principles of figure-ground assignment that characterize all human languages and probably originate with non-linguistic principles of spatial organization (Talmy, ; Landau & Jackendoff, ). As an example, typically, the smaller, more mobile object in a configuration is treated as the figure and the larger, more stable object as the ground (e.g. The laptop is on the desk); reversing this expectation makes a sentence sound odd (e.g. ?The desk is under the laptop). Second, infants, during their first year of life, already know a lot about the spatial properties of objects in the physical world. Studies using preferential looking time paradigms show that, at . months, children can reason about containment (e.g. Hespos & Baillargeon, ) and, at to months, they can already form a basic representation for the relations ‘above’ and ‘below’ (Quinn, , ; Quinn et al., ). At around months, infants can distinguish between containment, support and occlusion (e.g. Baillargeon et al., ; Aguiar & Baillargeon, ; Casasola & Cohen, ; Casasola et al., ; Casasola, ) and, at to months, they can form a category for the relation ‘between’ (Quinn et al., ). Preverbal infants can distinguish between relations that their native language does not encode. For instance, infants growing up in an English-speaking community can distinguish between tight-fit and loose-fit containment and support relations, although their native language does not systematically encode this distinction, while other languages do (Casasola & Cohen, ; Casasola et al., ; McDonough et al., ; Hespos & Spelke, ). Third, and relatedly, studies directly comparing linguistic and non-linguistic understanding of static spatial relations in slightly older children have found that non-linguistic understanding precedes the acquisition of spatial terms. For instance, E. Clark (a) demonstrated that children understood the notions of containment and support when playing with objects earlier than the age at which they fully acquired the meanings of the prepositions in and on. Levine and Carey () reported similar results with the axial terms front and back. Such findings suggest that concepts of location precede (and presumably structure) the acquisition of locative terms in language. Fourth, children acquire locative terms in a consistent order cross-linguistically (e.g. Ames & Learned, ; Parisi & Antinucci, ; Brown, ; Grimm, ; E. Clark, , ; Johnston & Slobin, ; Weissenborn, ; Johnston, ). In an influential study, Johnston & Slobin () found that children across different languages produced spatial adpositions close in meaning to the English terms in, on, under, and beside earlier than the prepositions between, in front of, and behind. It was proposed that this cross-linguistically robust timetable reflected the order in which children develop the corresponding nonlinguistic spatial notions: in, on, beside, and their synonyms rely on simple topological concepts such as containment, support, and proximity (see Piaget & Inhelder, ; but see Coventry & Garrod, , for a more complex picture). Other terms such as between concern the relation between three objects and may, thus, be more complex. Similarly, axial terms such as in front of and behind rely on spatial coordinate systems and involve complex computations of figure-ground relations (see section . on FoR terms; cf. Rabagliati & Srinivasan, Chapter in this volume, for a discussion of the relation between words and concepts in non-spatial domains). Despite these commonalities, languages differ greatly in the ways they express locative information. One difference concerns the formal devices used to mark locative meaning.
In English and many other languages, locative information is encoded in adpositions (prepositions or postpositions). Other languages lack adpositions (e.g. the Australian languages Jaminjung and Arrente) and encode figure-ground relations through locative case marking on the ground Noun Phrase (NP; and an optional positional case on the verb). Yet other languages (e.g. the Mayan languages, Tzeltal and Yukatek) have only a limited set of general adpositions and package locative information into a rich inventory of spatial verbs (Levinson & Wilkins, ). More importantly, languages differ in the way they carve up the semantic space of location. In a series of studies, Melissa Bowerman and her colleagues have documented such differences in the domains of containment and support or attachment (Bowerman et al., ; Bowerman, ; Bowerman & Choi ; Gentner & Bowerman, ; see also Levinson & Wilkins, ). For example, in English, the preposition in is used for containment (e.g. apple in bowl) and the preposition on is used for a series of support relations: (a) ‘support from below’ (e.g. mug on table), (b) ‘clingy attachment’ (e.g. stamp on envelope), (c) ‘hanging against’ (e.g. poster on wall), (d) ‘point to point attachment’ (e.g. apple on branch), (e) ‘encirclement with contact’ (e.g. ribbon on candle). In Dutch, as in English, a single preposition (in) is used for containment scenes (apple in bowl), but the English on space is partitioned into three prepositions: op used for support-from-below and clingy attachment (a b above), aan used for hanging support (c d above) and om used for encirclement (e above). In Spanish, a single preposition (en) is used to describe all the above relations. And in Korean, the degree-of-fit between figure and ground is marked in a way that cross-cuts containment and support: Korean speakers use the verb kkita for tight-fit containment and support relations (e.g. earplug in ear, top on pen) and the verb nehta for loose-fit containment and encirclement (e.g. ball in box, loose ring on pole; Choi & Bowerman, ; but see Kawachi, ). These cross-linguistic differences play an important role in the acquisition of locative terms. English-speaking children learn to encode support (via on) earlier than their Dutchspeaking peers (who have to learn a more complex, three-term system); by contrast, learners in both language groups acquire containment expressions (in) around the same time (Gentner & Bowerman, ). Furthermore, by age , children already adopt language-specific locative encoding patterns, with English learners organizing spatial meanings around the containment/support distinction and Korean learners organizing spatial meanings around the tight/loose fit distinction (Choi & Bowerman, ; see also Bowerman, ). Recently, a more systematic comparison of the ways that containment and support are described by children and adults cross-linguistically suggests that one has to look at detailed semantic profiles within each of these relations to capture the intricacies of spatial language and its acquisition (Landau et al., ). This work reveals a principled but highly complex interplay of shared and language-specific contributions to how spatial language is used and learned.
. M
.................................................................................................................................. Languages analyse motion events as the displacement of a moving entity (figure) in relation to a reference object (ground), along a trajectory (path), and in a specific manner
(Talmy, ). For example, in English the sentence The cat jumped from the couch into the basket includes a figure (the cat), a manner (jumped), and two path expressions specifying the source (from NP) and the goal or endpoint of the path (into NP), each with respect to a specific ground (the couch for the source path and the basket for the goal path). As with locative terms, there are good reasons to assume that linguistic-motion primitives correspond to pre-linguistic, probably universal, conceptual-motion primitives that shape motion vocabulary across languages. First, some basic motion concepts are available early on; infants in the first year of life detect changes in the path and manner of motion events (Pulverman et al., ) and are able to form a category for the invariant path or manner of motion across different exemplars of manner or path (Pruden et al., ). Interestingly, infants form path and manner categories independently of the encoding preferences of the linguistic environment in which they are growing up (Pulverman et al., ). Second, there are homologies between the way motion terms are used and acquired and the way humans process motion non-linguistically. A case in point is a well-documented asymmetry between goal and source paths. In language, both children and adults tend to mention goal path expressions (e.g. into the basket) more often than source path expressions (e.g. from the couch) when describing motion events (Landau & Zukowski, ; Lakusta & Landau, , ; Regier & Zheng, ; Papafragou, ). The goalsource asymmetry has also been documented in the speech of brain-damaged patients (Ihara & Fujita, ), and children with Williams syndrome, a rare genetic deficit that causes spatial impairment (Landau & Zukowski, ), as well as in the spontaneous gestures of congenitally deaf children who have never been exposed to conventional language (Zheng & Goldin-Meadow, ). Furthermore, across languages, goals are encoded with greater specificity than sources (Regier & Zheng, ; Johanson & Papafragou, ) and this asymmetry affects the way both child and adult learners generalize novel motion expressions (Papafragou, ). The linguistic source/goal asymmetry has its roots in non-linguistic motion cognition. Both children and adults are better at detecting changes of landmarks or spatial configurations in goal compared to source paths (Regier & Zheng, ; Papafragou, ). Furthermore, this non-linguistic source/goal asymmetry is already present in -month-old infants (Lakusta et al., ; cf. Lakusta & Carey, ). This evidence, thus, suggests a strong (albeit imperfect; Lakusta & Landau, ) homology between language and cognition. Despite being rooted in a shared conceptual typology, the linguistic encoding of motion is characterized by intense typological variability. Both the ways motion primitives are lexicalized in spatial vocabularies and the ways these primitives are conflated into sentential structure vary considerably cross-linguistically. For instance, some languages (e.g. Romance, Japanese, Greek, Turkish) tend to encode the path of motion in the main verb (e.g. in French entrer ‘enter’, sortir ‘exit’, descendre ‘descend’) and the manner of motion (optionally) in an additional clause or gerund (e.g. en courant ‘running’); by contrast, other languages (e.g. English, German, Russian, Chinese) package manner information in the main verb and path information in particles or prepositions (e.g. up, into; Talmy, ; see also Chen & Husband, Chapter in this volume). Several studies have confirmed these cross-linguistic preferences in motion encoding in both adults and children (e.g. Berman & Slobin, ; Slobin, , ; Naigles et al., ; Papafragou et al., , ; Hickmann, ; Allen et al., ; among others). For
instance, Papafragou et al. () found that English-speaking - to -year-old children and adults used primarily manner verbs to describe motion scenes (e.g. The frog is jumping into the room), while Greek-speaking participants used primarily path verbs (e.g. O vatraxos beni sto domatio ‘The frog is entering the room’). These language-specific preferences for encoding motion information affect how newly encountered motion terms are interpreted. In one study, adult speakers of English and Spanish watched simple motion events (e.g. a woman skipping towards a tree) and heard a novel motion verb describing the event. Spanish-speaking adults interpreted the novel verb as a path verb, while English-speaking adults interpreted it as a manner verb, thus following the motion lexicalization preferences of their language (Naigles & Terrazas, ). A similar bias towards language-specific interpretations has also been documented in children from different linguistic backgrounds, at least from age (e.g. Maguire et al., ; Papafragou & Selimis, ; Skordos & Papafragou, ; cf. Hohenstein et al., ).
. F
.................................................................................................................................. FoRs are abstract coordinate systems for locating a figure object in space in relation to the axes defined by or imposed onto a reference (ground) object. Languages distinguish three FoRs: the intrinsic, the relative, and the absolute (Brown & Levinson, ; Levinson, , ; Pederson et al., ; see also Shusterman & Li, a, for somewhat different terminology). The intrinsic FoR describes the location of a figure object in terms of the inherent properties (e.g. front/back, top/bottom) of the ground object, often one’s own body (e.g. The tree is in front of the house/me). The relative FoR describes the location of a figure with respect to a ground object that lacks inherent sides (e.g. ball, tree, bottle) in terms of the speaker’s or some other observer’s viewpoint (e.g. The ball is to the right of the table). The absolute FoR describes the location of the figure with respect to environmentbased coordinates such as cardinal directions, the solar compass, wind directions, mountain slopes etc. (e.g. The forest is to the north of the village). Different coordinate systems for locating objects are available in pre-linguistic infants (see Newcombe & Huttenlocher, ; Quinn, ). Early studies on spatial orientation suggested that infants in the first year of life use their own bodily coordinates to code object location and they become sensitive to environment-based coordinates such as landmarks only later (Bremner & Bryant, ; Acredolo, ; Rieser, ). However, subsequent work demonstrated that both coordinate systems are available early on and that the choice of system is context-dependent (Bremner, ; Acredolo, , ; Acredolo & Evans, ). Different types of coordinate systems are also available to non-human species (see Gallistel, , Gallistel & Cramer, , for reviews). Cross-linguistically, there is considerable variation in the availability or frequency of use of different FoRs. All languages have terms to describe the intrinsic FoR (if only in rudimentary form; Levinson & Wilkins, ) but the distribution of the other two FoRs differs. English and Dutch make use of both the relative and the absolute FoR but prefer the relative FoR for small-scale arrays (e.g. The ball is behind the table). Tzeltal and Arrente
mostly make use of the absolute FoR, even for small-scale arrays (e.g. The ball is to the north of the table), and lack relative terms altogether (Levinson & Wilkins, ). Even within the set of languages that share a frame of reference, there are differences in how FoRs work. For intrinsic FoRs, languages use different (and often fairly complex) criteria in assigning names to a reference object’s facets (Levinson & Wilkins, ). For instance, in English, the ‘front’ of an object is defined by canonical encounter (for people or animals), forward motion (for vehicles), functional orientation (for appliances), etc. (see H. H. Clark, ; Miller & Johnson-Laird, ; Harris & Strommen, ). However, in Tzeltal, object-part name assignment is completely dependent on the object’s internal geometry: for example, a stone lying down with a flat surface on the ground has its ‘face’ (i.e. the least flat side) upside down (Levinson, ; Levinson & Wilkins, ). For relative FoRs, one source of cross-linguistic variation is how the viewpoint of the observer is projected onto the ground object. In English, the sentence The ball is in front of the table typically means that the ball is located between the observer and the table. Thus, the table has acquired a ‘front’ by the observer through reflection, as if the table were ‘facing’ the observer. However, in Hausa, the same sentence means that the ball is in the region projected from the furthermost side of the table with respect to the observer (a position which would have been described by the term behind in English). Thus, in the Hausa relative system, the table’s ‘front’ faces the same direction as the observer (Hill, ; Levinson & Wilkins, ). Finally, absolute FoRs in the world’s languages are extremely diverse. Arrente has a fully abstract cardinal direction system (e.g. north, south, etc.), Tzeltal uses the terms ‘uphill’ (south) vs. ‘downhill’ (north), Yélî Dnye distinguishes between ‘up’ (east) vs. ‘down’ (west) and ‘hillwards’ vs. ‘seawards’, and Jaminjung between ‘upstream’ vs. ‘downstream’ (Levinson & Wilkins, ). Most acquisition studies of FoR terms have focused on learners of languages that have relative FoRs. Typically, in such languages, a single set of terms (e.g. front/back, left/right) marks both intrinsic and relative FoRs (Levinson & Wilkins, ) and the acquisition of these terms follows a cross-linguistically robust pattern (Johnston & Slobin, ; Rigal, , ). Children’s earliest knowledge of front and back emerges around age and corresponds to intrinsic FoR instances that take one’s own body as the reference object (Kuczaj & Maratsos, ; Levine & Carey, ), followed a year later by intrinsic FoR instances that take objects with intrinsic facets as reference objects (Goodglass et al., ; Grimm, ; Kuczaj & Maratsos, ; E. Clark, ; Tanz, ; Levine & Carey, ). Around the age of or later children show evidence of relative FoR meanings that involve applying their own viewpoint to objects without inherent fronts and backs and extending such uses to incorporate another person’s viewpoint (e.g. Goodglass et al., ; Grimm, ; Kuczaj & Maratsos, ; E. Clark, ; Tanz, ; Weissenborn, ; Levine & Carey, ; Johnston, ). The acquisition of left/right follows a similar sequence but lags considerably behind front/back, presumably because of additional computations required to differentiate the secondary left-right axis after the primary front-back axis has been defined (Elkind, ; Harris, ; Irwin & Newland, ; Rigal, , ; Shusterman & Li, b). This sequence of (sub-types of) intrinsic and relative FoRs has been attributed to the increasing conceptual demands on perspective-taking posed by the relative FoR in its various incarnations (cf. Piaget & Inhelder, ; but see Shusterman & Li, b, and Rubio-Fernández, Chapter in this volume, for more nuanced discussion).
Less is known about the acquisition of absolute systems. In a longitudinal naturalistic production study, Brown & Levinson () found that Tzeltal-speaking children start using absolute terms (‘uphill’, ‘downhill’) around age , but use them relationally (‘X is uphill of Y’) only by age ;, and integrate them into adult-like requests to others for manipulating objects in a tabletop array only by or (Brown & Levinson, ; see also De León, , on the acquisition of other absolute systems). Other work shows that absolute (environment-based) representations are also available to learners of languages that do not prioritize this FoR: Shusterman & Li (b) report that -year-old Englishspeaking children readily map absolute (north/south) meanings onto novel ambiguous spatial terms (e.g. It is on the ZIV/KERN side of the room; cf. also Haun et al., ). Beyond these basic patterns, the acquisition of the conventions regarding how a specific language community derives and uses FoR terms is quite protracted (Harris & Strommen, ; Abkarian, ), and there are key language-specific aspects in the profiles of different languages (see, e.g. De León, , and Brown & Levinson, , on the acquisition of intrinsic terms in learners of absolute languages).
. D ?
.................................................................................................................................. The evidence reviewed throughout this chapter suggests a tight causal relationship between spatial language and cognition, since the linguistic encoding of space builds on antecedently available, pre-linguistic spatial concepts in important ways. One might ask whether this causal relationship could be reversed—that is, whether spatial language itself might affect the way spatial categories are acquired, perceived, categorized, and remembered. If so, cross-linguistic differences in the encoding of space might create cognitive discontinuities among speakers of different languages. This topic has attracted a lot of recent attention within a larger discussion about the role of language in cognition (see Ünal & Papafragou, , for a recent review). In the domain of location, there is evidence that spatial semantic distinctions do not shape non-linguistic cognition. English makes a distinction between on and above but Japanese and Korean do not; nevertheless, categorization patterns for these spatial relations converge regardless of language background (Munnich et al., ). Other work has argued for the opposite conclusion. Recall that, unlike English that draws a distinction between containment and support in its prepositional system (in vs. on), Korean makes a distinction between tight fit and loose fit in its verbal system that cross-cuts the containment-support boundary (e.g. kkita ‘put tightly in/on/together/around’ vs. nehta ‘put loosely in/around’; Choi & Bowerman, ). Both English- and Korean-speaking infants distinguish between tight and loose fit when processing containment and support scenes; Korean-speaking adults continue to attend to the degree-of-fit when categorizing spatial relations of containment and support but English-speaking adults do not (McDonough et al., ; Hespos & Spelke, ; cf. also Choi, ). These results have been taken to suggest that linguistic encoding decreases the cognitive salience of degree-of-fit relations in mature English speakers. However, the empirical picture is complex and does not clearly support this conclusion: Norbury et al. ()
showed that both English- and Korean-speaking adults were sensitive to the non-linguistic dimension of fit; furthermore, for both groups, tight-fit relations were more salient than loose-fit relations. These results suggest that adults’ non-linguistic representation of fit does not depend on the language they have acquired (if anything, its structure is characterized by a bias that emerges in speakers of different languages and might involve a deeper perceptualcognitive asymmetry; Norbury et al., ). Similar results have been obtained in the domain of motion. Papafragou et al. () found that, although children and adult speakers of English and Greek described motion events differently, they did not differ in their memory and categorization of these events (see also Gennari et al., ). A further study (Papafragou et al., ) compared attention allocation to motion events (measured by eye movement patterns) in adult speakers of English and Greek. During a verbal description task, when people were preparing to describe these events, they allocated their attention to components of motion events in ways that reflected language-specific encoding patterns (i.e. speakers of English attended earlier to manner of motion and speakers of Greek attended earlier to path; see also Bunger et al., , for evidence of similar effects in young children). However, during a memorization task, when participants freely inspected the ongoing motion events, the languagespecific effects disappeared. These results suggest that motion event perception is not guided by the perceivers’ native language, even though language-specific patterns in attention emerge when the task specifically involves the recruitment of linguistic representations (as in language production). A striking finding from Papafragou et al. () was that, during the memorization task, after having seen the motion events unfold and when trying to memorize them, speakers of English and Greek attended to different components of the events (namely path for speakers of English and manner for speakers of Greek). A subsequent eye-tracking study (Trueswell & Papafragou, ) showed that these effects disappeared when participants were asked to perform a secondary task that engaged the language faculty (counting aloud) but not when they performed an equally taxing secondary task that did not engage the language code (tapping a rhythm). These results suggest that participants in Papafragou et al. () spontaneously recruited language online to support the representation of an event in memory: participants attended to aspects of motion events that were encoded outside the main verb in their native language and, thus, might be forgotten. However, the online recruitment of language to support cognitive operations was flexible and task-dependent; furthermore, such linguistic intrusions could be blocked by secondary tasks that interfered with linguistic encoding (see also Athanasopoulos & Bylund, ; Athanasopoulos et al., ). Finally, in the domain of FoR, Levinson and colleagues (Levinson, ; Pederson et al., ; see also Majid et al., ) compared how speakers of Dutch (a relative FoR language) and speakers of Tzeltal (an absolute FoR language) responded in various spatial tasks. For instance, Pederson et al. () tested participants in the Animals-in-a-Row task, in which participants first studied a line of toy animals on a table, then were rotated degrees and moved to a different table, where they were given the animals and were asked to make it ‘the same’ as what they saw before. Speakers of Dutch solved the task by applying a relative strategy (i.e. maintaining the left-right orientation of the animals in the row) but speakers of Tzeltal solved the task by applying an absolute strategy (i.e. maintaining the cardinal orientation of the animals in the row). Further studies have confirmed the presence of a correlation between the dominant FoRs in a specific language community and the use of
FoR representations in non-verbal cognitive tasks in members of that community (Haun et al., ; Haun et al., ). Such findings have been taken as indications that languagespecific preferences shape non-linguistic spatial cognition (Haun et al., ; Haun et al., ). Other work has challenged this view. Li & Gleitman () showed that speakers of English (a relative FoR language) could provide different responses in the Animals-in-aRow task depending on the testing conditions: when tested indoors without access to external landmarks, English speakers patterned with the Dutch speakers and used the relative solution; but when tested outdoors or indoors with landmark information present, English speakers were more likely to make use of the absolute solution, like the Tzeltal speakers. Similarly, Li et al. () showed that Tzeltal speakers could use relative strategies to solve rotation problems when given subtle hints about the solution that was sought by the experimenter; furthermore, Tzeltal speakers were more accurate in relative than in absolute solutions, elicited under similar conditions. This work suggests that participants may fall back onto the FoR linguistic conventions of the community when interpreting ambiguous instructions such as ‘make it the same’ (Li et al., ) but such biases do not limit the representation of FoR in cognition. Summarizing, despite the presence of cross-linguistic differences in the domains of locatives, motion, and FoR, the underlying cognitive representations are remarkably similar in members of different language communities. Nevertheless, spatial language regularly intrudes into cognitive processing, even when it is not explicitly invoked or necessary, especially when the task is cognitively demanding (Papafragou et al., ) or ambiguous (Li et al., ). As a result, cognitive and linguistic categories of the spatial world, although dissociable, are often highly correlated. This line of reasoning is consistent with evidence that overt use of spatial terms benefits children’s and adults’ performance on a variety of spatial tasks, including spatial categorization (Casasola, ; Casasola & Bhagwat, ), spatial analogy (Loewenstein & Gentner, ), spatial memory (Dessalegn & Landau, , ), and navigation (Hermer-Vazquez et al., ; Spelke & Tsivkin, ; Pyers et al., ; Shusterman et al., ). In all such cases, spatial language (whether covertly or overtly introduced) may augment representational or processing resources by helping identify, store, and/or manipulate spatial information.
. C
.................................................................................................................................. Decades of research on spatial terms have revealed a complex set of factors that shape the nature, use, and acquisition of spatial vocabularies. Several pieces of evidence support the conclusion that spatial language—at least in part—reflects a set of non-linguistic, potentially universal, cognitive spatial primitives. Nevertheless, detailed studies of individual linguistic systems make it clear that there are many differences in how individual languages talk about space. The literature we reviewed has highlighted the importance of conducting research with diverse populations (e.g. speakers of different languages, typically and atypically developing children, infants, non-human species) and studying spatial language and cognition with a variety of empirical methods (e.g. linguistic tests of production and comprehension, cognitive tests of memory and categorization, eye-tracking). The current state of the art calls for a nuanced position both on how spatial terms are acquired cross-linguistically (since learning
to speak about space does not involve a simple mapping between concepts and spatial terms) and on how spatial language connects to non-linguistic cognition. Future work should pursue the quest for linguistic-semantic spatial universals through both field work with speakers of many different languages and formal analyses of the semantics of space. Relatedly, future work needs to provide a fuller map of the non-linguistic cognitive presuppositions of spatial language through a variety of empirical methods (including neuroscientific approaches; e.g. see Burgess, ; Wolbers & Hegarty, ). A particularly rich avenue for further research in spatial language involves the role of pragmatic inference. Since language is limited and can only express certain aspects of nonlinguistic spatial representation (Talmy, , ; Landau & Jackendoff, ), pragmatic inference plays an important role in both how people choose spatial expressions as speakers and interpret such expressions as comprehenders (Herskovits, ; Levinson, b). For example, the English preposition in and related containment expressions across languages can be used to convey related but distinct relations such as full containment (‘coffee in a cup’) or partial containment (‘pencil in a cup’), depending on one’s knowledge about the specific objects in the scene (Herskovits, ; Levinson, , b). Addressees use implicature (Grice, ) or pragmatic enrichment (Sperber & Wilson, /; cf. Degen & Tanenhaus, Chapter in this volume) to add contextual refinement to coarse spatial meanings and speakers anticipate such means of reconstructing the exact spatial configuration conveyed through a spatial description. Such patterns of language use and interpretation have been argued to be impressively consistent across the world’s languages (Levinson, b; Levinson & Wilkins, ), although language-specific encodings may result in different pragmatic inferences for the same spatial configurations (Bowerman, ). Currently a growing body of experimental work documents pragmatic contributions to spatial meaning. Several cross-linguistic studies have shown that children’s and adults’ choice of spatial terms to describe space and motion scenes depends on whether these terms make an appropriate and specific informational contribution compared to other alternatives (Tanz, ; Papafragou et al., ; Papafragou et al., ; Grigoroglou et al., in press). A separate strand of research has shown that context provides sets of expectations that guide both the production of spatial descriptions and the interpretation of spatial language in conversation (Morrow & Clark, ; Coventry et al., ; Carlson & Covey, ; Carlson & Kenny, ; Coventry et al., ; Li et al., ; Ullman et al., ; Andonova et al., ). Finally, other work argues that general pragmatic principles affect the shape of cross-linguistic spatial systems themselves (Khetarpal et al., ; Khetarpal et al., ). The integration of these directions with research on spatial semantics and cognition is particularly promising for future research.
A Preparation of this chapter was supported in part by grant # from the National Science Foundation.
......................................................................................................................
......................................................................................................................
C reasoning, an understanding of events that are counter to reality, or false, is an essential ingredient of our everyday cognition. Counterfactuals are cases of possibly valid reasoning from premises that are false in actuality (Fauconnier & Turner, ), and require the comparison of reality to a model-based alternative. People understand a counterfactual statement, such as If money grew on trees then we’d all be millionaires, by keeping two possibilities in mind from the outset: the conjecture, that money grows on trees and we are all millionaires, and the presupposed facts, that money does not grow on trees and we are not all millionaires (Byrne & Tasso, ). The counterfactual thus requires a person to represent false information that is temporarily supposed to be true, and allows them to attach a label to their models to keep track of their epistemic status. Counterfactual situations are frequently depicted through language, yet it is only very recently that researchers have begun to empirically test how they are represented and accessed during language comprehension. The ability to reason counterfactually develops from a young age, as children as young as years old can understand the consequent meaning of nearly or almost, and are able to apply counterfactual thinking to pretend play (e.g. Riggs & Peterson, ). In fact, the development of fully functional counterfactual thinking is relatively protracted, continuing at least into middle childhood (Beck & Riggs, , ), and relies upon the development of domain-general processes including executive functions and language.
. H ?
.................................................................................................................................. Fauconnier () has catalogued a number of ways in which counterfactual worlds may be triggered, employing two dimensions of counterfactuality. The first is a lexical dimension, on which negatives are strongest as they automatically cancel one alternative (e.g. that John failed the exam) as in (a), followed by verbs like wish, as in (b). Lower on the lexical dimension, counterfactuality can be imposed or implied using conditionals (If-then constructions) (c) and modal terms, such as, could or might (d).
()
a. Luckily John did not fail the exam. He would have been thrown off the course. b. I wish Dave were strong like Superman. He could rescue me if I was in trouble. c. If Sarah were transferred to an office in France, then she would need to learn to speak French. d. John would have bought the expensive car if he had lots of money in the bank.
It is also known that tense influences the plausibility of counterfactual interpretation (e.g. Kratzer, ; Cowper, ). Consider, for example, the sentences in (). All three sentences express the same logical relationship between Bill coming tomorrow (the protasis B) and Susie making dinner. However, counterfactuality is produced in different ways according to the distribution of tenses. Specifically, for (a) to make sense, B must either be true (i.e. we know that Bill is coming) or undetermined, but it cannot be used if we are certain that Bill is not coming tomorrow. In contrast, (b) can be used if B is undetermined or if it has been established that Bill is not coming tomorrow (counterfactual), but not if B is true. Finally, (c) can only be used counterfactually as it directly implies that the speaker knows that Bill is not coming tomorrow. ()
a. If Bill comes tomorrow, Susie will make him dinner. b. If Bill came tomorrow, Susie would make him dinner. c. If Bill had come tomorrow, Susie would have made him dinner.
. C
.................................................................................................................................. An acceptance of some counterfactual alternative is commonly associated with story-telling and literature, where authors depict ‘what if ’ scenarios and the audience must make sense of events according to a non-real world. For example, readers must accommodate a counterfactual, magical world to make sense of flying broomsticks and talking animals in J. K. Rowling’s Harry Potter books, or Elves and Orcs in J. R. R. Tolkien’s Lord of the Rings trilogy. Counterfactual thinking also serves important social functions, for instance in reflecting on past events with negative outcomes (the ‘if-only . . . ’ effect of Kahneman & Tversky, ; see also Kahneman, ; Byrne, ). People tend to think not only about the events that actually happened but also how those events might have been different. For example, if you get stuck in traffic and are late for an important job interview, you might wonder whether you would have been on time if you had left earlier, taken a different route, or if you had taken the train. In fact, thinking through these alternatives is important in helping us to correct mistakes and avoid making them again. These sorts of counterfactual thoughts are pervasive in adult mental life and play an integral role in higher-level cognition such as deductive reasoning (Johnson-Laird & Byrne, ; Gilovich & Medvec, ). In sum, counterfactuals have been implicated in diverse cognitive activities, ranging from simple imagination beyond reality, and fantasy (e.g. Sternberg & Gastel, ), to the exploration of possibilities in reasoning (e.g. Byrne & Tasso, ; Johnson-Laird & Byrne, ).
There has been a very large amount of research on reasoning with counterfactuals (Byrne, , ), and on what sort of constraints there are on the kinds of counterfactual thoughts people are likely to produce in a variety of circumstances (e.g. Kahneman & Miller, ; Byrne, ; Markman & Tetlock, ). Many of these studies have revealed that counterfactuals are generated most frequently following a bad outcome, or goal failure (Roese et al., ). Specifically, ‘upward’ counterfactuals consider how a situation could have turned out better and thus serve a preparatory function to learning from mistakes. ‘Downward’ counterfactuals describe how a situation could have been worse, serving an affective function to make a person feel better. Psychological studies have indicated that there are considerable regularities in the sorts of counterfactuals people generate most readily, despite the infinite number of ways that past events could have happened differently. Their counterfactual thoughts tend to mentally undo the most recent event in an independent sequence. A temporal order effect has been proposed to explain why, for example, an individual is judged to be lucky when a good outcome is described after a bad one, or why a second penalty in a football match is well-rated after a first one that was poorly rated (Teigen et al., ). Walsh & Byrne () suggest five principles that explain the temporal order effect observed when people generate counterfactual statements. The first two assumptions are that people keep in mind the true possibilities, and that people think about just a subset of the possible counterfactual models. The third principle is that the subset of counterfactual possibilities that people keep in mind is guided by the winning conditions (i.e. the possibilities in which the players would have won, and not the circumstances in which they would have lost). The fourth principle is that people mutate the crucial elements of the facts to be like the winning conditions. Finally, the fifth principle is that the first element in the facts is an ‘anchor’ that is presupposed and remains relatively immutable (Byrne et al., ). Walsh & Byrne () carried out three experiments that provided the first empirical evidence that people think counterfactually about the same set of facts in different ways depending on the alternatives that have been made available. The participants’ task was to reason on the outcome of a colour card game between two players, where the facts were always the same (both players choosing black cards), but different descriptions of the winning conditions were given. Winning descriptions either referred to black cards (e.g. ‘if one or the other but not both picks a black card, each individual wins £,’) or to red cards (e.g. ‘if one or the other but not both picks a red card, each individual wins £,’). Results showed that participants were more likely to undo the most recent event (i.e. showing a standard temporal order effect) when the task instructions described the black card, and the first event (i.e. showing a reversed temporal order effect) when the task instructions described the red card. Thus, results showed that the temporal order effect can be produced or reversed by different descriptions of the winning conditions. In a similar study, Meehan & Byrne () investigated how the temporal order effect found in adults extends to children’s counterfactual thoughts. The experiment used the same colour card game as Walsh & Byrne (), and showed that children aged and years exhibited the standard temporal order effect when they were asked to think about what might have been, but they differed in their judgements of guilt and blame. While -year-olds, like adults, judged that the second player to select a card will feel more guilt and be blamed more by the first player, the temporal order effect does not occur for -yearolds’ judgements of guilt and blame. This dissociation of counterfactual thoughts and social
judgements in -year-olds indicates that their creation of counterfactual alternatives has not yet fully developed.
. H ?
.................................................................................................................................. In contrast to research within the framework of reasoning and its social concomitants, cognitive theories of how counterfactuals are understood during language comprehension are limited. To tackle this issue, we first focus on some issues that have been raised by philosophers concerning the logic of counterfactuals. Then, theories of language comprehension will be considered to explore what these models might be able to tell us about the comprehension of counterfactuals (see also Chen & Husband, Chapter in this volume, for relevant discussion of event structures).
.. Mental spaces Philosophers have tried to evaluate the logical truth conditions of counterfactual statements. Goodman () investigated the problem of determining which true statements are combined with the false premises to carry out the reasoning, while Lewis () considered the problem of determining when and which logical laws apply to counterfactuals. However, cognitive-semantic speculations about how counterfactual spaces are set up and structured were initiated by Fauconnier (, ) who applied a ‘mental spaces’ approach to the comprehension of counterfactuals. Mental spaces are defined as structured, incremental sets that include elements and relationships between them, with availability for new elements to be added and new interactions between the elements to be created. Thus, these mental spaces, and the relationships between them, are a way of specifying an interpretation of a discourse. According to Fauconnier, two mental spaces are produced in the case of counterfactual conditionals; one is the reality space and the other is the counterfactual hypothetical space whose structure is analogical, and not truth-functional (see Goodman, ; Lewis, ). Therefore, counterfactuality is seen as a case of forced incompatibility between these two spaces, since what is true in the counterfactual space is false in the reality space. Let us see how this approach can be applied. If we take the current space that a discourse occupies to be the parent space R, then an utterance of the form, ‘If p, then q’ generates a second space H in which p and q hold. So, in the example, If money grew on trees then we’d all be millionaires, the phrase If money grew on trees acts as a space builder for the counterfactual hypothetical space H. Therefore, real-world knowledge that money does not grow on trees is linked to the counterfactual’s consequence, we’d all be millionaires, in the H-space. Fauconnier (, ) suggests that there is no general linguistic algorithm to go from reality to counterfactual hypothetical space, but emphasizes that some pragmatic situations are likely to impose further constraints on the structural dependencies of the counterfactual hypothetical space with respect to reality. For example, if the arguments relate to physics, then all the laws of physics (e.g. gravity, time) that exist
in reality will also apply in the hypothetical counterfactual space, although the H-space may include some explicitly specified elements that are counter to reality. By this logic, a counterfactual conditional construction results in a hypothetical space that possesses the same constraints as those in reality. In sum, Fauconnier (, ) presents some very interesting analyses of what is entailed in counterfactual worlds, although his analyses do not provide any specific basis for predicting how propositions are processed with respect to real-world and counterfactual-world spaces.
.. Situation models On the other hand, evidence from the discourse processing literature suggests that comprehending a text involves the construction of a mental representation of the information, known as a situation model (Glenberg et al., ; Graesser et al., ; Zwaan & Radvansky, ). Situation models are non-linguistic representations that are composed of four primary types of information: a spatial-temporal framework, entities (e.g. people, objects, ideas), properties of those entities, and the relationships between the given information (e.g. spatial, temporal, causal). The information is then represented across three levels, from the situation model (event-specific), to the episodic model (coherent sequences of events), and finally to the comprehensive model (a complete collection of episodes). Thus, these mental simulations are considered to be experiential in nature as they are assumed to be grounded in perception and action. According to this proposal, understanding language entails recapitulating motor or perceptual activity as a ‘simulation’ of the language input. A number of findings have been produced over the last decade to demonstrate the validity of this mental-simulation view (see Zwaan, , for an overview). For example, neuroimaging studies have revealed that processing action words associated with specific body parts (e.g. arm—pull) activates consistent brain regions, including areas of the motor and premotor cortex, that are involved in perceiving or performing those motor actions with the same effectors (e.g. Pulvermüller, , ; Martin & Chao, ; Vigliocco et al., ). Interestingly, several fMRI studies (Tettamanti et al., ; Tomasino et al., ) have shown that this motor embodiment can be reduced when an action word is presented within a negated frame (e.g. ‘not grip’, ‘not push’). These studies support the view that understanding language can involve an embodied component, perhaps comprising a mental simulation of actions using specific effectors. Further evidence comes from behavioural tasks that show facilitation or interference effects depending on the degree of similarity between the language input and a response, based on the specific motor or spatial representations that the language activates. These studies rely on the idea that language understanding entails running a mental simulation of a described event, in which the comprehender activates motor representations corresponding to what participants in the scene might do and perceptual representations of images they might perceive. Therefore, processing information is facilitated if the perceived action matches an action required by the experimental task, but processing is impeded if these two mismatch or if processing of one interferes with processing of the other (for example in a speeded decision task). Glenberg & Kaschak () compared sensibility judgements as participants were presented with sentences such as He closed the drawer vs He opened the drawer. Importantly, half the experimental items described a
movement towards the participant (‘opened’), while half involved a movement away from the participant (‘closed’). These language input conditions were crossed with two response conditions, where participants responded to ‘sensible’ sentences by moving a lever away from or towards their body. Responses were faster when the response movement matched the movement implied by the language input (i.e. He closed the drawer with a movement towards the comprehender) than when there was a mismatch (i.e. movement away from the comprehender). Similarly, responses to a visually presented stimulus were faster when the depicted object’s shape and orientation matched information implied in a preceding sentence (Stanfield & Zwaan, ; Zwaan et al., ). Specifically, responses were facilitated when a sentence such as The ranger saw an eagle in the sky was followed by an image of an eagle with its wings outstretched, compared to trials where this information mismatched, such as an image of an eagle with folded wings. Another finding was reported by Zwaan & Yaxley (a,b), extending the mismatching results to single words rather than sentences. Participants were asked to judge the semantic relatedness of two words (e.g. attic—basement) presented one above the other on a computer screen. Responses were faster when the spatial relation on the screen was consistent with the spatial relationship in the real-world (i.e. attic on top) than when the two were inconsistent (i.e. basement on top). These studies therefore provide more support for the suggestion that thinking about the meaning of a word makes perceptual information available. Finally, studies on the comprehension of negatives reveals that information that is involved in a given situation is more available in the comprehender’s brain than information that is not directly involved (e.g. Glenberg et al., ; Zwaan et al., ; Kaup & Zwaan, ). Importantly, the issue of embodiment has been investigated in relation to counterfactual information (Urrutia et al., ; see also De Vega, ; De Vega & Urrutia, ). Experimental items described a transference situation, either away from me or towards me, within a factual or counterfactual sentence. Thus, sentences were either factual away from me (a), factual towards me (b), counterfactual away from me (c) or counterfactual towards me (d). ()
a. Because I am generous I’ve lent you the novel of Harry Potter. b. Because the jeweller is a good friend of mine he has shown me the imperial diamond. c. If I had been generous I would have lent you the novel of Harry Potter. d. If the jeweller had been a good friend of mine he would have shown me the imperial diamond.
Sentences were presented one word at a time, and crucially, the transfer word (e.g. ‘lent’ or ‘shown’) appeared to move visually towards or away from the participant (by getting smaller or larger on screen) during reading. Participants were asked to respond to this transfer word by pressing either a towards- or an away-key according to the perceived motion of that visually presented word. The Inter-Stimulus Interval (ISI) between the transient word presentation and its apparent motion was manipulated between participants ( vs milliseconds (ms)) so that the temporal overlapping between the word and perceived motion varied. Response times to the animated word showed a similar mismatching advantage for both factual and counterfactual sentences. In other words, at
these short ISIs, where word animation and semantic integration processes overlapped, responses were slower when the animation matched the meaning of the transfer word than when the perceptual and semantic information mismatched (e.g. towards-away conditions > away-away conditions). However, for factual sentences these effects were revealed in the ms interval whereas the effects for counterfactual sentences emerged in the ms interval and had disappeared by the ms interval. These results suggest that when a transference construction is embedded in a counterfactual sentence, an embodied simulation of its literal meaning is automatically generated, at least temporarily. Taken together, these findings suggest that the ‘simulation’ aspect of situation models plays a role in comprehension so that when comprehenders process some linguistic input they mentally simulate the described state of affairs using similar processes to those that would be activated if they were experiencing the described events directly. However, evidence to date does not support a view where embodied meaning is a necessary, realtime component of language comprehension in general (Sanford, ). As such, it is difficult to relate situation models to the representation of counterfactuals as they make no clear proposals about how possibly valid information that does not contain an embodied concept (e.g. ‘lend’) could be represented within this framework.
.. Mental models An alternative theory of mental models has been proposed by Johnson-Laird (; see also Garnham, ). Mental models are often referred to alongside situation models, with little distinction between the two, however, clear differences separate the two accounts. Specifically, while situation models describe representations of language as a mental simulation of the states of affairs described in the text, mental models focus on the way mental representations correspond to information about the real or imaginary world, thus focusing on the logical structure of events. As such, mental models are mental representations that are structurally close to a model of the world rather than a model of the language that describes it. Extensive research has been conducted on mental models and has demonstrated that mental models play an integral role in the comprehension and representation of written texts (e.g. Gernsbacher, ; Garnham, ; Oakhill, ; Bower & Rinck, ; Rickheit & Habel, ). The mental models theory has been developed within a simple information processing framework, spanning concepts such as position, spatial layout, and time. This model attempts to capture representations of reality and to enable inferences between concepts, thus it has proved particularly useful in accounting for deductive reasoning. However, growing evidence suggests that reasoning about factual and counterfactual situations relies on similar mental processes and that inferences are drawn from imaginary situations just as they are from real situations, by constructing mental models (e.g. Byrne, , ). Consider the example ‘If it had rained, the children would have played indoors’. In cases such as these, Byrne () suggests that readers do not represent the false alternative in which it rained and the children did not play indoors as this conflicts with the counterfactual conditional framework. Instead, representations are made of the factual situation, as indexed by the counterfactual conditional, where it did not rain and the children did not play indoors. This supports an earlier suggestion that typically people use mental models to represent information that is true (Johnson-Laird & Byrne, ).
In summary, because the mental models theory handles deductive reasoning, it naturally offers possibilities for representing counterfactuals. Furthermore, this theory has a ‘core’ extensional account of conditionals, making a conditional ‘if p then q’ logically equivalent to ‘not-p or q’. Consequently, when this account is applied to the case of counterfactual conditionals, it makes the clear proposal that both factual and counterfactual possibilities are represented by the reader.
.. Suppositional theory An alternative view to that of how mental models might be applied to counterfactual processing follows from work by Ramsey (). This account proposes that when comprehending a conditional statement, people ‘hypothetically add p to their stock of knowledge and argue on that basis about q’; a practice commonly known as the Ramsey test. Proponents of this account have challenged the mental models theory (Evans & Over, ; Evans et al., ), specifically referring to problems when participants are invited to express degrees of confidence in their conclusions to conditional arguments. In these cases, levels of confidence directly relate to their belief in the conditional premise (Stevenson & Over, , ; Liu et al., ; George, , ), suggesting that people’s mental representations of conditional statements involve varying degrees of association between p and q and between q and p. However, the mental models theory does not allow for both directional links and belief tendencies. As a substitute, Evans et al. (; see also Evans & Over, ) have proposed a suppositional theory where a conditional of the form ‘if p then q’ directs attention to possibilities following from p, and not to ‘not-p or q’ possibilities. According to the suppositional account, when a person encounters a counterfactual possibility she temporarily adds this possibility to her store of beliefs by creating a model via a process of minimal change to reality. So, in a counterfactual statement such as If cats were vegetarians, people will make a minimal change to their representation of what a cat is in order to suppose that cats are vegetarian. Then one can ask how confident they are in the consequent (e.g. you could feed a cat a bowl of carrots). Importantly, when applied to counterfactual reasoning, the suppositional view makes very different predictions to the mental models view as it suggests that people will, under some circumstances, consider a case where the supposed antecedent holds but the consequent does not. Thus, the supposition is about the closest appropriate possibility, according to a person’s judgement, meaning that counterfactual statements should be evaluated with respect to suppositional or hypothetical possibilities first.
. E
.................................................................................................................................. Until recently, research on counterfactuals was limited to philosophical and logical theoretical work, meaning that very little was known about how these counterfactual worlds are
set up and accessed during understanding. However, in the last decade or so, this question (and others) has begun to attract interest from psychologists, with an emerging body of empirical research testing comprehenders’ ability to make inferences about counterfactual events, and the activation of multiple representations of the world (Kulakova & Nieuwland, ). Much of this research has employed sensitive online research techniques (eye movements and Event-Related Potentials (ERPs)) that allow detailed exploration of the unfolding processes and representations that underlie counterfactual comprehension over time. These techniques are advantageous because they provide a great deal of multidimensional information about the temporal dynamics and specific cognitive functions involved in language processing. The majority of this empirical research has focused on understanding how the consequence of a counterfactual context is processed, and this question will be the focus of our present discussion. However, a handful of recent studies have begun to investigate effects during processing of the antecedent itself (e.g. Stewart et al., ; Haigh & Stewart, ; Kulakova et al., ; Kulakova & Nieuwland, in press). This early research suggests that interpreting a counterfactual antecedent incurs increased processing costs, due to the dual representations that are activated, although further research in this area is necessary to rule out alternative language-based explanations.
.. Evidence for rapid access to the explicit counterfactual world Empirical research has demonstrated that comprehenders have rapid access to the explicitly described counterfactual model of the world. The first piece of evidence for this comes from a study that employed the visual world paradigm to investigate people’s unfolding expectations of forthcoming language input. Visual world research involves tracking eye movements around a visual scene, closely time-locked to some related auditory input. Over twenty years of intense research using this paradigm has demonstrated that auditory information can direct eye movements towards appropriate objects in a visual display (Cooper, ; Tanenhaus et al., ). Moreover, growing evidence has shown that the eyes can move towards a critical object before a word referring to that object is available. It is commonly believed that such languagemediated eye movements reflect the cognitive processes that underlie language comprehension. For example, in a study by Altmann & Kamide (), participants listened to auditory sentences such as, ‘The boy will eat the cake’ or ‘The boy will move the cake’ alongside a visual scene depicting a boy, a cake, and some inedible objects. Results showed that people looked faster to the cake in the ‘eat’ condition than in the ‘move’ condition. More importantly, these eye movements were anticipatory (i.e. the difference was apparent before the auditory onset of ‘cake’), showing that visual attention is drawn to critical objects in the scene before they are actually mentioned in the auditory input. As such, the visual world paradigm is ideal to study top-down expectations in counterfactual thinking, as counterfactuals should rapidly establish a representation of the world that is contradictory to the readers’ real-world or factual knowledge. Counterfactuals may create strong predictive biases that interact with real-world knowledge during processing.
To examine this, Ferguson et al. () presented participants with auditory scenarios, crossing real-world (as in (a)) and counterfactual-world (as in (b)) contexts. Note that the critical word could either be consistent with the preceding counterfactual context (but anomalous with respect to real-world knowledge) or inconsistent with the preceding counterfactual context (but consistent with real-world knowledge). Participants’ eye movements around a concurrent visual scene depicting, among other things, a fish and some carrots, were monitored and time-locked to the onset of the disambiguating word (‘fish’ or ‘carrots’). Under these conditions, listeners were able to inhibit expectations based on realworld knowledge and instead anticipated reference to the contextually appropriate counterfactual-world object (i.e. carrots) at least ms before the auditory onset of this critical word (see Figure .). Moreover, anticipation of the counterfactual world emerged within the same time-course as anticipation within a real-world context, which suggests that the counterfactual-world representation is rapidly and preferentially activated (at least when visual cues are available to support this), even when this requires inhibition of realworld knowledge. ()
a. If cats are hungry, they usually pester their owners until they get fed. Families could feed their cat a bowl of [fish/carrots] and they would gobble it down happily. b. If cats were vegetarians, they would be cheaper for owners to look after. Families could feed their cat a bowl of [carrots/fish] and they would gobble it down happily.
Complementary evidence for rapid access to the counterfactual world has been found by examining eye movements during reading. This paradigm provides insights on the ease 2
‘feed their cat a bowl of ’
‘carrots/fish’
‘and it would...’
1.5
Log (P(RW)/P(CW))
1
Real-world Object
0.5 0 –0.5 –1
Counterfactual-World Object
–1.5 –2 –1500 –1300 –1100 –900 –700 –500 –300 –100 100
300
500
700
900
1100 1300 1500 1700 1900
Time from Target Noun Onset(ms) RW context—CW input
RW context—RW input
CW context—CW input
CW context—RW input
.. The average probability of fixating either the real-world (RW) or counterfactual-world (CW) object as a function of each condition in Ferguson et al. (: Experiment ). Note that the solid black vertical line in the figure (t = ) indicates the target noun onset and dashed lines represent relevant word onsets.
with which words are integrated during reading. Rayner (, ) suggests that the amount of time that readers spend looking at a word accurately reflects the ease or difficulty associated with processing that word. This effect is clearly demonstrated by variations of word length and word frequency, as well as constraints of the preceding context to ambiguous words (e.g. bank as a financial institution vs. edge of a river; Inhoff & Rayner, ; Rayner & Duffy, ; Rayner & Well, ). More interestingly for studies of counterfactual thinking, there is a growing body of research showing that costs associated with relating sentence meaning to real-world probabilities can be observed in eye movement patterns during reading. Specifically, difficulties integrating a particular word into the wider context or known reality are reflected in longer reading times and higher incidence of regressive eye movements (e.g. Ni et al., ; Braze et al., ; Rayner et al., ). Ferguson & Sanford () compared the influence of constraints from the counterfactual and real world in a task where participants were eye-tracked while they read the counterfactual materials as shown in (a) and (b) above. Results showed that readers can accommodate a novel counterfactual world, leading to a reversal of typical real-world anomaly detection effects when that information was presented within an appropriate counterfactual context. This pattern was evidenced by longer total reading times on critical words (e.g. ‘fish’ or ‘carrots’) that were inconsistent with the described counterfactual world, increased regressions out from the words that directly followed this critical word (showing that readers regressed back in the text to make sense of the inconsistency), and longer ‘go past’ reading times on this post-critical region (showing that readers reread earlier portions of text to make sense of the difficulty). Similar patterns of integration difficulty were found when the counterfactual premise was established by negating the real world (e.g. ‘If cats were not carnivores . . . ’; Ferguson et al., : Experiment ). The third source of evidence for rapid access to the counterfactual world comes from studies that have employed neuroimaging methods. Most of these studies have employed an anomaly detection reading paradigm while recording ERPs, and aim to exploit the brain’s clear sensitivity to stimulus predictability and semantic integration processes during language comprehension, evidenced by the N ERP component (Kutas & Hillyard, a). This component is a centroparietally distributed, negative-going deflection in the ERP, which peaks approximately ms after word-onset. Extensive research in psycholinguistics and cognitive neuroscience has shown that the amplitude of this ERP component is directly influenced by inconsistencies of local and contextual information (e.g. Van Berkum et al., ; Hagoort et al., ), as well as the degree of plausibility for the overall message (DeLong et al., ). Interestingly, research using this technique has shown that successful language comprehension does not necessarily require information to map directly onto our real-world expectations. A growing collection of evidence suggests that a strong discourse context can overrule local lexical-semantic factors when these two conflict and as such can immediately influence comprehension processes. For example, typical N responses to local semantic anomalies (e.g. the peanut was in love) can be reversed when presented within an appropriate discourse context that makes the described events plausible (Nieuwland & Van Berkum, ; Filik & Leuthold, ). Similarly, N effects are activated when a narrative describes a character’s inappropriate emotional response to a given social situation (Leuthold et al., ), when a statement conflicts with a person’s moral values (Van Berkum et al., ), or when their actions mismatch their (false) beliefs (Ferguson et al., ).
When applied to counterfactuals, later work by Nieuwland has shown that counterfactual events are just as easy to integrate as factual events, provided that the propositional truth-value of the statement is high. In this research, Nieuwland & Martin () recorded ERPs while participants read negated counterfactual sentences (a), and real-world sentences (b) that described (alternative outcomes to) known historical events. Results showed that words that were false with respect to the overall context elicited larger N components compared to true words. Moreover, the fact that these N responses were indistinguishable between real-world and counterfactual contexts suggests that readers can integrate the consequences of non-real events in a counterfactual world with no delay, and without interference from the implied factual (or known real-world) version of events. Interestingly, similar results have been found using counterfactual sentences that depict novel fictional events (e.g. ‘If dogs had gills, Dobermans would breathe underwater’), indicating that the ability to make counterfactual inferences does not rely on pre-existing knowledge about reality (Nieuwland, ; see also Ferguson et al., ). Nieuwland () has also presented evidence that different neural circuits might underlie these truth-value judgements for real-world and counterfactual-world utterances. Specifically, research using fMRI showed that while real-world and counterfactual false sentences elicited comparable activity increases in the left inferior frontal gyrus, counterfactuals elicited a larger truthvalue difference in the right inferior frontal gyrus compared to real-world sentences. ()
a. If NASA had not developed its Apollo project, the first country to land on the moon would have been [Russia/America]. b. Because NASA developed its Apollo project, the first country to land on the moon was [America/Russia].
.. Evidence for rapid access to the implied factual world Complementing this research showing rapid access to the explicitly described counterfactual world is empirical evidence showing fast access to the inferred factual world. Indeed, early research showed that participants’ memory for counterfactuals tends to be for the presupposed facts (Fillenbaum, ), since readers were more likely to report having previously seen a negative declarative sentence (e.g. ‘He didn’t catch the plane’) following a causal counterfactual (‘If he had caught the plane he would have arrived on time’) than when it was entirely new. More recent research using self-paced reading and probe word tasks provides further evidence for this factual representation following a counterfactual narrative (De Vega et al., ; De Vega & Urrutia, ). Here, participants read short stories that began by introducing an initial context (e.g. ‘Juan was in the office sitting in front of his computer. He was typing a report that his boss had asked him for.’), followed by a factual (e.g. ‘As Juan had enough time, he went to the cafe to drink a beer’) or counterfactual continuation (‘If Juan had enough time, he would have gone to the cafe to drink a beer’). Participants then responded to a probe word that was either related to the initial events of the story (typing), or to the recently described event (drink). Responses were faster to probe words that related to the initial situation when they followed a counterfactual context compared to a factual context, suggesting that readers maintained access to the real-world events within
a counterfactual story but had suppressed access to this initial scenario within a factual story. In addition, responses to situation-final probe words did not differ between counterfactual and factual contexts, suggesting that the final events were equally accessible following both types of context. De Vega and colleagues interpret this pattern as evidence that counterfactuals activate a representation of the factual (initial) situation alongside a representation of the counterfactual (final) situation. When a delay was introduced between the context and probe (De Vega & Urrutia, ), the final situation became less accessible following a counterfactual context (i.e. slower responses within a counterfactual than factual context), meaning that the ‘as if ’ situation was only momentarily represented. These findings suggest that readers cancel discourse updating following a counterfactual; they make the inference that the described events did not occur and instead favour the implied factual version of events. Complementary effects were found in an ERP task (Urrutia et al., ), where continuation sentences that referred back to the initial situation elicited more negative ERP waves (indicating integration difficulty) when they followed a factual discourse than a counterfactual discourse, suggesting that the counterfactual’s presupposition—that the events did not occur—precludes updating the hereand-now of discourse. Further evidence is provided by Ferguson () using an anomaly detection reading paradigm. Participants were eye-tracked while they read short narratives in which a context sentence set up a realistic counterfactual world (e.g. ‘If Joanne had remembered her umbrella, she would have avoided the rain’), and a subsequent critical sentence described an event that was either consistent or inconsistent with the implied factual world (‘By the time she arrived at school, Joanne’s hair was [wet/dry]’). Results showed that readers detected the contextually inconsistent word in the earliest moments of reading, leading to an immediate increase in regressive eye movements, and increased total reading times on the inconsistent word. Thus, readers evaluated incoming text according to an inferred representation of the factual world, even when this inference was based on locally available discourse context that was not supported by real-world knowledge.
.. Both worlds are represented simultaneously Intriguingly, emerging experimental evidence supports the previously described theoretical models, in showing that counterfactuals activate representations of both the counterfactual and factual versions of the world, possibly even simultaneously. In fact, behavioural evidence consistently supports this dual mental representation view. For example, Santamaría et al. (; see also Gómez-Veiga et al., ) employed a priming paradigm and observed that counterfactual statements such as ‘If it had rained, the plants would have bloomed’ prime faster self-paced reading of both negative (i.e. factual; ‘It did not rain, and the plants did not bloom’) and affirmative conjunctions (i.e. counterfactual; ‘It rained, and the plants bloomed’). In contrast, factual sentences only primed the affirmative conjunction. In addition, the results described earlier from De Vega and colleagues also support the dual representation of counterfactuals (De Vega et al., ; De Vega & Urrutia, ; Urrutia et al., ), since the alternative ‘as if ’ meaning was momentarily activated alongside the presupposed factual meaning, though it was later suppressed in favour of the factual version of events.
Indeed, in many of the eye-tracking and ERP studies reported earlier there was evidence of early interference from the alternative factual/counterfactual world, despite later measures reflecting the overall sentence meaning. For example, the eye-tracked reading study by Ferguson & Sanford () found that following a counterfactual context (e.g. ‘If cats were vegetarians . . . ’), continuations that violated real-world assumptions (i.e. cats eating carrots) led to disruptions during the early stages of critical word integration (increased first-pass reading times). Interestingly, this interference from real-world knowledge was longer lasting (extending into total reading times and number of fixations on the critical word) when the consequence was introduced using the modal verb ‘would’ rather than ‘could’ (e.g. ‘Families would feed their cat . . . ’, Experiment ), since would suggests that an event is a repeated or habitual, which activates stronger associations to real-world expectancies. Similarly, the eye-tracking and ERP studies reported in Ferguson et al. () showed that readers’ initial interpretation of events following a negated counterfactual context (e.g. ‘If cats were not carnivores . . . ’) reflected a conflict between real-world expectations and this fictional context. That is, reading times on the critical word (firstpass fixation duration and total reading time, Experiment ) did not differ between counterfactual-consistent and -inconsistent continuations, and the N response was larger for critical words that violated real-world knowledge than those that violated the counterfactual premise (Experiment ). Finally, in Ferguson’s () eye-tracking reading study, initial reading times on the critical word (first-pass reading times and first fixation duration) were increased even when the continuation was consistent with the implied factual meaning of the counterfactual context (but inconsistent with respect to the described hypothetical event). Alongside the increased regressions for counterfactual inconsistent narratives, this demonstrates that readers experienced distinct anomaly detection responses to both versions of the world. Importantly, this shows that they have early access to both factual and counterfactual alternatives upon encountering the critical word, which leads to different processing and recovery strategies in each condition. Taken together, these eye-tracking and ERP studies converge to suggest that readers maintain access to both the counterfactual and factual interpretations of events, and when encountering the critical word they simultaneously evaluate incoming information against both versions of the world, just prior to accommodation into the wider counterfactual proposition. This final interpretation of events appears to be guided by cues from the evolving linguistic input, with modal verbs such as ‘could’ and ‘would’ signalling a continuation of the counterfactual world, and factual cues (e.g. ‘That evening . . . ’) indicating reference back to the factual version of events. This proposal was recently tested by Ferguson & Cane (), who examined the relative availability of counterfactual and factual representations within a counterfactual premise, by assessing how easily readers can switch between these models of the world during comprehension. To this end, ERPs were recorded while participants read short narratives in which a counterfactual context set up a realistic hypothetical situation, as in (a). In order to assess the relative availability of each of these representations of the world, the linguistic structure of a second critical sentence that described a consequence of the preceding context was manipulated so that it either implied reference to the factual world (as in (b)), or included a modal inflection that signalled a continuation of events according to the counterfactual world (as in (c)). Crucially, the sentence-final word in this second sentence was either consistent or inconsistent with the overall preceding context.
()
a. If David had been wearing his glasses, he would have been able to read the poster easily. b. From this distance, David found that the words were [blurry/clear]. c. From this distance, David would have found that the words were [clear/blurry].
Results (see Figure .) showed that when reference to the counterfactual world was maintained by the ongoing discourse, readers correctly interpreted events according to the counterfactual world (i.e. showed larger N for inconsistent than consistent words). However, when cues referred back to the factual world, readers showed no difference in N between consistent and inconsistent critical words, suggesting that they simultaneously compared information against both possible worlds, despite offline cloze and plausibility ratings showing an eventual preference for the factual world. Interestingly, an analysis of individual differences in working memory revealed that only individuals with high working memory capacity detected the inconsistent critical word following a counterfactual-counterfactual context. Individuals with low working memory capacity did not show any signs of using the counterfactual representation to distinguish between world-consistent and -inconsistent continuations (though they did show appropriate inconsistency effects within a factual context). This finding demonstrates that holding multiple world representations in mind is cognitively effortful, and suggests that counterfactual inferences may not be made at all when working memory capacity is low. Moreover, the consistency effect was not apparent in either high or low working memory groups following a counterfactual-factual context, suggesting that switching between counterfactual and factual worlds is even more effortful. In summary, this growing body of empirical research demonstrates the powerful effects that counterfactuals can have on discourse comprehension. The findings largely support the proposal that counterfactual contexts activate dual representations of the world (e.g. Byrne & Tasso, ; Byrne, , ; Johnson-Laird & Byrne, ), comprising both the explicit hypothetical model (p and q) and the implied factual model (not p and
Pz
Pz μV –1
–200
1
Pz μV –1
200
400
600 ms –200
1
μV –1 200
400
600 ms –200
1
2
2
2
3
3
3
4
4
4
Factual consistent Factual inconsistent
Counterfactual counterfactual consistent Counterfactual counterfactual inconsistent
200
400
600 ms
Counterfactual factual consistent Counterfactual factual inconsistent
.. Grand average ERPs elicited by critical words in the target sentence for each of the conditions, in Ferguson & Cane (). Note that negativity is plotted upwards.
not q). Moreover, linguistic cues can guide the reader in selecting which model of the world to evaluate incoming information against. When reference to the counterfactual world is maintained by the ongoing discourse (i.e. through the use of modal inflections, ‘could’ or ‘would’), readers can interpret events according to this counterfactual world in the earliest moments of processing, and within a comparable time-course to factual events. However, this counterfactual world is grounded in knowledge about reality, meaning that maintaining and updating this hypothetical model over time is cognitively effortful. In contrast, when cues refer back to the factual world, or are ambiguous, readers perform a simultaneous comparison against both possible worlds, which has the effect of delaying evaluation of the overall discourse propositional meaning.
. C
.................................................................................................................................. Recent research has begun to examine the link between counterfactual thinking and social cognition. This research is largely based on work by developmental researchers who proposed that counterfactuals engage the same network of specialized cognitive processes that underpin understanding of others’ mental states (termed Theory of Mind (ToM); Leslie, ; Riggs et al., ). Similar to counterfactual reasoning, ToM is a case of possibly valid reasoning, based on the beliefs of other people, that might be false according to our own knowledge of reality (see Rubio-Fernández, Chapter in this volume). Thus, tasks involving ToM require an understanding of events according to the intentions, beliefs, and desires of other people, while maintaining an understanding of our own intentions, beliefs, and desires. In line with models of counterfactual thinking, it is therefore plausible to assume that ToM tasks also engage a dual comprehension process, where the comprehender must represent information about his own reality alongside a representation of reality according to another person’s beliefs. Indeed, recent evidence has shown that the same set of executive functions (e.g. working memory, inhibitory control, and cognitive flexibility) mediates successful counterfactual and ToM reasoning (Drayton et al., ; Van Hoeck et al., ), and that ability in counterfactual reasoning is a necessary but not sufficient component of successful false belief task performance (Peterson & Bowler, ). To date, the majority of investigations into this relatively novel relationship have been limited to developmental psychology, and have largely employed offline response-based methods. Two recent studies have employed online methods to examine the relationship between counterfactual thinking and social cognition in adults. Ferguson et al. (; see also Ferguson et al., ) used a visual world paradigm to examine the time-course of expectations based on the beliefs of others, in comparison to expectations according to a counterfactual world (as described earlier in section ..). In the ToM task, participants listened to contexts that described a character having a true or false belief about the location of an object (as in (a)). Their eye movements were then recorded as they listened to a target sentence describing the character’s looking behaviour (b) that was paired with a visual scene depicting the possible locations.
()
a. John washed the dishes after his breakfast and left his watch on the table. [Later John noticed Victoria move/While John was distracted, Victoria moved] the watch from the table to the bed. b. Later, John wanted to find his watch so he looked on the [bed/table] and yawned.
Results (see Figure .) showed that, similarly to counterfactual thinking, comprehenders were rapidly able to modify expectations based on their own knowledge about reality, and direct their visual attention to appropriate referents according to that character’s perspective (from ms before the target location onset). However, the relative availability of reality/alternative information appears to differ between counterfactual and ToM tasks. The counterfactual-world interpretation influenced attention as soon as the visual scene became available, but predicting others’ actions based on their beliefs was reliant on further linguistic cues to support the perspective inference. In addition, participants’ gender influenced predictions about others’ actions based on their beliefs (delayed prediction in males), but not when predicting events in a counterfactual world. Further investigations are needed to gain a detailed understanding of the specific and overlapping encoding and processing mechanisms employed in these tasks. The link between counterfactual thought and social communication has been further explored by Kulakova & Nieuwland (in press) in an ERP study. Participants listened to sentences that contained either a counterfactual (e.g. ‘If sweets were made of sugar . . . ’) or hypothetical (e.g. ‘If sweets are made of sugar . . . ’) antecedent that conflicted with realworld knowledge (i.e. that sweets are made of sugar) to explore how counterfactual comprehension relates to social communication skills. Results showed that counterfactual antecedents elicited a larger N response at the critical word ‘sugar’ compared to 2.5
[Object]
‘so he’
‘looked on the’
[Location]
[Wrap-up]
2
Log (P(Reality)/ P(Belief))
1.5 1
Reality Location
0.5 0 –0.5 –1 –1.5
Belief Location
–2 –2.5 –1500 –1300 –1100 –900
–700
–500 –300 –100 100 300 500 Time from [Location] Onset (ms)
Reality context—Reality input Belief context—Reality input
700
900
1100
1300
1500
Reality context—Belief input Belief context—Belief input
.. The average probability of fixating either the Reality or Belief location as a function of each condition in Ferguson et al. (: Experiment ). Note that the solid black vertical line in the figure (t = ) indicates the target noun onset and dashed lines represent relevant word onsets.
hypothetical antecedents, which reflects participants’ reduced expectancy for this factually consistent word within a counterfactual frame. More interestingly, the size of this N effect was modulated by individual differences in social communication skill, with higher pragmatic abilities leading to a larger expectancy violation effect on the N. This demonstrates that social skills play an important part in establishing counterfactual worlds and linking them with real-world knowledge online. Given this emerging evidence of a link between counterfactual thinking and ToM, it is remarkable that only a handful of published studies have empirically tested counterfactual thinking in individuals with Autism Spectrum Disorder (ASD). These studies have focused primarily on the production of counterfactual utterances, or on syllogistic reasoning based on counterfactual statements (e.g. ‘All cats bark. Rex is a cat—Does Rex bark?’). This limited work has revealed distinct counterfactual reasoning strategies between ASD children (who prefer additive statements, e.g. ‘If only I had . . . ’) and Typically Developing (TD) children (who prefer subtractive statements, e.g. ‘If only I had not . . . ’; Begeer et al., ), and has shown that while ASD children show impaired counterfactual reasoning relative to TD children, this impairment is not as great as that found on similar tasks that involve ToM (Grant et al., ). One potential source for these different counterfactual processing strategies comes from the need to distinguish reality from fiction; this is an important ability that develops early in TD children (e.g. Morison & Gardner, ; Skolnick & Bloom, ). It is particularly crucial to understanding counterfactuals, since one must accept events in a fictional world, yet maintain existing knowledge about the real world. Research suggests that children with ASD struggle with this fiction/reality distinction; they are less likely to identify which is ‘silly’ between statements like ‘I live in England’ and ‘I live on the moon’ (Surian et al., ), and include fewer imaginative elements in made-up stories compared to TD children (Scott & Baron-Cohen, ). In addition, adults with ASD show a significant preference for reading non-fiction stories over fiction stories (Barnes, ). An intriguing finding in the literature on ASD and counterfactual thinking is that being prompted to ‘imagine’ described events impairs ASD children’s counterfactual thinking, whereas TD children improve, suggesting that individuals with ASD adopt a response bias based on logic when they are uncertain about the communicator’s intentions (Leevers & Harris, ). These findings suggest that individuals with ASD may be less able to suspend reality than TD individuals, particularly when this requires vivid imagination of an alternative fictional world. Currently no cognitive model exists to interpret these patterns, or how they link with the broader social and linguistic difficulties seen in individuals with ASD. Future research exploring this link using online empirical research methods would provide valuable insights into counterfactual thinking, and inform more specific cognitive theories about dual representations, imagination, and language comprehension.
. S
.................................................................................................................................. In summary, the ability to update our knowledge using contextual information is a vital process during everyday language comprehension. Counterfactual language (e.g. If . . . then . . . ) establishes a special kind of context which requires a person to represent false information
as temporarily true (and vice versa). While a long tradition of research has explored counterfactual reasoning strategies, and the kinds of counterfactual thoughts people are likely to generate in a variety of circumstances, it is only very recently that researchers have begun to empirically test how counterfactuals are represented and accessed online during language comprehension. This emerging research has employed temporally sensitive cognitive neuroscientific methodologies alongside language comprehension tasks to demonstrate that healthy adult readers can make appropriate inferences following a counterfactual context, showing rapid, and possibly simultaneous, access to both the counterfactual and factual interpretations of events. Future research is needed to explore the link between counterfactual thinking and social cognition, including how counterfactual thinking is affected when social skills are impaired (as in ASD).
A This work was supported by a Leverhulme Trust research grant (RPG--) to HJF.
......................................................................................................................
......................................................................................................................
. I
.................................................................................................................................. D refers to the ability of sentences with plural subjects to give rise to an interpretation in which the property expressed in the sentence is predicated of the individuals of which the group or plurality is comprised, as opposed to predicating the property of the plurality itself (a collective interpretation). To illustrate, the sentence in () could be interpreted collectively (where the two men worked together to eat one pizza) or distributively (where each man ate his own pizza). ()
The two men ate a pizza.
While this sentence is ambiguous, there are other sentences that seem only to give rise to either collective or distributive readings. For example, () can only have a collective interpretation, where the group has the property of meeting, not a distributive one where each man has the property of meeting, while () gives rise to a distributive reading, rather than a collective one, where each man has the property of having dark hair. ()
The two men met in the alley by the warehouse.
()
The two men have dark hair.
The difference between the unambiguous () and () on the one hand, and the ambiguous () on the other, has been taken as evidence that certain predicates are specified to be collective, distributive, or mixed (allowing for the other interpretation) (Link, , ; Dowty, ; Lasersohn, ). Other inherently collective predicates are gather and scatter, while be a (wo)man, have brown eyes, be big, be round, and so forth, are among the inherently distributive predicates. One proposal for accounting for the distributive reading of () is that a covert distributivity operator D is inserted into the representation, resulting in the predicate being applied to the atomic individuals that make up the plurality the two men (see e.g. Link , ; Lasersohn , ; Schwarzschild ; Brisson ). The same would apply to a
conjoined subject, such as Frank and Dean. Gillon (), Schwarzschild (), and Verkuyl & Van der Does (), among others, have also proposed the existence of a cover, which partitions the plurality into subsets in a given discourse context (see discussion in Brisson, ). Lasersohn () has also proposed that a * operator applies to the predicate P to generate all of the individual sums of members, and *P is closed under sum formation.
. L-
.................................................................................................................................. While some predicates in sentences can be interpreted distributively, some quantifiers are known to lexically-encode distributivity. For example, in English, every and each are inherently distributive—the latter more so than the former—while all is not (Vendler, ; Beghelli & Stowell, ; Tunstall, ; Brisson, ). For example, () is typically interpreted with a distributive interpretation, whereby each of the men in the plurality has the property of eating a calzone. ()
{Every|Each} man ate a calzone.
This judgement may arise in part from a preference by the sentence processor to assign surface scope (every > a) to scopally ambiguous sentences (cf. Kurtzman & MacDonald, ), but scope is not the determining factor in a distributive interpretation. The difference between these inherently distributive quantifiers on the one hand, and plurals on the other, is also illustrated in (). The sentence in (a) requires that each individual man went to the racetrack, while (b) is weaker, and allows for non-maximality (Landman, a,b, ; Brisson, ). That is, not every man in the plurality needs to have gone. In sentences such as (), the quantifier every or each can be referred to as the distributive key, and the predicate as the distributive share. (Some languages have distributive-share quantifiers, but English does not.) ()
a. {Every|Each} man went to the racetrack. b. The men went to the racetrack.
The difference between the inherently distributive quantifiers every and each and the universal quantifier all is illustrated in (). Quantificational phrases involving every or each cannot compose with a collective predicate such as gather or scatter, while those with all can (as long as all has first composed with a definite plural phrase). () a. *{Every|Each} man gathered in a dimly lit alleyway. b. All the men gathered in a dimly lit alleyway. The contrast between every and each on the one hand and all on the other is attested in numerous other languages, including Georgian, Tagalog, Turkish, and others (Gil, ). That each is inherently more distributive than every can be illustrated with the set of examples from Brasoveanu & Dotlačil (b) that follows in (), which highlights what Tunstall () has proposed as the ‘differentiation condition’ of each: that sentences with
each require the event structures to be totally distributive—that is, each of the atomic individuals is associated with its own sub-event, and these sub-events can be properly differentiated. For example, in (a), it must be the case that each student participates in a separate, spatiotemporally distinct photographing event. The sub-events might also be differentiated with different event agents and/or resultant states (see Tunstall, ; Landman, ). () a. Jake photographed every student in the class, but not separately. b. #Jake photographed each student in the class, but not separately. If there is a covert distributivity operator D, one would expect that some languages would lexicalize this operator overtly. Indeed, it has been proposed that each is in fact such an overt distributivity operator (Link, ; Schwarzschild, ). By way of illustration, () forces the distributive interpretation of (). It must be the case that each of the men eats his own pizza; they cannot share one pizza. If the example were different, such that a pizza was not consumed, but was instead moved from one table to another, the pizza could be the same pizza, but there would have to be separate moving events, with each of the two men as an agent of his own event. ()
The two men each ate a pizza.
By contrast, () brings out the collective interpretation of ()—although it does not force it in the same way that () forces the distributive interpretation, since it is possible that each man could eat his own pizza while they sit together in a booth in an Italian restaurant.1 ()
The two men ate a pizza together.
Not only quantifiers lexically encode distributivity. Some predicates obligatorily apply to the atomic individuals composing a plurality. Schwarzschild () has called such predicates ‘stubbornly distributive’. Distributive predicates include those such as be round/ square, have brown/blue eyes, be tall/big, and be intelligent. When composed with a plurality in subject position, these predicates must modify the individuals in the group, rather than the group itself (as collective predicates like gather or scatter would). Thus, distributive predicates cannot compose with mass nouns, such as bread or water, since there are no individuals to be predicated of, but they can compose with object mass nouns, such as furniture or jewelry. However, if the mass noun appears in a phrase like drops of water or loaves of bread, the distributive predicate can then apply. This pattern is illustrated in (). ()
1
a. The pillows are {round/big}. b. #The water is {round/big}. c. The drops of water are {round/big}. d. The furniture is {round/big}.
For discussion of together, see Parsons (); Hoeksema (); Dowty (); Lasersohn (, ); Schwarzschild (, ); Verkuyl (); Brisson (); among others.
Some individual lexical items, such as heavy, lend themselves easily to either distributive or collective interpretations. For example, () could mean that each of the individual boxes is heavy, or that all together, the boxes are heavy (but each of them is relatively light). In this case, the context would serve to disambiguate. ()
Those boxes are heavy.
In the sections that follow, I present the findings from a number of experimental studies conducted over the last few decades demonstrating how children and adults interpret both ambiguous and unambiguously distributive sentences in off-line experimental tasks investigating preference and/or acceptability, and how adults respond to overt cues to distributivity in tasks designed to collect real-time behavioural data. The collective findings from these studies indicate that children as young as and adults access multiple readings (collective, distributive, cumulative) of ambiguous sentences. However, while adults take each as an unambiguous cue to distributivity, children do not appear to be sensitive to this lexical property of this quantifier until well after years of age. Moreover, when adults do take the presence of each as a signal to access a distributive interpretation, this has a detectable effect in reaction time. While the vast majority of studies have been conducted in English, the growing interest in experimental work on distributivity has led to recent crosslinguistic explorations, which will no doubt continue in subsequent years, resulting in a clearer picture of the theoretical and developmental landscape of distributivity.
. D
.................................................................................................................................. In this section, I focus on the overt distributivity marker each, and what children and adults know about the meaning of this lexical item. While each, every, and all in English are all universal quantifiers, they differ in the degree to which they require exhaustivity and distributivity in the predication of the plurality to which they apply. However, though a contrast rooted in distributivity between each and every on the one hand, and each and all on the other may be clear for adults, children as old as to years of age fail to appreciate this distinction. Thus, it appears that the difference among these quantifiers is a lexically-encoded contrast that must be learned over the course of development, and this appears to take a fair amount of time. This observation follows from a number of studies over the years.2 2
Since the focus of this chapter is on distributivity, I will not review other literature on children’s other non-adult-like responses to questions with universal quantifiers, such as the so-called ‘quantifier spreading’ cases. For discussion of these findings and related references, see, for example, Brooks & Sekerina (/ and Brooks & Parshina, Chapter in this volume); Roeper et al. (); and Syrett (c). Although this phenomenon may ultimately be related in some way to the patterns of performance summarized there, the connection is not transparent enough to make a connection here, given the range of work reviewed in this chapter.
To begin, a contrast between adults and children with each was observed by Ferenz & Prasada () in an act-out task in which they gave children and adults instructions to put all/each (of the) objects in a certain location. When presented with the instructions with all, adults placed all of the relevant objects in the location roughly at once, but when presented with the instructions with each, they typically placed the objects in the location one at a time (as Vendler, , and Tunstall, , might have predicted they would). Children, however, did not interpret each as a cue to a ‘distributive’ response, and did not place the objects in the location one by one. One might argue that since Ferenz & Prasada’s experiment used an act-out task that children’s failure to provide a ‘distributive’ response with each in this task is not sufficient evidence to demonstrate they lack understanding—that is, the argument might be that their performance underestimates their competence. This would not be surprising. However, further evidence that children lack an understanding of the strong distributivity property of each comes from their answers to wh-questions involving the universal quantifiers every and each. Achimova et al. () presented child and adult participants with scenarios followed by a question involving the interaction between a which phrase and a universal quantifier (either every or each). The wh-phrase and the quantifier either occupied subject or object position, as shown in (). () a. Which game did {every/each} friend play? b. Which friend received {every/each} toy? When the quantifier was in subject position (as in (a)), the prediction arising from the theoretical literature was that either a single answer or a Pair-List Answer (PLA) would be possible, with either universal quantifier. When the quantifier was in object position (as in (b)), the prediction was that a PLA should be barred with every (for theoretical reasons discussed therein), but permitted with each, given each’s strong distributivity. Interestingly, consistent with previous research findings, children produced more PLAs than adults with every in the subject position (although PLAs were licensed here), but also with every in object position (where PLAs were not licensed). Adults conservatively produced single answers, even when either answer type was licensed. Interestingly, though, children under-produced PLAs relative to adults both when each was in subject position and when it was in object position. That is, adults took the presence of each to indicate strong distributivity, producing PLAs to questions such as the above as a result (in contrast to their consistent single answers with every), while children seemed oblivious to this lexical property of this quantifier. Brooks & Braine (), Brooks et al. (), and Brooks & Sekerina (/) also found that children diverged from adults in their treatment of each but patterned along with adults in their treatment of all. The divergence with each appears to be rooted in both exhaustivity and distributivity. Brooks and her colleagues presented child and adult participants with sentences following the template in () (), accompanied by black and white drawings in which some combination of Ns (e.g. people, flowers) were performing an action or were in some configuration with respect to some number of Ns (e.g. cakes, boats, vases). In one version of the task, participants were given a forced choice between two different images and asked to choose the best sentence to match the context
presented in the image, while in the other version, they were asked whether the sentence could accurately describe the individual picture. ()
{All of the Ns are/Each N is} Ving an N.
()
There is an N Ving {all/each} of the Ns.
Adults had no difficulty making correct decisions about sentences with all and each, given the various contexts. Children patterned with adults with the all sentences and scenarios. For example, when presented with (), they preferred a scene where all of the Ns were engaged in a Ving event with an N, and when presented with (), they preferred a scene in which each N was being acted on by an N. However, with the each sentences, children showed no such preference. In the complementary acceptability judgement task, while adults did not accept the each version of () applied to a scene where there were some Ns doing nothing, children accepted such sentences. Finally, given an image where three Ns are acting on one N, and an image where each of three Ns is each acting on a separate N, adults accepted () in the latter, but not the former, context, while children accepted () a significant amount of time in both contexts. Thus, these results show that children do not reliably appreciate the role of each as a strong distributivity marker until well after age –—a surprisingly late age in language development. In a similar vein as Brooks and colleagues, Musolino () was interested in assessing children’s and adults’ judgements of sentences with each, but his findings were set against the backdrop of sentences with numerical phrases. Sentences with plurals or numerical phrases in both subject and object position, such as (), are multi-ways ambiguous, taking into account the presence of the D operator, quantification over events, scopal interaction between the subject and object phrases, and the possibility of a cumulative reading (Scha, ; Aone, ; Link, ; Schwertel, ). For example, here, the five men could have each purchased three calzones, all five could have chipped in together to purchase the three, some combination of the subsets of the five could have purchased some combination of the subsets of the three, or all together, the five could have purchased the three with it still being possible that one man purchased one calzone in the group. ()
Five men purchased three calzones.
Musolino wanted to determine whether children (and adult controls) could access multiple interpretations of ambiguous test sentences that had numerical phrases in both subject and object position, as in (), probing the availability of distributive, collective, and cumulative readings. ()
Three boys are holding two balloons.
In an acceptability judgement task, both age groups accepted sentences such as this one when there were three boys, each of whom was holding two balloons (a distributive reading), and when there were two balloons, the group of which was being held by three boys (a collective reading). Neither age group accepted the sentence when there were two balloons, each held by three boys (the inverse scope reading where the object takes wide
scope over the subject). Only adults accepted the sentence in a context favouring the cumulative reading, where there was a total of three boys and a total of two balloons, each of which was being held by at least one boy. However, when () was changed to the minimally different sentence () with each, both age groups accepted the new version of the sentence in the object-wide-scope context, and where there were two balloons, the group of which was being held by three boys. Thus, the presence of this overt distributivity marker made available an inverse scope reading that was not accessible before. ()
Three boys are holding each balloon.
This finding thus reveals that children allow each to take wide scope in a linguistic environment in which they do not permit the same for an object indefinite. However, children may be too permissive in this regard, since they—unlike adults—accepted () in the first context, where there were three boys, each of whom was holding two balloons. Thus, both children and adults can assign multiple readings to sentences that give rise to distributive, collective, and cumulative interpretations, and constrain these interpretations such that not just any interpretation is licensed. At the same time, children fail to constrain the range of interpretations allowed by sentences with each in the same way that adults do. However, because the range of predicates in this experiment was rather minimal, and since the control items in this experiment did not involve extraneous boys or balloons and nonholding events, this experiment was not able to fully probe children’s understanding of the range of permissible readings for the target sentences, or the distributivity property of each. Syrett & Musolino () followed up on these results with a wider range of predicates, presenting child and adult participants with prerecorded videos of actors performing distributive and collective events, described by ambiguous sentences, such as those in (), and their unambiguous counterparts, which were created by inserting one or the other of the lexical items in parentheses. () a. b. c. d.
Two boys (each) pushed a car (together). Two boys (each) built a tower (together). Two girls (each) pushed a car (together). Two girls (each) drew a circle (together).
As was predicted, children and adults accepted the unambiguous sentences in both the distributive and collective contexts. When participants were given these sentences and still images of the distributive and collective events in progress in a forced-choice preference task and asked to make a choice between the two scenes, adults vastly preferred the collective context, while children had a slight but non-significant preference for the distributive context. But while adults reliably took together as a cue to collectivity, and each as a cue to distributivity, children—as might have been predicted, based on previous findings—appeared to be oblivious to this fact, accepting these lexically modified sentences in both contexts. However, in the forced-choice preference task follow-up, children showed that they preferred the collective context over the distributive one for sentences with together, showing that they, like adults, appreciate the collectivizing function of this modifier. We will return to this finding momentarily.
The fact that children diverge notably from adults in their interpretations of sentences with each, failing to fully appreciate the strong distributivity of this universal quantifier and accessing interpretations that are not licensed by sentences in which each appears, led Pagliarini et al. () to predict that, in sentences without each, children should also diverge from adults in accessing a distributive interpretation more often than a collective one, similar to the trend Syrett & Musolino () noticed in their forced-choice preference task. Their reasoning was as follows. Take a sentence such as (), which allows for a distributive reading (where each boy lifts his own two boxes), a ‘dependent’ distributive reading (where each boy independently lifts the same two boxes), and a collective reading (where the boys work together to lift the two boxes, either simultaneously or sequentially). ()
The boys lifted two boxes.
When presented with such a sentence, adults resolve the ambiguity in favour of a collective interpretation, considering the quantificational alternative that the speaker could have uttered featuring each (Each boy lifted two boxes). Since children do not appreciate the role of this more informative, disambiguating quantificational alternative, they do not readily conclude that a speaker who utters () most likely favours the collective reading, and therefore access the distributive reading, in contrast to adults. Pagliarini et al. () therefore presented participants with sentences with each in subject position and sentences without each in subject position (and instead featuring a plural definite description), accompanied by a still image portraying either a distributive or collective context, as noted in () (). (Their experiment was conducted in Italian, not English.) ()
Each girl is building a sandcastle.
(accompanied by picture of distributive context)
()
Each boy is building a snowman.
(accompanied by picture of collective context)
()
The girls are building a snowman. (accompanied by picture of distributive context)
()
The boys are building a tower.
(accompanied by picture of collective context)
As expected, adults accepted () and rejected () in the respective context, and accepted () in the collective context. They were split on acceptance of () in the distributive context. By contrast, children beginning at age accepted all sentences in the given contexts. That is, consistent with the findings reported above, children incorrectly accepted () in the collective context, dropping below chance level acceptance only after age . And consistent with the pragmatic ‘alternatives’ hypothesis, they accepted () in the distributive context. Although we have focused on the fact that the accumulated findings across tasks indicate that children are insensitive to the distributive property of each until well into late childhood, there is something to be said about their interpretations of sentences with together. While they accept such sentences in both distributive and collective contexts, Syrett & Musolino () showed that they prefer the collective contexts when given the choice. Syrett & Musolino () took this result as a starting point to probe the reason for children’s acceptance of sentences with together in distributive contexts.
In a task similar to Syrett & Musolino (), Syrett & Musolino () presented participants with sentences such as those in () not only in collective contexts, but also in distributive contexts in which the two individual events occurred sequentially. Their hypothesis, inspired by Lasersohn (, ), was that children accepted the sentences with together in Syrett & Musolino () because the events were spatiotemporally contiguous sub-events of a larger event, and therefore licensed a non-collective reading supported by together. ()
a. Two boys pushed a car together. b. Two girls read a book together.
Indeed, in contrast to the previous findings, children rejected the sentences with together when the events were no longer time-locked—and even more so when children engaged with a puppet. While children do not appear to recognize early in development that each is an overt cue to an unambiguous distributive interpretation of sentences, they do appear to recognize that certain predicates are ‘stubbornly distributive’ and obligatorily apply to individuals. Syrett (d) presented children age to , and adult controls, with a binary forced-choice task in which they were introduced to different kinds of objects that would be conceptually and linguistically familiar to very young children (e.g. buttons, cups, fruit, blocks, pillows), along with properties that were also familiar. The individual properties of these objects were highlighted, as were the properties of a group (such as shape or size). Participants were then given a choice between two images on the screen: one where the group had the target property (e.g. being round) but the individuals possessed another (e.g. being square-shaped), and another where the properties were reversed. A set of brief instructions as in () accompanied these images, asking participants to point to the side of the screen as indicated. ()
a. Point to the round {furniture/buttons}. b. Point to the big {fruit/cups}.
Syrett (d) found that every age group was significantly more likely than chance to select the individuals with the target property upon hearing this prompt. Even by age , children’s selection percentage closely resembled that of adults. The one difference between children and adults came from the predicate be tall, where children in the youngest age group consistently selected the group with the property, and then this pattern dropped off to chance or a little over chance level at years of age.3 Importantly, in a follow-up condition, when the instructions included the collectivizer group, as in (), participants were significantly less likely to choose the side of the screen where the individuals had the target property, and were pulled toward the side of the screen where the group or plurality had the target property instead. ()
Point to the round group of {furniture/buttons}.
3
A similar finding for tall was also reported by Scontras & Goodman ().
These findings thus demonstrate that young children are not completely oblivious to the feature of distributivity. While they may not be sensitive to it as it is encoded in each, they are sensitive to it in certain predicates, which obligatorily apply at the level of the individual, rather than to the plurality of which these individuals are members. Thus, children appreciate that pluralities have a decomposable structure with individuable parts, and that distributive predicates modify these parts, regardless of whether the group is referred to with a plural count noun.
. R-
.................................................................................................................................. Earlier, we reviewed a range of findings from off-line act-out, acceptability judgement, and preference tasks revealing how children and adults respond to ambiguous sentences with potentially distributive interpretations and unambiguously distributive sentences featuring each. To complement these findings, this section reviews the findings from on-line tasks revealing how adults interpret sentences with each, compared to key baseline sentences. Results from these experimental tasks with adults, which measure eye gaze, self-paced reading, and reaction times in response to certain prompts, provide behavioural evidence that each serves as a reliable cue to distributivity (and the presence of a D operator) in real time, that distributive interpretations prompted by each have an immediate effect on how subsequent singular indefinites are processed, and that in specifying distributivity, each requires that the relevant property apply to atomic individuals composing the plurality such that the relevant events are differentiated from each other.4 Frazier et al. () monitored participants’ eye gaze while reading sentences with mixed predicates in order to determine whether positing a D operator for distributive interpretations signaled by each slowed down the sentence processor. Target sentences in their experiment had the structure of those in (). The slashes below, which were not presented to participants, represent regions of interest. () a. Edna and Milton / mailed one care-package each / after hearing about / the severe droughts in Africa. b. Edna and Milton / mailed one care-package together / after hearing about / the severe droughts in Africa. c. Edna and Milton / each mailed one care-package / after hearing about / the severe droughts in Africa. d. Edna and Milton / together mailed one care-package / after hearing about / the severe droughts in Africa. The (a) and (c) versions resulted in distributive interpretations, whereas (b) and (d) favoured a collective interpretation. Crucially, the (a) and (b) versions were locally See section 11.3 of the chapter by Nouwen et al. (Chapter in this volume) for evidence for ‘variation effects’ that arise when modified numerals are embedded under the distributive quantifier each. 4
ambiguous, as the lexical item each or together did not appear until after the predicate, while the (c) and (d) versions were locally unambiguous, since these lexical items appeared immediately after the conjoined subject and before the predicate. Frazier et al. () first normed their experimental sentences in advance in a naturalness rating task to ensure that none had an inherently collective or distributive bias. A separate group of participants participated in the eye-gaze experiment. Participants saw one version of these sentences, and sixteen test sentences in total, along with a number of fillers. Two values were measured: first-pass reading time (the sum of fixations in a given region between the time of first entering and first leaving the region), and total reading time (the sum of all fixations in a given region, including the first-pass time and subsequent rereading of the material in the region). Frazier et al. () found an interaction in both the first-pass reading time and in the total reading time, such that distributive predicates were read more slowly than collective predicates in ambiguous sentences ((a) vs (b)). There were also more regressions out of the third region (and into the second region) in the distributive versions than in the collective versions, when the predicates were locally ambiguous. Frazier et al. () took their findings as support for their hypothesis that the processor is slowed down with the locally ambiguous sentences when postulating the presence of a distributivity operator triggered by the presence of each. These findings might also be seen as consistent with Pagliarini et al. ()’s proposal that the default for adults is a collective interpretation, for pragmatic reasons, and that this is the reading they access until explicitly cued to access a distributive one. Patson & Warren () gathered evidence from reaction times that there is a processing cost when a singular indefinite phrase (e.g. a box) is interpreted as conceptually plural, when participants are given a cue by each to interpret the sentence distributively. Following a methodology used by Berent et al. (), they presented participants with a self-paced reading task, and then had them make rapid judgements of whether they saw one or two words in a key region. They predicted that singular indefinite phrases in a distributively interpreted sentence would be interpreted as conceptually plural, and therefore, that when asked about the number of words they saw on the screen, participants would be slower to render ‘one-word’ judgements for indefinites in these environments than with singular indefinites in collectively interpreted sentences. Target sentences were of the form in (), with either each or together in subject position, and a singular or plural direct object (the latter meant to control for differences in reading time in the spillover region). () a. b. c. d.
Each of the men carried a box. Together the men carried a box. Each of the men carried some boxes. Together the men carried some boxes.
Participants saw thirty-six target sentences, in a self-paced reading paradigm. The first part of the sentence was in black, while the last word in the indefinite phrase (in bold above) was in blue. Participants pressed the space bar to click through chunks of the sentence, and were told that upon seeing the blue text, they should decide whether there were one or two words on the screen and press ‘’ or ‘’ on the keyboard accordingly. (Filler items had two words displayed, so there was an equal number of ‘’ and ‘’ answers.)
Patson & Warren () found that decisions about the number of words in the blue region (the sentence-final indefinite phrase) took longer when the decision was ‘’ and the phrase was grammatically plural-marked, and importantly, that there was an interaction between noun marking and distributive/collective interpretation such that with singularmarked indefinites, a ‘’ decision took longer for distributive interpretations than collective interpretations. There was no reliable effect of distributivity or plural-marking otherwise. The fundamental distinction between each and every proposed by theoreticians discussed earlier has been further probed experimentally in both off-line and on-line sentence processing tasks. (See extensive discussion in Brasoveanu & Dotlačil, Chapter in this volume.) Following seminal work on the interpretation and processing of quantificational phrases by Tunstall (), Brasoveanu & Dotlačil (b) designed a binary forced-choice task and a related self-paced reading task to test the prediction from Tunstall’s differentiation hypothesis that participants should prefer the inverse scope of the quantifier each over the indefinite subject in () more than with every, and further, that the presence of a resultative, as in (), should neutralize this preference, since it indicates the resultant state of the event (‘be blue’), which reflects a distributive property of an individual.5 (See the earlier discussion of stubbornly distributive predicates.) Thus, of all four sentences, (b) should show the highest rate of inverse scope readings. (The area in parentheses, which was identical among the sentences, only appeared in the self-paced reading task, as a spillover region for analysis.) () a. A helper dyed every shirt (without thinking about it). b. A helper dyed each shirt (without thinking about it). () a. A helper dyed every shirt blue (without thinking about it). b. A helper dyed each shirt blue (without thinking about it). In the binary forced-choice task, participants were given a choice between two interpretations, as in (). There were eighteen test items in both experiments, distributed into four lists that represented four different item conditions, with every experimental item appearing exactly once in each list. These test items were interspersed with a number of fillers. ()
A helper dyed every shirt blue. a. Each shirt was dyed by a possibly different helper (until it was the color blue). b. All the shirts were dyed by the same helper (until they were the color blue).
Brasoveanu & Dotlačil (b) found that inverse scope readings were highest with each, and without a resultative present. In the self-paced reading task, Brasoveanu & Dotlačil (b) found a significant slowdown in reading times for the each sentences starting at the third word (about above), but they found a significant interaction of quantifier and presence/absence of a resultative, such that the presence of the resultative led to a decrease in reading time for each, but not for every. This finding was to be 5
For extensive discussion of the encoding of event structure in lexical semantics and verb frames, see Chen and Husband (Chapter in this volume).
expected: since the resultative satisfied the differentiation condition, each needed to do less work.6 Complementary sentence-processing work by Brasoveanu & Dotlačil (a) also revealed a processing cost for each in sentences with same, arguing that the reason is that the event differentiation requirements of each are at odds with the semantics of sentence-internal same.
. C
.................................................................................................................................. The purpose of this chapter was to review the findings from a range of experimental studies on the topic of distributivity. The findings reported in this chapter reveal the following. First, although children as young as years of age recognize, along with adults, that some predicates apply distributively to atomic individuals in a plurality, they, unlike adults, fail to appreciate the lexically-encoded distributivity of the universal quantifier each. In task after task, children accept sentences with each in non-distributive contexts and fail to produce a distributive response when given the opportunity to do so. Second, the presence of each has real-time consequences for adults processing sentences with distributive interpretations. Since the vast majority of these studies, with very few exceptions, have focused on English and the quantifiers every, each, and all (and the modifier together), it seems fitting that recent research, motivated by theoretically-grounded hypotheses, has begun to extend these methodologies to languages beyond English. These investigations include languages in which particles or other discourse markers, rather than (or in addition to) universal quantifiers, signal distributivity, although in different ways, and the variety of ways in which the conditions required for distributivity can be satisfied. For example, po in Serbian and Russian has been argued to be a strong distributive share (or pluractional) marker, distinct from the Serbian term for every, svaki, and the Russian term for every, kazhdyi (Knežević ). The particle dou in Mandarin Chinese has been claimed to be similar to English each in its status as a distributivity marker, although it also has a meaning similar to English even, and functions as a propositional operator, taking a proposition as its argument, which is unlike a universal quantifier (Lee, ; Lin, ; Liu, ). Probing these and other languages further will allow us to get a much more fine-grained understanding of how distributivity is captured at the lexical level, and identify the universal and language-specific aspects of how distributivity is acquired and processed.
A This writing of this chapter was made possible by startup funding from Rutgers, The State University of New Jersey—New Brunswick.
6 Brasoveanu & Dotlačil (b) also reported interesting differences among participants in their preference for scopal relations—hence the title of their paper. However, as those findings are not directly connected to the goals of this chapter, I leave them aside here.
......................................................................................................................
......................................................................................................................
-
. I
.................................................................................................................................. ‘B fly’, ‘The fox is a sly animal’, and ‘A cat lands on its feet’ are all generic generalizations that allow speakers to talk about classes, types, or kinds of individuals—in this case, about the kinds bird (Aves), fox (Vulpes vulpes), and cat (Felis catus). They are used to convey characteristic properties of kinds, and/or to express knowledge about the world including beliefs, stereotypes, and prejudices (e.g. ‘Women talk more than men’). Generics can be seen as one of the building blocks of human cognition, as they allow us to conceptualize of properties of kinds and to organize our experience of the world, while they are ubiquitous in everyday communication and feature in run-of-the-mill conversation. Generic generalizations have fascinating properties, especially when compared to other types of generalizations such as overtly quantificational generalizations. Quantificational generalizations are expressed in quantitative terms. Statements like ‘all cats eat mice’, ‘some lions live in cages’, and ‘most tigers have yellow eyes’ refer to the quantity that satisfies the relevant property. In semantic theory, these generalizations can be relatively easily modelled in terms of set-inclusion relations (Barwise & Cooper, ). For the processing of quantifiers, see Brasoveanu & Dotlačil (Chapter in this volume). On the other hand, generic generalizations do not seem easily reducible to these terms, but seem to reflect richer and more complex relations between the kind and the property. Generics often refer to characteristic or essential properties of a kind; while ‘most books are paperbacks’ is true, the generic ‘books are paperbacks’ is not because publication format is not an essential property of books. Furthermore, unlike a comparable generalization (‘all birds fly’), generics (‘birds fly’) tolerate exceptions (e.g. penguins), a characteristic that has proved difficult to account for. Given their complex yet fundamental nature, it is not surprising that generics have attracted the interest of linguists and philosophers of language since the s, while they have also recently become the focus of concentrated interest by cognitive and developmental psychologists. In this chapter, I will review the phenomenon of genericity and I will argue that our understanding is enriched by research that integrates the tools and perspectives of theoretical and experimental approaches. In section ., I first introduce some basic facts about genericity, then I discuss the main characteristics of generics and finally I illustrate the
range of possible statements one could make when using generics. Section . presents how genericity has been treated within three different disciplines (linguistics, psychology, and philosophy) and discusses at length two types of approach, the formal semantics, and the psychological Generics-as-Default approach. In section ., I present an overview of the experimental research on the topic in two different disciplines, cognitive psychology and linguistics, focusing on adult data. Bilingual and second language adult acquisition as well as child language acquisition are also briefly discussed. Section . discusses whether and how the experimental investigation has advanced our understanding of the phenomenon of genericity and considers other experimental methods that could be used in future research while pointing out some under-researched areas. In section ., I present some conclusions.
. W ?
..................................................................................................................................
.. Main definitions and facts Traditionally, two types of phenomena have been classified as generic (Krifka et al., ). The first one involves kind-referring Noun Phrases (NPs), as in (), where the subject NP the cat does not state something about some particular cat, but rather about the kind cat (Felis catus) itself. In this instance the cat does not refer to an ordinary individual or object, but instead is kind-referring (contrast this to ‘the cat is sleeping on the mat’ that refers to an ordinary object, a cat, where the NP here is object-referring). The second one involves propositions, called characterizing or habitual sentences, which describe a general property or regularity that does not refer to a specific or isolated fact, but rather summarizes groups of particular episodes or facts. () below does not describe an isolated fact, but a habit, some kind of characteristic event or behaviour (that refers to what Neko usually does after hunting mice). These two phenomena can co-occur, as in (). Here, the subject NP can be analysed as a kind-referring NP and the sentence itself as expressing a generalization that holds for the kind cat. Thus, () expresses a generalization about the kind as a whole, as well as a regular event. A kind-referring NP like the cat in () has been called a D-generic (Krifka, ), as these are commonly expressed with the definite singular in English, whereas a sentence that combines both, as in (), has been called an I-generic, because these are commonly expressed with indefinite NPs like bare plurals. D-generics also include generalizations made with kind-level predicates (Carlson, ), such as be extinct, die out, be common, which predicate a property directly of the kind in question. For example, in uttering () one states something about the kind dodo (Raphus cucullatus), namely that this kind of thing has ceased to exist. ()
The cat was first domesticated in Cyprus.
()
Neko takes a nap after hunting mice.
()
Cats are fed with fresh fish or cat food.
()
The dodo is extinct.
-
It is also important to note two linguistic facts about the expression of genericity: a) crosslinguistically, no language has a unique, unambiguous marker of genericity equivalent to a quantifier or determiner (Dahl, : ); and, arguably as a consequence, b) within a language, we may find different kinds of NPs in a generic statement (see Krifka et al., : , where they state that ‘characterising sentences may contain virtually any NP’). Among them, the most common types of NPs that appear in characterizing sentences are bare plurals, indefinite singulars, and definite singulars, as we see below:1 ()
Cats have whiskers.
bare plural
()
A cat has whiskers.
indefinite singular
()
The cat has whiskers.
definite singular
Despite the interesting fact that different languages employ different grammatical devices to express genericity (Longobardi, ; Chierchia, ), the cross-linguistic aspect of genericity has rarely been tackled systematically (see however Dahl, ; Behrens, , ; Dayal, ).
.. Main characteristics of generics The main characteristics of generics to be discussed in this section are the following: a) temporal unboundedness; b) law-like or nomic character; c) association with dispositions; d) resistance to contextual restriction; and e) exception tolerance.
... Temporal unboundedness As Mari et al. (: ) state, a property of characterising sentences which has often been regarded as essential to their generic meaning is temporal unboundedness, namely the property by which such a sentence is not true relative to a time interval with definite bounds, but rather relative to an indefinitely large interval, or even in a timeless way.
Thus, it has been argued that generics cannot be felicitously modified by adverbs that denote particular temporal locations such as today, as in (), unlike sentences that talk about a singular episode, as in (), where the time interval according to which the sentence is true is bound by the relevant time adverb, in this case today. The temporal unboundedness of generics means that they are taken to be timeless statements akin to eternal truths. Crucially, someone evaluating a generic statement like () does not need to have any information about the context of utterance, as the truth or falsity of the generic statement does not rest upon any specific state of affairs, but on general knowledge about cats and their characteristics.
1
The differences between them are still a matter of controversy, but see Krifka et al. (), Greenberg (), Mari et al. () among others for insightful discussion.
()
?Cats meow today.
()
Cats are meowing today.
()
Cats meow.
Even though all theories seem to recognize that unboundedness is a main characteristic of generics, it is not a well-defined property. For more discussion, see Declerck (); Krifka et al. (); and Mari et al. ().
... Law-like or nomic character Another property of generics, which seems to be linked to their unboundedness, is their law-like or nomic character. Generics express regular patterns of occurrence of certain kinds of events, rather than singular events (Dahl, ). As a consequence, they are not true with respect to a definite time interval, but they are eternal truths with a law-like nature as seen in (). Nomic statements also concern possible, non-actual situations or cases, in contrast to accidental generalizations that concern only actual cases. Under this perspective, a sentence like () above is a nomic generalization about cats that concerns cases in which a law or rule applies, restricting us to ‘all normal cases’, predicating the property of meowing of all normal cats both actual and future. This could be paraphrased as ‘if something is a cat, it meows’. This characteristic is related to exception tolerance, which will be discussed in section .... Similarly, a sentence like () is a nomic generalization that concerns sugar under normal circumstances or situations. ()
Sugar dissolves in water.
... Association with dispositions The nomic character of generics is related to another characteristic property of theirs, namely the fact that they express dispositions and abilities, which do not depend on particular circumstances (Dahl, ; Krifka et al., ). A property might only show under certain conditions, which may never be actualized. So, we do not need any corroborating past instances of the property being actualized to be able to judge that () and () are true. () might be true even if the machine never has and never will have crushed a single orange and () can be paraphrased as in (). ()
This machine crushes oranges.
()
A Rarámuri runs miles in one session.
()
A Rarámuri can run miles in one session.
... Resistance to contextual restriction The last two characteristics have been used to highlight the differences between generics and universals. Standardly, Krifka (: ) argued that unlike sentences with explicit quantifiers like every, generics cannot be contextually restricted. Thus, in a discourse like
-
the following, the nominal argument of every (lion in ()) can be interpreted as ‘every lion in this cage’, whereas lions cannot be interpreted as ‘lions in this cage’. Contextual restriction is proposed as the mechanism by which a quantified, but not a generic, statement can be interpreted with respect to specific individuals explicitly mentioned in the previous discourse. ()
(Context: There are lions and tigers in this cage.) a. Every lion is dangerous. b. Lions are dangerous.
... Exception tolerance Finally, the most distinctive characteristic of generics is arguably the fact that they tolerate exceptions (Krifka et al., ). () can be truthfully uttered in the face of exceptions, such as flightless birds like penguins, ostriches, emus, or birds with clipped wings, whereas the universally quantified generalization in () is false given the existence of exceptions such as the above. ()
Birds fly.
()
Every bird flies.
Pelletier (: ) posed the following question recently: ‘How many exceptions can a generic statement tolerate and still be true?’ Using a ‘squish’ of examples as Pelletier () does in order to illustrate the issue, we see that we can speculate about the percentage of exceptions ranging from per cent, as in (), to a few abnormal cases, as in (), to around per cent, as in ()(), and even higher, as in (), and even to per cent, as in (), where the property is truthfully predicated of a tiny proportion of mosquitoes. ()
Snakes are reptiles.
()
Tigers have stripes.
()
Telephone books are thick.
()
Guppies give live birth.
()
Lions have manes.
()
Italians are good skiers.
()
Mosquitoes carry the West Nile virus.
Here lies the difficulty when one tries to translate generics as quantified statements in an attempt to give a quantificational analysis of generics. Some generics might be better paraphrased with ‘all’, others with ‘most’, and others with ‘some’. This challenge was identified by Carlson (: ), who used the term ‘fluctuating truth conditions’ to refer to this property of generics. As we will see in the next section, even though there is variation in the percentage of exceptions generics allow, this variation is not random, but it depends on the type of property predicated of the kind.
.. Genericity: A mixed bag or a single phenomenon? As becomes obvious from the earlier discussion, the question of what type of properties generics can express is not trivial. For a statement to qualify as a generic generalization, the property must bear a certain relationship to the kind. Generics can express generalizations about any type of kind, be it a natural kind, as in (), an artefact kind as in (), or a social kind as in (). The majority of generics make claims about characteristic or essential properties, that is, properties that tend to have high prevalence for the kind in question, as well as being typically associated with it. These properties can be true of all the members of the kind, as in (), of the vast majority of the members of the kind, as in (), or of a minority of the members of the kind, as in (). Moreover, as Leslie () first discussed in detail, generics can furthermore express non-characteristic properties of kinds (that is, properties that are not prevalent), when they make claims about a property that is noteworthy, dangerous, or striking even though it has low prevalence, as in (). Lastly, there exist generalizations that resemble generic generalizations but concern accidental properties of kinds. These generalizations might be highly prevalent and seem (at least statistically) true, as in (), while others are highly prevalent, but seem false, as in (): ()
Squirrels eat nuts.
natural kind
()
Needles are sharp.
artefact kind
()
Artists are creative.
social kind
()
Foxes are animals.
()
Robins fly.
()
Deer have antlers.
()
Sharks attack people.
()
Cars have radios.
()
Books are paperbacks.
All these facts highlight the difficulty one faces when one tries to provide a unified account for all different types of generics. In the linguistics literature, different kinds of statements have been used to exemplify the wide range of statements one might express with generics, but it is in the psychology literature that we find the first attempt to classify different types of generics by giving them distinct names and discussing the conditions that would make them true or false in more detail. I have showcased the wide range of statements that can be made via generic generalizations and I will classify them following recent advances in the experimental literature in section ....
. A
.................................................................................................................................. I turn now to discuss how generics have been treated in linguistics (in formal semantics) and psychology (cognitive and developmental work), and I will conclude with a brief presentation of related work in philosophy.
-
In formal semantics, generic generalizations have been studied since the s (see Lawler, , ; Dahl, ; Carlson, ) and genericity is frequently viewed as a species of quantification. Nevertheless, how to characterize their semantic interpretation and how to model their truth conditions remains controversial (see recent discussion in Carlson, ; Mari et al., ). Generics have also become of key interest to cognitive psychologists. In contrast to the linguistic approach, the main psychological proposals treat generics as categorically different from quantifiers (see Leslie, , ; Gelman, ). These proposals posit a generic bias and argue that generics are a cognitive default, because they have priority both in terms of ontogeny and in terms of cognitive complexity.
.. Formal semantics ... Carlson’s monadic generic operator The first semantic theories of generics attributed generic meaning to a verb phrase operator that took an ordinary verbal predicate as an argument and yielded a characterizing predicate. Carlson () treated bare plurals as names for kinds and took the logical form of generic sentences to be that of a subject-predicate form, whereby there was a ‘generic operator’ Gn that had the effect of mapping episodic predicates (in his analysis, ‘stage-level predicates’) to their habitual counterparts. So, the generic statement below would have the following logical form: ()
Cats have whiskers.
()
Gn(have whiskers)(cats)
Heim (a) and Farkas & Sugioka () argued against Carlson’s monadic operator Gn and proposed that generics should receive an analysis based on a tripartite structure.2 Their proposal formed the basis of the received view of generics discussed in the next section.
... The modal approach, aka the received view: Krifka et al. () Krifka et al.’s () version of the modal approach has become the received view of generics. This view treats generic sentences as modalized conditional statements that involve a universal quantifier. This proposal was an answer to the following puzzle: even though generics seem similar to universals, they are both more restrictive and less restrictive than universals. Generics are more restrictive, because they are law-like. This is why mere accidental generalizations like ‘books are paperbacks’, even though statistically true, do not qualify as true generics. But generics are also less restrictive than universals, given that they allow for exceptions. Modal accounts assume a phonologically null Q-adverbial quantifier ‘GEN’ that is an unselective variable binding operator similar to adverbs of quantification like ‘usually’, 2 A major motivation for the tripartite structure implicit in quantification is that it readily accommodates intuitions of ambiguity, such as those associated with sentences like ‘typhoons arise in this part of the Pacific’, which can be interpreted as either ‘typhoons in general have a common origin in this part of the Pacific’, or as ‘there arise typhoons in this part of the Pacific’ (see Krifka et al., , for discussion).
‘typically’, and ‘always’, as analysed in Lewis (). This operator is sentential and is represented by a tripartite structure (Krifka et al. : ). Thus, a sentence like () is represented as in () below: ()
GEN [restrictor]
[matrix]
()
GENx [cats (x)]
[whiskers(x)]
Krifka et al. () propose an intensional analysis of GEN, according to which a sentence with an indefinite singular is interpreted as a conditional sentence with the if-clause providing the restriction for GEN. GEN is interpreted as an intensional unselective universal quantifier meaning ‘must’. On the assumption that indefinites contribute a free variable ranging over individuals (cf. Heim, a), this variable can be bound by the available universal quantifier as well. As Mari et al. (: ) illustrate, Krifka et al. () follow a classical modal framework, in which W is a set of worlds, D is a domain of entities, and ≤ an ordering source on worlds according to normality. Thus, a generic sentence like () is represented as follows: () a. A dog barks. b. If something is a dog, it barks. c. 8w’ ≤w, x [dog(x,w’)] [barks(x,w’)] Paraphrase: in all worlds, which are ‘normal’, if something is a dog in those worlds, then it barks in those worlds. As seen earlier, in the modal approach to generics a ‘normality’ condition is often implied. Thus, generics are argued to be true if it is normal for the members of the kind to have the property in question.
... The probabilistic approach: Cohen A probabilistic approach has been defended by Cohen (, ), according to which it is probability rather than modality that forms the basis of the semantics of generics, at least for ‘absolute generics’ such as ravens are black. This account holds that As are B is true just in case the probability of an arbitrary A being a B is greater than .. Cohen (: ) introduces a homogeneity condition, according to which ‘the generic gen(ψ,φ) presupposes that its domain, ψ, is homogeneous, in the following sense: for any psychologically salient criterion by which ψ may be partitioned into subsets, the conditional probability of φ ought to be roughly the same given every such subset of ψ.’ Salient partitions are, for example, space, numerical scales, gender, subject matter, and abstract domains. Illustrating Cohen’s () account, which proposes that there is a covert generic quantifier GEN, the statement cats have whiskers would receive the following representation: ()
Cats have whiskers. GEN (cat(x), have.whiskers(x)) P(have.whiskers | cat) > . (the probability of an object having whiskers given that the object is a cat is greater than .)
-
... Recent proposals questioning the quantificational view Carlson (: ) summarizes the received view for generics in the linguistics literature by arguing that ‘while the details of various analyses that have employed it [GEN] may be called into question, that there is some kind of operator akin to GEN in generics is a reasonably secure claim at this point’. The exact semantics of GEN are nevertheless highly disputed—the following proposals are reviewed in Krifka et al. (: ff.): a) relevant quantification (Declerck, ); b) prototypes (Nunberg & Pan, ); c) stereotypes (Geurts, ); d) modal interpretations (Krifka et al., ); e) situations (Ter Meulen, ); and f) non-monotonic inferences (Asher and Morreau, ). Within formal semantics, the view that generics are not quantificational is either considered to have been surpassed by the modal account or has not yet been shaped into a fully spelled-out account. Recent discussion in Mari et al. () and further references therein, though, point out that an account based on noun ambiguity à la Dayal () could provide an account for genericity without assuming a generic operator. They do not provide the details of such an account, however, thus such a proposal still remains to be evaluated. Furthermore, Deo & Madiman () offer a Neo-Carlsonian account based on gradability and probability theory (cumulative distribution functions and stochastic comparison) and analyse generics as predicating a (non-)gradable property directly of a kind.
.. Cognitive psychology As we have seen earlier, even though within formal semantics there seems to be a received view, the meaning of generics is still a matter of controversy. In psychology, in recent years, a different kind of approach to genericity emerged. The main view to which most psychologists ascribe to is the Generics-as-Default (GaD) view, which treats generics as an innate and default mechanism.
... The Generics-as-Default view In contrast to the quantificational analysis of generics, a growing body of experimental cognitive and developmental psychological work on the topic proposes that genericity is categorically different from (and significantly simpler than) quantification (Leslie, ; Gelman, ). The GaD view postulates that generics are an innate and default mode of thinking. This idea is linked to the view of cognition that assumes two different systems, argued for by Kahneman & Frederick () among others, which includes a distinction between System , a fast, automatic, effortless lower-level system, and System , a slower, more effortful higher-level rule-governed system. According to this view, the reason there is no overt ‘GEN’ in any language is that generics do not, after all, involve any kind of overt quantifier or operator. In the psychological GaD approach, because generic generalizations are understood to be a basic, pre-linguistic mode of thinking, some of the specific challenges for the quantificational analysis are avoided. For instance, according to this view, there is no overt generic operator in any known language because generics are the unmarked, System , case. On this view, only effortful, non-default quantificational generalizations require overt linguistic
exponence. Leslie (, ) put forward an account that offered a treatment to what have been called ‘troublesome generics’, like the following: ()
Mosquitoes carry the West Nile virus.
()
Books are paperbacks.
()
Birds lay eggs.
These sentences pose a problem to the quantificational account of generics, because () is true despite the fact that only a tiny proportion of mosquitoes have that property, () is false in spite of the vast majority of books having that property and () is true even though the egg-laying property is held by less than half of birds (only fertile adult female birds might do so). Paraphrasing the above sentences would make them look similar to ‘some’ (), ‘all’ () or ‘most’ (), making their truth conditions look ‘quirky’. What Leslie proposed was that in order to understand the meaning of generic sentences we should think of the kind of generalizations they express. Their quirky truth conditions are not because of the semantics of generics (or GEN), but because generics are more cognitively primitive than quantificational generalizations. The quirkiness of generics is thus a consequence of cognition, not of generic sentences, which just ‘give voice to our most primitive, default generalizations’ (Leslie, : ). The default mechanism of generalization employed by the cognitive system is further influenced by the ‘strikingness’ of a property along some characteristic dimension. Leslie (: ) proposes the following circumstances under which Ks are Fs is true: ()
. . . circumstances under which a generic of the form Ks are F is true are as follows: The counterinstances are negative, and: If F lies along a characteristic dimension for the Ks, then some Ks are F, unless K is an artifact or social kind, in which case F is the function or purpose of the kind K; If F is striking, then some Ks are F and the others are disposed to be F; Otherwise, almost all Ks are F.
Initial motivation for the GaD view was mainly conceptual and relied on the following observations: a) the proposed ‘unmarked’ surface form of generics in the sense of Chomsky (); b) evidence from congenitally deaf children who employ gestures that can be understood as generic without any exposure to spoken or sign language (Goldin-Meadow et al., ); c) the Pirahã language, which lacks universal quantifiers like ‘all’ yet features generics (Everett, ); and d) results from reasoning studies (Jönsson & Hampton, ), which could be interpreted as errors with universal statements being treated as generics. The GaD view argues that generics are a cognitive default and thus makes the following predictions: generics should have priority both in terms of ontogeny (children are expected to understand and produce generics before quantified statements) and in terms of cognitive complexity (quantified statements are misunderstood more often than generics in experimental tasks). The above predictions have been tested in a growing literature to be discussed in section .. Building on the earlier proposal, Leslie and colleagues have focused on the adult processing of generics trying to find evidence that corroborates such a view, while Gelman and colleagues have sought to find evidence from a developmental perspective.
-
.. Philosophy The discussion of whether quantificational approaches to generics are correct has had repercussions in recent work in philosophy too, which we cannot discuss in detail because of space limitations. Within philosophy, the view that generics are not quantificational has been defended in various places: a) generics are simple subject/predicate sentences that predicate properties of kinds (Liebesman, ); b) genericity has a psychological, rather than a linguistic, basis (Collins, ); and c) a sophisticated kind-predicate view à la Carlson () needs to be revisited (Teichman, ). Sorensen () has also argued that understanding generics can elucidate long-standing philosophical issues such as the sorites paradox. Sterken () has emphasized the context-sensitive character of generics and has argued that GEN is not quantificational, but an indexical, composed of two free variables: one representing quantificational force and one representing lexical restriction. Nickel () integrates semantics with metaphysics and stresses the relevance of an explanatory mechanism. Finally, there is interesting work related to how generics concerning social kinds like ‘women are submissive’ or ‘Muslims are terrorists’ are applied in stereotyping and prejudice (see, e.g., Haslanger, ; Saul, ; Leslie, forthcoming). The interested reader is referred to Nickel (forthcoming) for a recent overview of genericity through the perspective of philosophy of language.
.. Summary and some considerations Summarizing, we can divide theoretical accounts of genericity into two broad categories: the ones that treat generics as quantificational and those that do not. We see from the above overview that each kind of approach makes some predictions about how generics are interpreted or learned. Each kind of approach faces its own challenges; for instance, on the one hand for the formal semantics account the fact that GEN is never pronounced calls into question its explanatory power, especially when one thinks of how a learner would come to postulate such a covert element, while on the other hand the psychological GaD view could be seen as exchanging one problem for another given that generics do not receive any special treatment but are given for free. In this latter respect, though, as soon as one looks at generics cross-linguistically, it becomes apparent that the acquisition of genericity needs to be studied alongside not only quantification, but also definiteness and specificity, as they all seem to play an important role. As we will see in the next section, some issues have already been tested in experimental settings, while others remain for future research.
. E
.................................................................................................................................. The recent surge in interest to provide new accounts of genericity has helped identify theoryneutral and theory-critical issues. The spate of theories that postulate modal operators in
possible-world semantics has posited several different relevant factors. The apparent theoretical impasse might be partly due to the lack of reliable data, since usually the issues have been addressed through researchers’ reflective intuition, which prohibits consensus about even the most fundamental facts about genericity. The fluctuating truth conditions I discussed earlier make generics particularly suited for experimental investigation. As Krifka et al. (: ) noted much of our knowledge of the world, and many of our beliefs about the world, are couched in terms of characterizing sentences. Such sentences, we take it, are either true or false—they are not ‘indeterminate’ or ‘figurative’ or ‘metaphorical’ or ‘sloppy talk’. After all, we certainly would want to count the classic Snow is white as literally having a truth value!
A convincing way to answer this is to start collecting Truth-Value Judgements (TVJ). From this point of view, it is in fact surprising that the experimental investigation of generics did not start earlier.
.. Experimental psychology ... Off-line methods: Adult judgement data The main bulk of experiments that has been conducted on generics falls under the umbrella of the GaD view or is in general cognitively-driven. With respect to the methods used in the experimental investigation of genericity we observe the following variety of tasks: a) TVJ and agreement task, either binary or with a Likert scale (Khemlani et al., ; Cimpian, Brandone, & Gelman, ; Leslie et al., ; Meyer et al., ; Khemlani, Leslie, & Glucksberg, ; Prasada et al., ); b) prevalence estimation task (Cimpian, Brandone, & Gelman, , Khemlani, Leslie, & Glucksberg, ); c) confidence rating task (Khemlani et al., ); d) paraphrase task (Leslie et al., ); e) cue validity task (Khemlani, Leslie, & Glucksberg, ); f) memory/recall task (Leslie & Gelman, ); g) sentence-picture matching task (Cimpian, Gelman, & Brandone, ); h) naturalness evaluation task (Prasada et al., ); and i) default inference task (Khemlani, Leslie, & Glucksberg, ). These experiments have mainly studied genericity by contrasting it with overt quantifiers (universal such as ‘all’, proportional such as ‘most’ and existential such as ‘some’). The main question they seek to answer is whether generics are a default mechanism and whether there exists a generic bias when it comes to generalizations. In the remainder of the section, I will review some representative studies that use the methods mentioned earlier. Leslie and colleagues. As we discussed in section ., determining which properties can be generically predicated has proven very challenging. The degree of exceptionality seems to be one of the defining parameters of the relevant categories. Based on Leslie (, ), Leslie and colleagues (Khemlani et al., ; Khemlani et al., ; Leslie et al., ) started collecting judgement data on generics. Leslie et al. (: , table ) offer the following more elaborate classification, where they define the following five distinct sub-types:
-
a) quasi-definitional: property must be universally true of all the members of the kind with no exceptions, for example ‘triangles have three sides’ b) majority characteristic: property must be central, principled or essential (Gelman, ; Medin & Ortony, ) and directly related to the nature of the kind. It must be highly prevalent, while allowing some exceptions (e.g. albino tigers), for example ‘tigers have stripes’ c) minority characteristic: property must be central, principled, or essential, but only be held by a minority of the kind. Restricted to methods of gestation, methods of nourishing the very young, and characteristic physical traits exhibited only by one gender, for example ‘lions have manes’ d) striking: property must only be exhibited by a small minority of the kind, and must signify something dangerous which is to be avoided, for example ‘pit bulls maul children’ They further distinguish these types of generic generalizations from majority generalizations and false generalizations that superficially share the form of generics, but are either accidental or false, like the following: a) majority: property must be prevalent among members of the kind, but must not be a principled connection (Prasada & Dillingham, , ), for example ‘cars have radios’ b) false generalization: property must be prevalent among members of the kind and there must be a sufficiently salient alternative property (e.g. being left-handed), so that the generic form of the predication sounds false or mistaken, for example ‘Canadians are right-handed’ Illustrating with data from Leslie et al. ()’s experiments and , these different types of generics yield different acceptance rates, thus lending some support to the above classification. The acceptance rates in their Experiment vary as follows: quasi-definitional (%), majority characteristic (%), minority characteristic (%), majority (%), striking (%), and false generalizations (%). Thus, people judge different types of generalization as true most of the time, unless they express something that is a false generalization, which should be read as ‘false as a generic generalization’. In their Experiment a, Leslie et al. provided participants with information about population estimates presented before the TVJ. Compared to Experiment , in Experiment a, the acceptance rates were almost the same for quasi-definitional, majority characteristic and minority characteristic statements ( %), but they varied at a greater degree for striking (dropped from % to %) and for false generalizations (dropped from % to %) and to an even greater degree for majority statements (dropped from % to %). In their Experiment b, an additional second phase followed the TVJ task, where participants had to provide paraphrases of each statement they had judged in the first phase. Even though participants first evaluated the statements as in Experiment , we do see some differences. Acceptance rates were almost the same for quasi-definitional, majority characteristic, minority characteristic, and striking statements, but they varied at a slightly greater degree for false generalizations (dropped from % to %) and to a greater degree for majority (dropped from % to %). Taken as a whole, these judgements concerning generics are not surprising; nevertheless, they are important since they are the first judgement data found on the topic and indicate
that the task participants are asked to perform, or the context against which a generic is evaluated might have an effect at least for some types of generics. Collecting acceptance rates is only one way to make use of the above classification, which can also prove useful in order to identify whether specific phenomena are relevant for only some types of generalizations. It can further help to address generalizations that predicate a property of only a minority of the members of a kind, such as minority characteristic and striking generics, which pose a challenge to the quantificational accounts of genericity. The generic overgeneralization effect. One of the outcomes of the Leslie and colleagues’ studies was that they argue there is evidence for a ‘generic bias’, an effect they called the ‘Generic OverGeneralisation’ (GOG) effect. I will focus on Leslie et al. () to illustrate it. Similar results have been reported in other experiments that used TVJ tasks (Khemlani et al., ; Khemlani et al., ; Khemlani, Leslie, & Glucksberg ; and Meyer et al., ) or recall tasks (Leslie & Gelman, ). Leslie et al. (: ) use GOG to refer to ‘the tendency to overgeneralise from the truth of a generic to the truth of the corresponding universal statement’. As we discussed earlier, in their Experiment , participants judged the truth or falsity of a list of generic as well as ‘all’-quantified statements that were presented one after the other without any background context. In more than half of the trials when the ‘all’-statements involved characteristic properties, participants judged these statements to be true: per cent for majority characteristic such as ‘all tigers have stripes’ and per cent for minority characteristic statements such as ‘all ducks lay eggs’. By contrast, ‘all’-statements, which did not involve characteristic properties, such as ‘all cars have radios’, were only judged true per cent of the time. The authors argue that these high acceptance rates for the characteristic-property ‘all’-statements are due to participants interpreting the ‘false’ universally quantified statements as if they were their ‘true’ generic counterparts, and are thus a clear case of GOG. As the authors acknowledge, however, these elevated acceptance rates might be due to alternative explanations, which they sought to address in subsequent experiments. They considered three possible explanations for why their participants were so prone to accept statements like ‘all horses have four legs’ as true: a) ignorance of the relevant facts; b) a subkind (taxonomic) interpretation of ‘all’; and c) quantifier domain restriction in the sense, for example, of Stanley & Szabó (). The authors discarded these alternative explanations on the basis of additional experiments they ran and argued that they had support for a strong generic bias, according to which people sometimes treat universally quantified statements as if they were generic. It is important to note that the GOG data concern how universal quantifiers behave in a specific context, not how generics behave. Given that this view takes generics as a given, a default which requires no further explanation, the GOG effect affects generics only indirectly in that quantifiers are sometimes interpreted as if they were generics.3 Prasada and colleagues. Prasada & Dillingham (, ) propose that our conceptual system distinguishes at least two types of connections between kinds and their properties, namely principled and statistical connections. Principled connections concern properties
3
For a closer look at the GOG effect and alternative explanations for the pattern of judgement data that have been claimed to support it, see Lazaridou-Chatzigoga, Stockall, & Katsos ().
-
(k-properties) that instances of a kind have by virtue of their being the kinds of things they are (e.g. we think that dogs, by virtue of being dogs, have four legs). Principled connections license formal explanations and include a normative as well as a statistical dimension. Statistical connections on the other hand concern properties (t-properties) that are merely prevalent; that is, properties that only bear a factual connection to the kind in question (e.g. being red for a barn). Statistical connections do not support formal explanations or give rise to normative expectations. By devising a series of experiments that asked participants to judge whether certain paraphrases were appropriate for statements that involved either a principled or a statistical connection, they were able to provide empirical support for the distinction drawn. To illustrate, participants judged that principled connections like ‘dogs are four-legged’ could be paraphrased as either ‘dogs, by virtue of being the kinds of things they are, are four-legged’ or ‘dogs, in general, are four-legged’, whereas statistical connections like ‘barns are red’ could only be paraphrased as ‘barns, in general, are red’ and not as ‘barns, by virtue of being the kinds of things they are, are red’. Prasada et al. () elaborate on these conceptual distinctions. They used a TVJ task with a seven-point Likert scale in experiments and and a naturalness evaluation task in Experiment to investigate the conceptual foundations of generics as well as the role of prevalence, cue validity, and normalcy in licensing generics, which have been proposed as licensors in formal semantic approaches to generics. Cue validity is seen as ‘the probability that an item is part of a category given a particular cue’. In their Experiment , looking only at bare plural generics, these are the ratings they obtained (on a scale from to +): quasi definitional (.), majority characteristic (.), minority characteristic (.), majority (.), striking (.), majority false generalization like ‘books are paperbacks’ (.) and minority false generalization like ‘cars are yellow’ (.). These rates are similar to the ones obtained by Leslie et al. (). Prasada et al. () further focused on striking and minority characteristic generics that pose a challenge to quantificational approaches. They argue to have found support for the following: a) striking generics involve a causal connection between a kind and a property (and not a principled or a statistical connection); b) minority characteristic generics exhibit the characteristics of principled connections; and c) prevalence is dissociated from the acceptability of generics. They further argue that their data prove problematic for normalcy accounts and for the idea that cue validity is the licensor of low prevalence generics. The normalcy account, they argue, works only for majority characteristic generics like ‘tigers have stripes’. They conclude by pursuing a conceptually based approach to the semantics of generics building on Leslie (, ).4 Cimpian and colleagues. Cimpian and colleagues investigated the relation between genericity and prevalence, which put to the test intuitions that have been pervasive in the generics literature at least since Carlson (). Cimpian, Brandone, & Gelman () investigated an asymmetry in generic meaning, namely the fact that generic sentences have strong implications but require little evidence to be judged true. In order to do so, they devised four experiments with adult participants,
4
For a study focusing on minority characteristic statements and their gender-specificity, which is not discussed here because of space limitations, see Passanini & Hampton ().
who had to judge sentences involving novel kinds such as ‘lorches have purple feathers’. Participants judged these sentences as referring to nearly all lorches, whereas they also judged them as true given low prevalence levels (e.g. even when only % or % of lorches had the relevant property). They also investigated whether generic sentences about dangerous (‘% of lorches have dangerous purple feathers. These feathers are as sharp as needles and can easily get lodged in you, causing massive bleeding’) or distinctive properties (‘% of lorches have distinctive purple feathers. No other animals on this island have wide, smooth feathers like these’) would be more acceptable than generic sentences without similar connotations, concerning plain items (‘% of lorches have purple feathers’), and they found this to be the case for the sentences in question. Their results revealed that the proportion of ‘true’ responses was higher for the dangerous (%) and the distinctive (%) items than for the plain items (%), thus showing that both danger and distinctiveness have independent effects on generics’ truth conditions. The danger and distinctiveness information had an effect on generics’ acceptability mostly at the lower prevalence levels, particularly at the per cent level. Cimpian, Gelman, & Brandone () investigated the hypothesis that generics are not only about prevalence or frequency, but that their acceptance is influenced by naïve theories speakers have. They report three experiments in which they presented adult participants with novel categories with key features that concerned physical features characteristic of a biological kind. The participants preferred to map generic sentences (e.g. ‘dontrets have long tails’) onto novel categories for which the key feature (e.g. long tails) was absent in all the young but present in all the adults rather than onto novel categories for which the key feature was at least as prevalent but present in some of the young and in some of the adults. Furthermore, they were able to show that this mapping is specific to generics and does not hold for quantifiers like some or most. In sum, Cimpian and colleagues provide empirical proof for the suggested asymmetry in generic meaning: the fact that generic sentences may be judged as true at low prevalence levels yet at the same time may be associated with implications of high prevalence. Furthermore, they show that generics are not only about prevalence/frequency, but are also influenced by how one conceptualizes about different categories and how one uses causal and/or explanatory knowledge when thinking about the relation of properties and kinds.
... Online methods: Real-time processing and processing in the brain Within the GaD approach, generics are claimed to be part of System of Kahneman’s cognition system, which is fast and effortless, while quantified statements are part of System , which is slower and requires more cognitive effort. Therefore, generics are predicted to be faster to process than quantified statements. One experiment that manipulates task demands in the processing of generics is Meyer et al. (). The rationale behind their experiment was that, if generics are default, it should be easier to judge that a property is characteristic of a generic (i.e. dogs) than of a quantified set (i.e. all dogs), and they expected this generic advantage to be particularly evident when participants were under time pressure. Meyer et al. () recorded TVJs and judgement times in response to majority characteristic generic vs. ‘all’ quantified statements like ‘{all} dogs
-
have four legs’ and definitional statements like ‘{all} giraffes have long necks’, and varied whether participants were told to answer as quickly as possible, or to take as much time as they wanted. The main results were that a) participants in the speeded condition were more accurate at making a TVJ (% for generics vs. % for quantified statements) and faster to respond to generics (~ milliseconds (ms)) than to universals (~,ms); and b) for definitional statements, participants were faster to respond to generics than universals in both the speeded (~ms vs. ~,ms) and the unspeeded (~,ms vs. ~,ms) condition. Both results seem to be consistent with the GaD hypothesis: participants were, indeed, more likely to say true to ‘all dogs have four legs’ in the speeded condition that in the non-speeded condition, and were slower to judge quantified than generic statements, which suggests quantification is effortful, and participants were more susceptible to a GOG error when they were under time pressure. In sum, Meyer et al. () argue to have evidence that participants did sometimes judge universally quantified statements (‘all dogs have four legs’) as true when participants were instructed to respond as quickly as possible, but not when there was no time pressure, consistent with their predictions. The single study that investigates the processing of generics in the brain is Prasada et al. (), who used a combined behavioural and ERP study investigating the modulation of the N amplitude (an electrophysiological response that has been shown to reflect the processing of semantic information, specifically the integration of a word in the context of a preceding word, sentence, and discourse) as a function of subtly different morphosyntactic environments that condition either a generic (‘bananas are yellow’) or a non-generic (‘this banana is green’) readings. In Experiment , they measured whether the predictability of characteristic and uncharacteristic properties varies alongside constraints on generic/non-generic interpretation. In order to do so, they used (a) a Cloze procedure (Experiment a), where a participant has to fill in the blank of a sentence fragment such as ‘bananas are ____ or this banana is ____’, where the critical words were shown not to differ in predictability, and (b) a predictability rating task on a seven-point scale that showed that the predictability of (un)characteristic properties was in accordance with the constraints of (non-) generic interpretation. In Experiment , they used the same set of stimuli and recorded the electrophysiological response of the critical word, for example ‘yellow’ or ‘green’. They found a robust N effect. The N amplitude was shown to be sensitive to whether the critical word was interpreted as characterizing a kind or an instance of that kind, with the generic interpretation showing a significantly larger N. Additionally, the same words in the characteristic condition elicited a larger N in the generic condition (‘bananas are yellow’) than in the non-generic condition (‘the banana is yellow’). In sum, both true and false generics elicited an increased N in comparison to the non-generic statements, which led Prasada et al. () to reject the possibility that the truth-value of the statement had any effect. They then outlined some possible interpretations of their results that involve a) the differential involvement of semantic memory; b) the distinction between principled and statistical connections; and c) semantic relatedness, and they conclude that it is still unclear which difference between the interpretation of generic and non-generic sentences is reflected by N amplitude.
.. Experiments with generics in linguistically-driven research ... Typical adult populations Starting with work on English, Lazaridou-Chatzigoga & Stockall () compared generic generalizations with universally quantified statements with ‘all’, ‘all the’ and ‘every’ in a timed TVJ task that measured statement reading time and TVJ time. Statements involved majority and minority characteristic statements read after a preceding context that made exceptions salient. The study sought to measure the relevance of quantifier domain restriction and context for the statements in question. With respect to the reading time data, they found two effects. The first result is that the time to read the statements was significantly faster for the generic majority characteristic condition than any other condition. This might be taken as evidence that some generic statements are easier to process than the corresponding universally quantified statements, consistent with the GaD view. However, the background statement did not make the exceptions to the majority characteristic statements salient. Hence, it is not possible to determine whether this difference in reading times is due to the difference between majority vs. minority characteristic properties, or to the difference in the degree of relevance of the background context. The second result was that the time to make the TVJ for minority characteristic generic and ‘all’ statements was much longer than for the other two universal quantifiers (‘all the’, ‘every’), or for majority characteristic statements, contrary to the GaD, but consistent with participants engaging in a costly process of quantifier domain restriction. This result is also an indication that generics might be sensitive to context under certain circumstances, which is in line with recent work (Greenberg, ; Sterken, ) that claims that generics display some context sensitivity, but contrary to the received view that takes generics to strictly resist contextual restriction (Krifka et al., ). This study, however, did not manipulate enough features of the context to be conclusive. Building on the abovementioned study, Lazaridou-Chatzigoga et al. (a) addressed the relevance of quantifier domain restriction and context for generic and universally quantified statements in a TVJ task. They presented statements preceded by one of three levels of context: a) neutral, where the information in the context does not interact with the truth-value of the critical statement; b) contradictory, where it presents an exception which should rule out a universally quantified statement; and c) supportive. In their study, context did not only affect acceptance rates for ‘all’ and other universal quantifiers (‘all the’, ‘each’), but it further predicted the levels of quantifier domain restriction depending on the level of context. The effect of context was greater for ‘all the’ and ‘each’, two quantifiers that require quantifier domain restriction, while it was smaller for ‘all’, whose domain is only optionally restricted. Thus, Lazaridou-Chatzigoga et al. (a) advanced the quantifier domain restriction view that predicts the attested differences, and argued that proponents of the GaD view ruled out context prematurely. Furthermore, they find that context matters for generics too. In sum, these results show that context is a viable alternative explanation for much of the so-called GOG effect. Fuellenbach & Husband () studied the role of determiners in generalizing principled and statistical properties, focusing on how such information may be used to generalize about newly acquired properties of pseudo-words (similar to Cimpian, Brandone, &
-
Gelman, ). In their first study, they used bare plural generics and indefinite singular generics and in their second study bare plural generics and definite singular generics. Their results showed that indefinite singular generics are associated with increased expectations that a property is a principled connection to a kind, while definite singular generics are associated with decreased expectations that a property is a statistical connection to a kind. Regarding English, Spanish, and Brazilian Portuguese, Ionin, Montrul, & Santos () investigated genericity in characterizing sentences (e.g. ‘lions are dangerous’) and with kindlevel predicates (e.g. ‘dodo birds are extinct’) cross-linguistically in order to test the following issues, where disagreement in the theoretical literature exists: i) the status of bare (without a determiner) NPs in generic environments in Brazilian Portuguese; and ii) whether singular and/or plural generics are restricted to canonical kinds cross-linguistically. They used an acceptability judgement task, whose results provide support for Dayal’s () proposal that plural generics cross-linguistically denote kinds, whereas definite singular generics denote taxonomic entities. Regarding Brazilian Portuguese, the issue concerning bare nouns remains open, but see the discussion in Ionin, Montrul, Kim, & Philippov () and references therein for more details. Barton et al. () investigated definite article use with generic reference given the claims in the literature that there is variation in article use in the expression of generics in German (Tiger sind gefährlich vs. Die Tiger sind gefährlich ‘Tigers are dangerous’). Their results show that definite articles are optional with generic plural subjects and that bare subjects are preferred; definite plural subjects are accepted more often with kind-level predicates than with individual-level predicates. Age, regional background, and educational level seem to influence the attested variation, but the study did not include enough participants to make it possible to determine the effect of each potential factor.
... Bilingual and second language adult acquisition In recent years, genericity has been also studied in bilingual speakers and/or second language learners. This literature mainly contrasts some Germanic language (English, German) with some Romance language (Spanish, Italian, French), which differ in the way they typically express genericity, either with a bare or a definite plural. In a series of papers, Ionin, Montrul, and colleagues investigate genericity and language transfer in bilinguals and second language learners of English and Spanish (Montrul & Ionin , ; Ionin, Montrul, Kim, & Philippov, ; Ionin et al., ). For instance, Ionin et al. () examine the interpretation and judgements of both definite plural and bare plurals in L acquisition bidirectionally, with Spanish-speaking learners of English and English-speaking learners of Spanish. First language transfer was found in both directions when proficiency was lower, while there was more target-like performance when proficiency was higher. Kupisch and Barton investigate genericity in adult bilinguals and heritage speakers who speak German and some other language (see Kupisch, ; Kupisch & Barton, ; Barton, ). For instance, Kupisch () studies definite articles in specific and generic subject nominals in Italian spoken by adult simultaneous bilinguals and second language learners and argues that there are clear differences between the two groups of individuals which depend on which language is dominant.
Slabakova () examined the interpretation of bare plurals by L-Italian L-English and L-English L-Italian learners using a TVJ task, finding evidence for L transfer as well as recovery from L transfer, and proposed a mechanism for this recovery.
.. Generics in child language acquisition studies Related to first language acquisition, one of the questions that emerges from the overview of the theoretical accounts in section . is how children acquire generics given the absence of dedicated words or morphemes that encode genericity cross-linguistically. The formal semantics account posits the covert operator GEN, but how children come to posit such an operator is not explained. The GaD approach, on the other hand, argues that children do not face such a challenge because the unmarked and cognitively simpler generic is acquired by default and does not need to be learned. This picture set up a flourishing experimental agenda, which I will not discuss here in detail due to space limitations, but see Pérez-Leroux () for a recent comprehensive overview, as well as Lazaridou-Chatzigoga et al. () for discussion and a useful appendix. Summarizing very briefly, related to production, there are claims that generics can be found in the speech of children at the earliest multi-word stages. Gelman et al. () examined data from the CHILDES project and argue that all children for whom there was data produced generics at age , and by age children produced generics as frequently as adults. Regarding comprehension, Gelman () argues that children by the age of are sensitive to subtle morphological cues to distinguish generic from specific reference (e.g. ‘the Xs’ vs. ‘Xs’), by the age of they appropriately read contextual cues to determine that an utterance may be generic (e.g. interpreting a plural NP in the context of a single instance as signalling a generic intent), and by age they distinguish generics from the quantifiers ‘all’ and ‘some’ and have adult-like understanding of generics. Regarding second language acquisition and child bilingualism, the interested reader is referred to Pérez-Leroux et al. (), Kupisch (), and Serratrice et al. (), among others.
. R
.................................................................................................................................. It should be evident from the preceding review that the recent concentrated interest in genericity has deepened our understanding of the phenomenon and that experimental data can be used in order to advance a theory of generics. For instance, the classification of generics discussed in section ... seems particularly useful when one tries to decide whether different types of generics should be treated as a single phenomenon or whether we are dealing with several phenomena with a superficial similarity. In the process of doing so, different frameworks can be proposed and validated against both theoretical and datadriven arguments. The recent emergence of experimental studies on genericity is welcome, as these initial steps are crucial in order to point to some of the possible directions this research can follow. The majority of the studies reviewed here used off-line judgement data with only a few
-
studies addressing the real-time processing of generics. On top of that, the majority of the TVJ tasks gave the participants a binary choice, whereas Likert-scale rather than binaryjudgement tasks are essential for the study of subtler aspects of meaning (see Katsos & Bishop, ). This is because the less fine-grained binary judgement tasks (which require a ‘yes’/ ‘no’ or ‘accept’/ ‘reject’ type of response) are liable to misrepresent participants’ competence by conflating mildly infelicitous and downright unacceptable statements (since both are lumped together as ‘no’ or rejections). Given recent analyses of generics that place them at the heart of human cognition, one would not only want to study conscious (i.e. judgement data) procedures, but also subconscious mechanisms. Consequently, other methods that could shed light on the phenomenon would include self-paced reading studies, priming, visual world paradigms, and brain-based methods including electroencephalography (EEG), functional Magnetic Resonance Imaging (fMRI), and magnetoencephalography (MEG). To illustrate just one potential study, a priming paradigm could be used in a study in order to evaluate the relevance of prototypes for exceptionality in generics. The participants would be asked to merely provide TVJs related to generic statements, but exceptions would be used as primes before the judgement. For instance, before judging ‘tigers have stripes’ the participants would see a) a picture of a striped tiger (a prototypical instance); b) a picture with an exceptional tiger, such as an albino tiger; or c) an unrelated picture used as a control. This study would evaluate whether there are priming effects on people’s judgement, as well as whether the different conditions influence the time-course of processing. Furthermore, while research has started to emerge on the neural basis of quantifiers (see for instance Shetreet et al., , and Troiani et al., , using fMRI), there is only one study that investigates generics in the brain (Prasada et al., ; see earlier for discussion). It is important to study the basic composition of generic statements with neuroimaging techniques in order to lay the foundations of the phenomenon on the neural level. Only with that kind of work would we be in a position to increase our understanding of the neural bases of genericity and to subsequently compare the brain regions involved in genericity to the regions involved in quantification. Before concluding, I would like to highlight four main outstanding issues. First, more research is needed on the relationship between genericity and quantification and the role of genericity in human cognition more broadly. Is there indeed a ‘generic bias’ for interpreting quantificational statements as generic? If so, does this propensity originate in cognition or in language, and in either case, what are the mechanisms of the bias? This investigation seems pressing given the prominence of generics as our major means to conceptualize about the world, the attention the GaD approach has received recently, and the use of generics to express stereotypes. Second, despite the fact that this chapter did not review the child acquisition data in detail, the question of how children learn generics seems to play an important role when one is to judge the explanatory power of one theory over another: the formal semantics approach cannot offer any clear explanation for the fact that generics seem to be early and easy to acquire, whereas the GaD view relies on acquisition data in order to advance the idea that generics are a default interpretation. How adult-like children’s interpretation is of generics and whether different aspects of genericity are acquired at different ages remain still open issues (see Lazaridou-Chatzigoga et al., b). Third, variation within a language and cross-linguistic variation need to be addressed more consistently and systematically. Although genericity is common in all languages,
languages make use of different morphosyntactic means to express genericity. To illustrate, within languages that have articles, some languages like English may use a bare plural in the subject position to express a generic like ‘tigers have stripes’, while other languages like Greek cannot use a bare plural and express the relevant statement with a definite plural i tighris ehun righes ‘(the) tigers have stripes’, which is furthermore ambiguous between a generic and a specific (definite) interpretation. Even though there is theoretical and descriptive work in several languages, the main bulk of experimental studies have been performed in English (with only a few exceptions). Much work needs to be done in order for one to evaluate claims about the relative (un)markedness and complexity of generic vs. quantified statements and to determine the language-specific and language-general features of generics. Fourth, another under-researched area that is amenable to experimental investigation is the study of the frequency of generics in everyday speech and other kinds of register. This is relevant both in order to appreciate the pervasiveness of generic language, but also in order to characterize the input that a child or an adult second language learner receives and how they can use it in order to figure out the intricacies of generic meaning. Related to that, we would not only want to establish the absolute frequency of generics, but also measure how frequent the different types of generics within a language are, as well as compare these frequency measures to the frequency of other kinds of generalization (universal, existential, proportional, etc.), as well as to specific or particular statements.
. C
.................................................................................................................................. This chapter investigated recent experimental research on genericity, placing it within the context of the most prominent theories in two different disciplines, namely in formal semantics and cognitive psychology. Both the theoretical and the experimental research on the topic have substantially provided further insight into what generics are and how they work. New questions emerge from the experiments reviewed here, which will hopefully be taken up by investigators across disciplines leading to an ever-larger explosion of research on the topic. I hope that this overview will stimulate further work and that more scholars will turn their attention to generics, especially in order to do interdisciplinary work that integrates the tools and perspectives of both strands of investigation.
A Preparation of this chapter was supported by AL /- (DFG Gottfried Wilhelm Leibniz Preis to Artemis Alexiadou). I would like to deeply thank my two collaborators, Napoleon Katsos and Linnaea Stockall, for countless insightful discussions on the topic. My understanding of the issues presented here has also benefited from discussions with former and current colleagues at Queen Mary, the University of Cambridge and the RUESHeL group at Humboldt-Universität zu Berlin. The author is also grateful to Chris Cummins for his comments on the first version of this chapter, as well as to Alda Mari for organizing the ‘GENIUS: Genericity: Interpretation and Uses’ conferences that brought together linguists and psychologists. I feel lucky to have participated in its third edition in , as it both sparked and consolidated my interest in an interdisciplinary approach to the topic. All errors or misunderstandings are my own.
......................................................................................................................
......................................................................................................................
, ,
. I
.................................................................................................................................. A least since Barwise & Cooper (), the dominant view on the semantics of natural language quantifiers has been that they express relations between sets. For example, all expresses the inclusion relation: all men are mortal is true if and only if the set of men is included in the set of mortal beings. Similarly, some expresses the overlap relation: some politicians are corrupt conveys that there is a non-empty intersection between the set of politicians and the set of corrupt individuals. While the general idea behind this style of analysis for expressions of quantity is still in place as one of the pillars of formal semantics, the scope of what is collectively known as Generalized Quantifier Theory (GQT) (following Barwise & Cooper, ) has turned out to be quite limited. For instance, simply considering expressions like many or most to be relations between sets obscures the fact that at some level, such expressions are interpreted via mechanisms from the adjectival domain: the semantics of many is (related to) that of an adjective in its positive form (Solt, , for instance); most is inherently superlative (Hackl, ). Perhaps the clearest example of an area where the literature has diverged from the GQT conception of quantifiers is that of modified numerals. It is very easy to give a semantic analysis of modified numerals such as more than fifteen in terms of relations between sets: more than fifteen corresponds to the relation between sets where the overlap between the sets consists of more than fifteen individuals. There are two (independent) issues with such an analysis. First, it is hard to understand how modified numerals can correspond to relations between sets of individuals given that there are good arguments that a Determiner Phrase (DP) like more than fifteen biscuits has the structure [more than [ biscuits]] rather than [[more than ] [biscuits]] (Krifka, a; Geurts & Nouwen, ). Only the latter is compatible with the GQT view. The second issue, and one that is central to this chapter, is not so much whether modified numerals express relations between sets of entities, but rather, irrespective of how modified numerals express their cardinal content, whether their meaning can indeed be assumed to merely involve the counting of individuals. At the core
of this part of the discussion is the observation, due to Geurts & Nouwen (), that more than n and at least n + are not semantically equivalent. Clearly, if x is a natural number, then x > n if and only if x ≥ n + , and so if modified numerals simply concern the cardinality of the overlap between two sets, then it remains a mystery as to where the source of the non-equivalence lies. Geurts & Nouwen () identify several semantic differences between more than and at least. Of these, the fact that they support different inference patterns has proven to have the most enduring appeal in the literature. Take for instance the premise that The final paragraph contained three typos. Geurts & Nouwen () observe that while it clearly follows from this that The final paragraph contained more than two typos, it is not so clear that we may conclude that the paragraph in question contained at least three typos. The intuition behind what blocks the inference is that somehow The paragraph contained at least three typos suggests that there may have been more. Intuitively, then, at least conveys epistemic information about numerical alternatives. This is supported by examples like (), which are odd since they appear to suggest that the speaker is ignorant about how many children she has. This oddness disappears in the parallel sentence with fewer than. () ()
?I have at most three children. I have fewer than four children.
Nouwen () argues that the difference between superlative (at least/most) and comparative (more/fewer than) modifiers is more accurately described as the former, but not the latter, imposing a on the values under consideration (compare Schwarz et al., ; Nouwen, ). For instance, () is felicitous as opposed to the infelicitous () since the number of sides of the base of a pyramid is one of a number of possible values (anything that exceeds two), while there is just one possible number of sides to a triangle (namely three). This means that there is no salient way to interpret (), except perhaps to interpret it with respect to a range of epistemic values (values that the speaker holds possible), but this suggests that the speaker does not know what a triangle is. Hence the oddness. ()
?A triangle has at least three sides.
()
A base of a pyramid has at least three sides.
Nouwen () furthermore observes that the difference between comparative and superlative modifiers extends to a larger set of modifiers, thereby establishing two distinct classes: class A modifiers are modifiers that lack ignorance (or, more generally, range requirement) effects while class B modifiers display these inferences. It turns out that languages tend to have comparative modifiers and locative prepositions (over, under) as class A modifiers and superlative adverbs (at least/most, maximally/minally), directional prepositions (up to, from) and disjunctions (n or more/fewer) as class B modifiers. A dominating theoretical trend has emerged over the past decade encompassing approaches that assume a partially pragmatic analysis for the observations earlier; one in which the interpretation effects observed with superlative modifiers are related to the
, ,
implicatures of disjunctive statements (Büring, ; Cummins & Katsos, ; Schwarz, ; Coppock & Brochhagen, a; Kennedy, ; Nouwen, ). The idea shared by these approaches is that the mechanism responsible for ignorance and free choice implicatures in disjunctive statements is also responsible for the meaning component that distinguishes class B modifiers from class A. Consider, for example, a disjunction like (), which gives rise to the strong inference that the speaker does not know in which city John is. ()
John is in Brussels or Antwerp.
The ignorance inference for () is usually explained in terms of simple Gricean reasoning, as follows. Utterances come with sets of (stronger) alternatives. In accordance with the maxim of quantity, for each alternative p to (), the hearer can assume that the speaker lacks the belief that p is true. If the speaker is furthermore assumed to be knowledgable, then the hearer can conclude that the speaker must believe the falsity of p. Some alternatives are symmetric. For instance, for () both John is in Brussels and John is in Antwerp are more informative. However, the hearer cannot assume that the speaker is knowledgeable about both: for in that case, given that neither of these alternatives was used, the speaker must lack the belief that either is true. And so, if the speaker were knowledgeable, then she must have had the belief that both of these alternatives are false, but then the utterance () itself would have been false as well. The conclusion the hearer can draw now is that the speaker must not know which of these alternatives is true (Sauerland, ; Fox, ). Pragmatic theories of superlative modifiers assume that at some level of description, modified numerals like at least three correspond to a disjunction of values, say three or four or five or six or . . . One way to do this is to assume that at least three has the alternatives of such a disjunction. That is, the pragmatically active alternatives are given by the set {, , , , . . . }. The result of using at least three is now that it is associated with a large number of symmetric alternatives. This explains why such modified numerals trigger ignorance inferences. The parallel with disjunction goes further, however. Just as the ignorance implicatures for disjunctions can be obviated in the scope of quantifiers, the ignorance effect of at least and at most disappears in exactly the same environments. The most salient reading of () is that some people ate an apple, some ate a pear, and no one ate neither an apple or a pear. Similarly, () says that some found three typos, some found more, but no one found fewer. ()
Everyone ate an apple or a pear.
()
Everyone found at least three typos.
Given the parallel, the idea in the literature is that the pragmatic mechanisms responsible for ignorance and ignorance obviation with disjunction govern the interpretation of superlative quantifiers too. A slightly different way to implement the pragmatic approach is to assume that the alternatives are not pragmatic but semantic: like disjunction, class B modified numerals have an alternative semantics. This alternative-inducing semantics triggers pragmatic effects like ignorance based on reasoning why the speaker used an alternative-inducing expression (Coppock & Brochhagen, a).
As we said earlier, the pragmatic approach (in its various guises) dominates the field. We want to mention two views, though, that differ in interesting respects from what is now mainstream. First of all, Geurts & Nouwen () propose an essentially semantic account in which superlative modifiers are inherently modal. According to their proposal, the relation to the speaker’s epistemic state is an explicit part of the lexical semantics of at least. The fact that this epistemic effect may be obviated is accounted for by means of assumptions about how modal statements may change flavour when embedded in other quantificational environments. There are some clear issues with this proposal, which we won’t review here. In the context of experimentation, however, the theory is interesting since it generates predictions about what would happen if we were not to assume a pragmatic analysis of the inferences at issue. An altogether different account is proposed by Cohen & Krifka (), who assume that superlative modifiers ‘are interpreted at the level of speech act’. In a nutshell, according to Cohen and Krifka, at least three expresses that three is the minimal number for which the speaker is unwilling to assert the negation of the sentence in question. For reasons of space, we cannot go into the details of this analysis, but where relevant we will try to evaluate its predictions later.
. I
.................................................................................................................................. A large series of experiments has recently been carried out to investigate the extent to which numeral modifiers give rise to ignorance inferences and whether these arise semantically or pragmatically. The experimental investigations differ in whether they tested these inferences directly or indirectly and in the designs and methodologies used.
.. Experimental support for the semantic account and the class A/B distinction The first experiment where ignorance effects seem to be attested with superlative numeral modifiers is reported in Geurts & Nouwen (). Geurts & Nouwen carried out two paper-and-pencil experiments in Dutch testing the inference patterns of the relevant superlative and comparative numeral modifiers. Participants had to judge the validity of arguments like the following: ()
Premise: Beryl had n sherries.
()
Conclusion: Beryl had {at least n / at most n / more than n / fewer than n + } sherries.
Conclusions with comparative modifiers were almost unanimously accepted, while conclusions with at most were accepted in per cent of the cases and those with at least in half of the cases. This difference between comparative and superlative modifiers can be attributed to the fact that the latter give rise to ignorance inferences and thus the relevant conclusions are at odds with the competence and certainty conveyed by the premises, for
, ,
example, (). Geurts & Nouwen (), who take this method to test for entailment relations, interpret these results as favouring their semantic account, whereby ignorance is encoded in the lexical semantics of superlative modifiers. Geurts et al. () found a similar contrast between comparative and superlative modifiers in a similar experiment. While the responses to statements with comparative numeral modifiers were at ceiling, only half of the participants judged statements with the corresponding superlative modifiers as valid. This distinction manifested itself in an online verification study, in which they found that items with superlative modifiers incurred longer decision times than items with comparative modifiers. Geurts et al. () too take this contrast to indicate a semantic difference between comparative and superlative modifiers in favour of the semantic account in Geurts & Nouwen (). Before we proceed to survey the experimental studies that aim to adjudicate between a semantic and a pragmatic account, we would like to mention two studies that found a difference in the acquisition time of superlative and comparative modifiers. Musolino () tested -year-olds’ understanding of the numeral modifiers more than, at least, at most, and exactly in a truth-value judgement task. While the children performed well on items with more than and exactly, they performed at chance level on items with superlative modifiers. Geurts et al. () partially replicated Musolino’s () findings with -year-old participants in an action-based task. Geurts et al., however, did not find a clear-cut distinction between comparative and superlative modifiers; rather, the significant difference was between the upward-monotone modifiers at least and more than, in which the children scored better, and the downward-monotone modifiers at most and less than, for which scores were lower. The only significant superlativecomparative comparison was for at most and less than. Taken together, these acquisition studies show that the meaning of superlative modifiers is acquired later than that of comparative modifiers.
.. Evidence for a pragmatic account In order to determine whether ignorance inferences are semantic or pragmatic, Cummins & Katsos () employ a more elaborate method that distinguishes between logical contradictions and pragmatic infelicities. They tested whether sentence continuations in which an utterance like () with a numeral modified by exactly follows an utterance with a numeral modifier are judged as entailing, contradictory or pragmatically degraded continuations. Participants were asked to judge how coherent an utterance consisting of a pair of sentences is on a Likert scale from – (incoherent) to (coherent). Katsos & Bishop () find that when asked to provide gradient, rather than binary, ratings for utterance felicity, participants successfully differentiate between semantic entailments, logical contradictions, and pragmatically infelicitous statements. Based on these observations, Cummins & Katsos () predicted that logical contradictions will be judged as incoherent, scoring at the lower part of the Likert scale, continuations where the second sentence entails the first (henceforth logical entailments) would be judged as coherent, with scores at the upper part of the scale, and continuations that deny pragmatic inferences (i.e. implicature cancellations) would be judged more coherent than the former, but yet less coherent than the latter, scoring at the middle of the scale.
at least = more than n houses: at most = fewer than Specifically n þ =n ; she has exactly houses In fact n =n
() Jean has
()
Jean has some houses. 8 9 < none = Specifically ; she has half of the houses: In fact : ; all
(target item)
(control item)
The control conditions (i.e. logical contradictions, logical entailments, and pragmatic infelicities, respectively, in ()) obtained scores as predicted. The continuations at least n . . . exactly n and at most n . . . exactly n+ received higher scores than the test logical contradictions and lower than their comparative counterparts (i.e. more than n . . . exactly n+ and fewer than n . . . exactly n ), the latter scoring as high as control logical entailments. The difference between superlative and comparative modifiers was found for items in which the second sentence was introduced by specifically or by in fact, with the latter conditions receiving higher coherence scores in general. The fact that participants found the second sentence less acceptable when following a sentence with a superlative modifier rather than a comparative modifier can be attributed to the ignorance effect triggered by the former, which clashes with the knowledgeability expressed by the sentence with the numeral modifier exactly. Moreover, the earlier comparison of superlative modifiers with the test logical contradictions suggests that ignorance effects with superlative modifiers are pragmatic inferences, whose cancellation yields infelicity, similarly to the control pragmatic infelicities with some in (), although less robustly, as the cancelled ignorance inferences were judged as more felicitous than the cancelled scalar implicature. Cummins & Katsos’s () findings go against the semantic account by Geurts & Nouwen () and are taken to support their own pragmatic account of ignorance, where superlative modifiers are interpreted as disjunctions and ignorance arises as a quantity implicature. Recall that Geurts & Nouwen’s () and Geurts et al.’s () results did not exhibit wholesale rejection of utterances that deny ignorance inferences and could therefore be accommodated, too, by an account where ignorance is derived via a pragmatic mechanism. McNabb & Penka () use a similar methodology, where the task also involves judging coherence on a Likert scale. In contrast to Cummins & Katsos (), McNabb & Penka () use rich discourse contexts as in (), followed by the target sentence (). ()
After the soccer practice, Cassandra sent Eduard to collect all of the soccer balls lying around in the court. He collected the n balls that were scattered around the court. When he was done, he went back to Cassandra and said:
() ‘I collected { at least / at most } { n / n / n þ } balls.’ The results replicated the findings of Cummins & Katsos () in that combinations like () () were judged to be more coherent than logical contradictions (i.e. at least n þ /at most n ). Again, this finding challenges the semantic account in Geurts & Nouwen ().
, ,
There is also evidence against the semantic account that does not involve coherence judgements. Coppock & Brochhagen (b) argue that Geurts et al.’s () methodology might not be exclusively testing for entailment relations, as what seems to matter in validity judgement tasks is whether the conclusion and premise(s) share common information, be it truth-conditional or pragmatic (see also Ariel, , on the difference between lexical semantic content and semantic-pragmatic truth compatibility). Thus, it is very likely that pragmatics might have played a role in Geurts & Nouwen’s () and Geurts et al.’s tasks, calling into question those authors’ interpretation of the relevant findings in favour of Geurts & Nouwen’s semantic account of ignorance. Coppock & Brochhagen (b) use a truth-value judgement task, under the assumption that this is a better method for identifying entailments as it blocks pragmatic intrusion. In their task, native speakers of English had to judge whether a sentence like () was true or false given a picture of n objects of type X. ()
There are { at least n / at most n / more than n / fewer than n þ } Xs in the picture.
Although their target sentences were very similar to those used in Geurts & Nouwen () and Geurts et al. () and had to be assessed against a similar context, albeit this time visually depicted and not written, Coppock & Brochhagen’s (b) results were very different. All (superlative and comparative) test conditions yielded surprisingly high percentages of true judgements, and did not significantly differ from each other, but crucially differed from the control logical contradictions (i.e. at least n þ , at most n , more than n, fewer than n), which received very low percentages of true judgements as expected. These results go against Geurts & Nouwen’s () semantic account of ignorance effects, which would predict that sentences with superlative modifiers would be judged as contradictions against a depiction of an exact number of objects, leading to superlative conditions patterning with contradictions but not with comparative conditions. Coppock & Brochhagen (b) therefore conclude that their findings lend support to a pragmatic account of ignorance.1
.. (More) direct investigations of ignorance inferences Although ignorance inferences are a central characteristic that all theoretical accounts of numeral modifiers aim to capture, the experiments discussed so far did not include a direct 1 It is worth mentioning that both Coppock & Brochhagen (b: Experiment ) and McNabb & Penka (; in the experiment described earlier) found a contrast between at least and at most, which one could not explain under the previously mentioned assumption: at most n þ was considered significantly less coherent than the similar at most n, at least n, and at least n test conditions. The authors explain this asymmetry in the following terms. Participants have difficulties accepting the speaker’s utterance at most n þ given a quantity n introduced in the context, because she is stating the possibility that a greater quantity (i.e. n þ ) is true while she knows that it is false given n in the preceding context. This is not the case for at least n, although this too allows for the possibility of a higher quantity than n that is not true given n in the context. The contrast arises because at most n þ , as opposed to at least n, explicitly states, or highlights in Coppock & Brochhagen’s (b) inquisitive semantic terminology, the possibility that a greater number than n is true.
investigation of ignorance inferences. McNabb & Penka () were the first to directly test for ignorance inferences and arguably to provide direct evidence that ignorance effects are pragmatic inferences. McNabb & Penka investigated how speakers interpret superlative modifiers within an ignorance context by manipulating the speaker’s epistemic state. In their experiment, superlative modifiers were embedded under universal and existential deontic modals, as the authors’ greater aim was to determine which combinations of superlative modifiers and modals obviate ignorance inferences. In their experimental task, participants had to judge whether the speaker’s utterance was compatible with a preceding context that explicitly described her epistemic state, as illustrated in (), on the same type of modified Likert scale used by Cummins & Katsos (). () a. þ knowledgeable speaker: the secretary, who was involved in the selection process, said: b. knowledgeable speaker: the secretary apologized for not knowing the requirements for the application, and said: ‘You are {allowed / required} to have {at least / at most} three works in the portfolio you send us.’ In the +knowledgeable speaker condition, the combinations allowed to + at most and required to + at least received significantly higher coherence ratings than the combinations allowed to + at least and required to + at most. The authors argue that this difference arises because the latter combinations trigger ignorance inferences, which are in conflict with the epistemic state of the speaker in that particular context. They further conclude that the difference of these latter combinations from the control logical contradiction items suggests that ignorance with superlative modifiers is a pragmatic inference. Although the control items were completely different items, this contrast indirectly shows that ignorance should not be accounted for in the style of Geurts & Nouwen (), but rather by a pragmatic analysis. McNabb & Penka () expected that in the knowledgeable speaker condition all superlative modifier-modal combinations would be well-formed or at least not significantly different, under the assumption that all combinations have a potential parse that gives rise to an ignorance inference. However, the results in this condition were similar to the ones in the + knowledgeable speaker condition: allowed to + at least and required to + at most still got lower scores. The authors speculate that this could be because participants might have made the assumption that the speaker did in fact have knowledge of the exact required or permitted amounts by virtue of them being, for example, a secretary in the relevant company in (), and thus this confound may have prevented the licensing of ignorance inferences. Westera & Brasoveanu () conducted two self-paced reading experiments in English, each followed by a validity judgement task, aiming to directly study the context-sensitivity of ignorance effects with numeral modifiers. They assume that in precise contexts, namely when there is a how many Question Under Discussion (QUD), superlative and comparative modifiers trigger ignorance effects to the same extent. Their next assumption is that, if there is an underspecified QUD, superlative modifiers are more likely to give rise to ignorance than comparative modifiers. They base these assumptions on a corpus study in which they find that superlative modifiers are more frequent in precise contexts, hence in underspecified contexts people are more likely to accommodate a precise context in the case of superlative
, ,
modifiers, which is the right setting for triggering ignorance. Finally, in imprecise or coarse contexts no ignorance is expected to arise with superlative or comparative modifiers. Participants in Westera & Brasoveanu () were asked to read short exchanges between a judge and a witness in a court setting, as illustrated in (). First, the judge’s question, which set up the QUD, was presented in its entirety. Then, the witness’s response, which contained the superlative modifier at most or the comparative less than and a perception verb, was presented word-by-word. The witness’s statement was followed by the judge’s conclusion that the witness did not know the exact quantity of the objects in question. Participants then had to assess how justified the judge’s conclusion was on a Likert scale from to , where was not justified at all and was strongly justified. ()
The judge asks: a. Imprecise QUD: ‘Did you find {at most / less than} ten of the diamonds under the bed?’ b. Underspecified QUD: ‘What did you find under the bed?’ c. Precise QUD: ‘How many of the diamonds did you find under the bed?’ Witness: ‘I found {at most / less than} ten of the diamonds under the bed?’ Judge’s conclusion: ‘The witness doesn’t know exactly how many of the diamonds she found under the bed.’
Participants found the judge’s conclusion to be more justifiable when the QUD was precise (how many QUD) or underspecified (what QUD) than when it was imprecise (polar QUD). In addition, when just considering the responses to a precise QUD, the judge’s conclusion about the witness’s ignorance was more justifiable when the witness had used at most vs. less than. Furthermore, when all the data were taken together, no interaction between the QUD and the numeral modifier factors was found, nor a main effect of the numeral modifier. These results indicate that the QUD does play a role in how people interpret sentences with numeral modifiers and more specifically influences the likelihood of ignorance readings arising, regardless of the type of the numeral modifier. Interestingly, these results are not expected given an account that takes ignorance to be part of the lexical semantics of superlative modifiers, which predicts ignorance to arise in all QUD conditions to the same extent and only with at most. In order to make sense of the difference between the superlative and comparative modifiers in the how many QUD condition, Westera & Brasoveanu () hypothesize that this QUD is underspecified to some extent, citing corpus data that lend support to this hypothesis and further suggest that what questions might be interpreted more precisely. In order to further test for QUD-sensitivity, they modified the QUD conditions, as shown in (). ()
The judge asks: a. Imprecise QUD: ‘Approximately how many of the diamonds under the bed?’ b. Precise QUD: ‘Did you find eight, nine, ten, or eleven diamonds under the bed?’ c. Precise QUD: ‘Exactly how many of the diamonds did you find under the bed?’ Witness: ‘I found {at most / less than} ten of the diamonds under the bed.’Judge’s conclusion: ‘The witness doesn’t know exactly how many of the diamonds she found under the bed.’
Similarly to the first experiment, precise QUD conditions (exactly how many and multiple-choice questions) led to higher ignorance inferences than the imprecise QUD condition (approximately how many questions). Also, there was no difference between the numeral modifiers. These results too indicate that ignorance inferences are contextsensitive and that they arise with both comparative and superlative modifiers. In the pursuit of a link between the degree of ignorance inferences and the time-course of these (and related) inferences, Westera & Brasoveanu () analysed reading times from their first experiment in the regions following the modified numeral in the witness’s response. Reading times for the numeral region (e.g. ten) and (some) subsequent regions were slower in the witness’s reply to the underspecified (b) and precise (c) QUDs than in the reply to the imprecise QUDs (a). The word under was also read slower when a precise QUD was answered by the witness using a superlative modifier, but this was not the case for the precise QUD & comparative condition. Westera & Brasoveanu () link the increase in reading times in the underspecified and precise QUDs vs. imprecise QUD to the increase of ignorance readings in these conditions in the off-line task and the interaction they find between precise QUD and superlatives. They attribute the increased reading times to the costly on-line calculation of pragmatic ignorance inferences or to the relevant silent intonational effects during reading depending on the context. Surprisingly enough, in the second on-line task, which included more clear-cut QUD conditions, there was no significant effect of QUD condition. Westera & Brasoveanu () blame this on a possible habituation effect, as in their second experiment the filler items included a judge inferring a witness’s ignorance, too. This study shows that ignorance inferences are (i) available with both superlative and comparative modifiers (depending on the context) and (ii) context sensitive, and as such they should be captured in pragmatic terms. However, apart from the replication failure of the findings of the first on-line experiment, another concern this study raises has to do with the use of perception verbs in the target sentences, which lead to a strong epistemic inference, thereby possibly clashing with the use of a numeral modifier in the underspecified and precise QUD conditions and thus affecting reading times. Note also that it is not entirely clear whether participants generated ignorance inferences on-line in the target sentences or whether they did so when prompted by the judge’s conclusion about such inferences. Alexandropoulou et al. () carried out an eye-tracking reading study in Dutch with the aim of obtaining precise data on the real-time processing of ignorance inferences in knowledgeability vs. + knowledgeability contexts. They manipulated the context preceding a target sentence with the superlative modifier at least in a way similar to that in McNabb & Penka (), as illustrated in () (translated from the original Dutch). ()
Sophie is a figure skater and very dedicated. Last weekend she was going to try to train as intensively as possible. (+knowledgeable:) I can tell you how much because I talked to her yesterday. ( knowledgeable:) I’m not sure how much exactly, but this is what I think: Sophie practiced at least seven hours on the ice. (target)
As is clear from (), the knowledgeable speaker context forces an ignorance reading, which is compatible with the core meaning of at least (e.g. n ≥ in the example item) as well
, ,
as with the ignorance inference.2 The + knowledgeable speaker context, in contrast, is compatible with the core meaning of at least, but at odds with the ignorance inference. This context manipulation was inspired by Breheny et al. (), who used a self-paced reading task to measure reading times of scalar terms (or and some) in a context triggering a scalar implicature, which was compatible both with the semantics of the scalar term and the scalar implicature, and in a context that did not trigger that scalar implicature but was compatible with the semantics of the scalar term. Breheny et al. () found a slowdown at the region of the scalar expression in the former (semantic and pragmatic compatibility) condition, which they associated with scalar implicature calculation occurring on-line and being costly. A similar effect was found by Panizza et al. (), who tested the interpretation of bare numerals (one-sided vs. two-sided) with an eye-tracking reading task with a similar context manipulation. According to this context design and Alexandropoulou et al.’s () prediction, if ignorance with at least is a pragmatic inference, which has to be computed on-line and is costly, parallel to run-of-the-mill scalar implicatures (following Bott & Noveck, ; Breheny et al., ; Huang & Snedeker, a, inter alia), a similar effect should be attested in the ignorance-triggering context in which the speaker is knowledgeable.3 The region of seven hours, where the interpretation of the whole modified numeral is completed, exhibited longer reading times when the target sentence was uttered by a knowledgeable speaker as opposed to a + knowledgeable speaker (significant effect in total reading times, marginal in first pass and rereading probability).4 There was also a spillover effect of speaker’s knowledgeability in the region of the Prepositional Phrase (PP) on the ice, whereby the PP was more likely to be re-read in the knowledgeability speaker condition than in the +knowledgeability speaker condition. These results are parallel to those found by Breheny et al. () and Panizza et al. () for other scalar expressions. Similarly to those studies, they are taken to be due to the on-line derivation of a pragmatic inference, especially an ignorance implicature. As such, this study offers a direct indication of the pragmatic status of ignorance inferences with superlative modifiers, disfavouring semantic accounts of ignorance. According to semantic accounts, an opposite effect should have been attested, namely a slowdown in the + knowledgeability speaker condition, due to the contradiction of the speaker’s epistemic state set up by the context and of that set up by the lexical semantics of at least.5
2 Note that the speaker in knowledgeability contexts has partial knowledge of the number at stake rather than complete ignorance, cf. I’m not sure how much exactly in (18). The assumption is that it would be extremely odd for a speaker with complete ignorance to utter any number. 3 In addition to manipulating the speaker’s epistemic state, Alexandropoulou et al. () also added stimuli that contained a modal. For instance: Sophie wanted to practice at least seven hours on the ice. The idea was to test a prediction made by Büring () and Kennedy () that, in order for ignorance to arise with a modal verb, at least needs to take wide scope. Since this is presumably a costly operation (cf. Hackl et al., —see also Brasoveanu & Dotlačil, Chapter in this volume), an interaction effect was predicted between verb type (modal versus non-modal) and speaker’s epistemic state (ignorance versus knowledgeability). However, no such interaction effect was found. 4 The numeral was grouped together with the noun to comply with the predominant syntactic structure, namely [more than [15 biscuits]] rather than [[more than 15] biscuits], as discussed in the introduction. 5 An on-line experiment we haven’t discussed so far, but which deserves mention nonetheless, is Hacohen et al. (), which was designed to test the predictions made by Cohen & Krifka ().
In summary, experiments investigating the availability of ignorance inferences with numeral modifiers, both off-line and on-line, converge on the finding that ignorance inferences of superlative modifiers are derived by a pragmatic mechanism, be it in the neo-Gricean style of Büring (); Cummins & Katsos (); Kennedy (); Schwarz (), who derive them as scalar implicatures, or in the style of Coppock & Brochhagen (b), who derive them via a different independently motivated pragmatic principle. In what follows, we discuss another type of inferences of numeral modifiers, the so-called variation effects, which are triggered when numeral modifiers appear in certain embedding environments.
. V/
.................................................................................................................................. In the previous sections, we discussed experimental investigations of ignorance inferences that modified numerals give rise to. If, as such investigations suggest, these inferences are indeed pragmatic, and if they are indeed linked to similar inferences attested for disjunctive statements, then it is expected that obviation patterns may occur: whenever a superlative quantifier is in the scope of another quantifier (be it nominal or modal), the ignorance inference is no longer obligatory. When modified numerals are embedded under universal nominal quantifiers (as well as plurals and generics), they give rise to variation effects (Nouwen, ). To repeat the example from earlier, Everyone found at least three typos seems to have a reading on which it conveys that some found three typos and some found more. Most accounts of such variation inferences for modified numerals under universal nominal quantifiers derive these inferences pragmatically, though they differ in details (Büring, ; Mayr, ; Schwarz, ; Coppock & Brochhagen, a; Nouwen, ). Cummins & Katsos () investigated statements with comparative and superlative modifiers embedded under the distributive quantifier each. Cummins & Katsos sought to find whether participants thought that (a) implies (b) and (a) implies (b), thereby indirectly investigating variation inferences. Recall that the inference that () and () give rise to is that the speaker knows precisely how many children each one of Anne and Brian has, as long as neither of them has fewer than three children. ()
a. Anne has three children but Brian has at least four children. b. Anne and Brian each have at least three children.
According to this approach, at least three expresses that three is the minimal number for which the speaker is unwilling to assert the negation of the sentence in question. This inference, which is derived by implicature, predicts a difference between true and false responses: a sentence containing at least three would be semantically false if the actual exact number of the object in question is and pragmatically true if the actual exact number is . Hacohen et al. () did indeed find that reaction times for true judgements were significantly longer than reaction times for false judgements. While Hacohen et al.’s () assumptions on how the pragmatic inferences are derived are different to those in the dominant family of accounts of numeral modifiers (Büring, ; Cummins & Katsos, ; Schwarz, ; Coppock & Brochhagen, a; Kennedy, ; Nouwen, ), the results still support a pragmatic rather than a semantic account.
()
, , a. Anne has three children but Brian has more than three children. b. Anne and Brian each have more than two children.
Cummins & Katsos () argue that the variation inference in () can be easily derived from their pragmatic, disjunctive-inference account of superlative modifiers but not from the semantic, modal account in Geurts & Nouwen . They find that while the inference with more than in (b) is accepted in per cent of the cases, the inference with at least in (b) is accepted in per cent of the cases only (and even less frequently, at % of the time, with at most), showing that variation inferences are robust but the extent to which they are generated depends on the type of numeral modifier. Alexandropoulou () tested the likelihood and strength of variation effects in superlative modifiers under universal nominal quantifiers as a way to gauge whether these inferences should be characterized as pragmatic or semantic inferences. In her experiments (conducted in Dutch), participants read statement-question pairs like the one in () and were asked to evaluate how reasonable the question was in response to the answer on a modified Likert scale. ()
a. According to a random sample every bag contains at least 22 liquorice candies. b. Do they all contain the same number of liquorice candies?
While the question in (b) is not incongruent with the assertion in (a), which can be paraphrased as ‘for every bag the number n of liquorice candies is ≥ ’, it is at odds with the potential inference that ‘there is no specific n such that all bags contain exactly that many liquorice candies’, which is the variation reading. Following the methodology of the coherence experiments in Cummins & Katsos (), responses for target items like () were compared to ratings given to contradictory and semantically and pragmatically noncontradictory controls. Items like () were rated significantly better than semantic contradictions but worse than semantically and pragmatically non-contradictory controls, showing that variation effects indeed arise in sentences like (a) and strongly suggesting that these inferences are pragmatic. Alexandropoulou et al. () replicated Alexandropoulou’s () results for constructions with at least and found them for more than as well. They further investigated the implicature mechanism involved in the generation of variation effects. Specifically, they tested two hypothetical strategies differing in the type of alternative propositions entertained. Both strategies assume that an expression like at least ten in () conveys a disjunctive choice of values, but they differ in which alternatives make up the disjunction. In the first strategy, illustrated in (), the relevant alternatives are made up by individual numerical values, a strategy similar in spirit to, though differing in its details from, Coppock & Brochhagen (a) and Nouwen (). The output of this strategy, illustrated in () for (), is that not everyone ate the same number of tacos. In other words, this strategy predicts that at least two different values are attested in the situation described by the sentence, but it does not say anything about which values these may be. ()
Everyone ate at least ten tacos.
() Partial variation with infinite disjunctions Assertion
Alternatives
Implicatures
8x½ fx ≥
8x½ fx ¼ 8x½ fx ¼ 8x½ fx ¼ . . . etc.
:8x½ fx ¼ :8x½ fx ¼ :8x½ fx ¼ . . . etc.
It follows that: :∃n8x½ fx ¼ n (partial variation)
The alternative strategy to viewing at least ten as an infinite disjunction of values assumes that at least ten is parallel to the binary disjunction exactly ten or more, similarly to Büring (); Cummins & Katsos (), and Kennedy (), as illustrated in (). The output of this strategy is that someone ate ten tacos and that someone ate more. As such, (), but not (), predicts the inference from () that someone at ten tacos. This difference in prediction allow Alexandropoulou et al. () to test which alternatives modified numerals are associated with. () Partial variation with binary disjunction Assertion
Alternatives
Implicatures
8x½ fx ≥
8x½ fx ¼ 8x½ fx >
:8x½ fx ¼ :8x½ fx ¼
It follows that: ∃x½ fx ¼ 10 ∃x½ fx > 10 (partial variation)
Using the same methodology as Alexandropoulou (), Alexandropoulou et al. () had participants read dialogues between a researcher and an interviewer, the latter following up on the researcher’s statement with a question. Participants were asked to evaluate whether the question made sense given the preceding statement. In order to test whether participants utilized the strategy in () or in (), they compared a condition in which the interviewer’s question included the smallest number asserted in the researcher’s assertion (n in at least n and n+ in more than n) to a condition in which the number was higher than the one asserted. ()
Smallest number condition: During the event all streets were guarded by at least four / more than three policemen. Interviewer: How did you find out that there was a street guarded by four policemen?
()
Higher number condition: During the event all streets were guarded by at least four / more than three policemen. Interviewer: How did you find out that there was a street guarded by five policemen?
Alexandropoulou et al. () found that follow-up questions with the lowest number compatible with the modified numeral scored better than semantic contradictions and worse than the non-contradictory controls (as in Alexandropoulou, ), and, most importantly, they were rated higher than the questions with higher numbers both in utterances with at least and more than—though the ratings for at least were higher than those for more than. These results show that the strategy in () is the more likely one for variation inferences and thus lend support to the accounts that predict it, such as Büring
, ,
(); Cummins & Katsos (), and Kennedy (). This family of analyses, however, predict these inferences for at least only, to the exclusion of Mayr (), who predicts them for more than, too. This series of experiments is an example of how experimental studies can shed light on competing theoretical analyses, which may sometimes differ drastically in terms of the assumptions they make but lead to very subtly different predictions. The results of these studies did not only help adjudicate between the analyses but also led to additional observations. In contrast with previously detected differences between at least n and more than n with respect to obligatory ignorance inferences in unembedded contexts, under universal nominal quantifiers, these two numeral modifiers both give rise to variation effects, but do so in varying degrees.
. D/
.................................................................................................................................. The theoretical and experimental crossover has proven beneficial to our understanding of the semantic-pragmatic sources of ignorance inferences and variation effects of numeral modifiers. There is now a large set of results pointing to the pragmatic, rather than semantic, nature of ignorance inferences. Moreover, by looking at finer-grained properties of ignorance and variation effects, it is possible to draw conclusions on what that pragmatic mechanism involves in more detail. Experiments have also allowed the discovery of factors that are not components of any of the standard theoretical accounts. One example we discussed earlier is how ignorance inferences are dependent on the QUD. The experimental route has also posed some interesting challenges. We would like to briefly discuss two that are salient in light of the experiments discussed earlier, one on the experimental side of things and one on the theoretical side.
.. Probing pragmatics The first challenge stems from the subtlety of experimental results, when it comes to distinguishing semantic and pragmatic content. The methodologies developed for investigating the meaning of modified numerals are for the most part adopted from psychological studies and adapted to the study of linguistic meaning. The interpretation of behavioural data as (a way of) reflecting how speakers interpret propositions in context and make additional inferences is a challenging pursuit. There are two sides to this. First of all, quite a few of the studies discussed earlier use Likert scales to obtain coherence / naturalness scores for combinations of sentences or combinations of a sentence and a context (Cummins & Katsos, ; Westera & Brasoveanu, ; McNabb & Penka, ; Alexandropoulou et al., ; Alexandropoulou, ). As we explained earlier, following Katsos & Bishop (), the idea is that logical contradictions can be identified by scores clustering at the bottom of the scale, logical consistencies can be identified by clusters at the top and combinations that are semantically compatible, but not pragmatically compatible, can be identified by scores that fall between these extremes. In practice, many studies find not just that combinations that are pragmatically odd are associated with the middle region of the scale, but also that
they come with much more variance. This poses a danger of overinterpretation, since the pragmatic realm is associated with a very large grey area. Any response pattern that is not absolutely contradictory or unequivocally coherent is going to be classed as pragmatic in nature. This ties in with a general issue: few theoretical semantic and pragmatic analyses make explicit predictions regarding the likelihood of inferences, implicature strength or processing (though see Hackl, ; Panizza et al., ; among others). A second issue has to do with the inherent subtlety of pragmatic phenomena. While most experimental research is on scalar expressions like some and their relatively straightforward quantity implicatures (cf. Skordos & Barner and Breheny, Chapters and , respectively, in this volume), the kind of inferences associated with modified numerals have properties that are not necessarily well-understood. The extent to which ignorance inferences are defeasible, for instance, is part of an ongoing debate (Aloni & Port, ). Moreover, the experiments in Chemla & Bott () suggest that the free choice inferences that come with disjunction have a processing profile that is different from that of runof-the-mill quantity implicatures. If there is indeed a parallel between disjunction and modified numerals, as the dominant theories suggest, we may not expect all inferences associated with at least and its kin to compare directly with findings about, say, the implicatures of some. In sum, probing pragmatic inferences is made difficult by variation in the prominence of the inferences. In turn, this makes the connection to theory difficult, since the dominant pragmatic frameworks simply distinguish defeasible inferences, which are pragmatic in nature, from non-defeasible ones, which are semantic. Such frameworks do not factor in anything like the likelihood of the inference. As such, the findings discussed in this chapter are in line with an emerging realization that, at least from the perspective of performance, not all implicatures are alike (Doran, Baker et al., ; Doran, Ward et al., ; Van Tiel et al., ).
.. The comparative–superlative (A–B) distinction revisited While there has been an incredible amount of work on modified numerals in the past decade, much of that work is devoted to explaining the inferences of superlative quantifiers only. The original motivation to study modified numerals, however, sprung from the stark contrast between superlative and comparative quantifiers. While we are definitely closer to understanding why superlative modifiers trigger ignorance inferences, it is not so clear why comparative modifiers do not. Some of the experimental studies discussed earlier add an interesting twist to this. While many studies find a clear contrast between comparative and superlative modifiers, others suggest that the inferences we detect for at least/most may be attested for more/ fewer than. For example, it was shown that ignorance inferences can also arise with comparative modifiers, depending on the context (Westera & Brasoveanu, ), and that variation effects can arise with comparative modifiers, albeit to a lesser degree than with superlative modifiers, contrary to most analyses (Alexandropoulou et al., ). Moreover, other studies unearth unexpected differences between lower-bound and upperbound superlative modifiers (Coppock & Brochhagen, b; McNabb & Penka, )
, ,
that do not lend themselves to analyses that assume a symmetrical treatment of, for example, at least and at most. Speculating somewhat, the experimental route has the potential to provide us with a crucial insight. What we originally thought to be about the absence or presence of inferences turns out to be about something much more subtle. The difference between superlatives and comparatives is due to the degree to which the various inferences are hardwired in the lexical entry. If this kind of finding turns out to be robustly supported by future experimental evidence, then it is up to the theoretical frameworks to make sense of this. What could this hard-wiring be? Could it be a form of conventionalisation or fossilization (cf. Aloni & Port, )? Why is there no cross-linguistic variation in what kind of modifiers end up having stronger inferences? That is, why do, cross-linguistically, superlative modifiers express ignorance inferences much more strongly than comparative modifiers do? As we hope such questions and speculations make clear, the challenges we identified for the experimental investigation of numeral modification directly set the agenda of future challenges for pragmatic-semantic theory.
A We gratefully acknowledge a grant from the European Research Council under the European Union’s Seventh Framework Programme (FP/–) / ERC Grant Agreement no. .
......................................................................................................................
......................................................................................................................
N is a universal phenomenon in human communication. All natural languages have negation, yet it seems to be absent in animal communication systems (Altmann, ). We use negation frequently for a wide range of purposes, including rejection, denial, and talking about non-existence. According to the Oxford English Corpus, ‘not’ is the thirteenth most frequently used word in the English language. If we look at the combined frequency of ‘not’ and ‘no’, negation is in the top ten list. Negation enables us to entertain the truth value of a proposition, which is one of the central aspects of language use (Kant, ). It has long been of interest to logicians and philosophers, and relatively recently to psycholinguists and pragmatists. In terms of the processing of negation, psycholinguistic studies have shown two effects that call for an explanation: the first is the asymmetry between its frequent use in natural language and its apparent processing costs, often reported in psycholinguistic studies; the second is the finding that in early stages of processing, negation seems to be ignored and attention seems to be focused on its positive argument. In terms of pragmatic functions, it has been shown that despite its simple semantic meaning, negation interacts with context to produce rich pragmatic effects. Both negation processing and its pragmatic functions present puzzles that cannot be explained by its semantic function. This chapter reviews experimental research on sentential negation. We first survey studies on negation processing (section . .) and the pragmatic effects of negation (section .). We then discuss theories of negation processing and evaluate them against empirical findings (section .). Finally we wrap up the discussion by looking at what negation processing can tell us about the role of pragmatics in sentence processing in general.
. N : T
.................................................................................................................................. Psycholinguistic research into negation focuses on the asymmetry in processing cost between negative sentences and their positive counterparts. () is an example of a negative
sentence and () is its positive counterpart. Here negation scopes over the embedded positive sentence ‘the door is open’, which is called the ‘positive argument’ or the ‘positive counterpart’ of the negative sentence. ()
The door is not open. > Not (the door is open).
()
The door is open.
Negative sentences are found to be more difficult to process than positive sentences. Negation-related difficulty has been found in paradigms such as sentence completion (e.g. Wason, ), sentence verification (e.g. Clark & Chase, ; Carpenter & Just, ; Mayo et al., ), logical reasoning (e.g. Oaksford & Stenning, ; Evans et al., ; Prado & Noveck, ), and inference drawing (e.g. Hovland & Weiss, ), manifested by longer response times and higher error rates for negated sentences compared to their positive counterparts. Also, the memory of negated information is worse than that of positive material (e.g. Cornish & Wason, ; Howard, ). The difficulty of negation observed in these behavioural studies is also reflected in Event-Related Potential (ERP) and functional Magnetic Resonance Imaging (fMRI) studies (e.g. Reichle et al., ; Tettamanti et al., ; Tomasino et al., ). Psychologist Paul Wason conducted a series of experiments on negation processing in the s and s studying negative sentence verification and completion. In his study (Wason, ), participants were asked to complete sentences such as () to () to make them true or false, or to verify sentences in this form. Though () and () are logically equivalent, and so are () and (), negative items such as () and () induced higher error rates and longer response times compared to positive items such as () and (). ()
. . . is an even number.
()
. . . is not an even number.
()
. . . is an odd number.
()
. . . is not an odd number.
A similar effect has been found in other sentence verification studies. These studies either involve verifying positive and negative sentences against world knowledge, such as ‘an elephant is not a mammal’ (e.g. Wason, ; Wales & Grieve, ; Mckinstry et al., ), or involve verifying sentences against pictures, such as the sentence ‘the dots are not red’ against an image of red or black dots (e.g. Gough, ; Trabasso et al., ; Clark & Chase, ; Carpenter & Just, ). In these studies, negative sentences have longer response times and higher error rates than positive sentences. A neuroimaging study by Carpenter et al. () found that in a sentence-picture verification task, compared to positive counterparts, negative sentences induce increased activation in areas of the brain that are responsible for language comprehension (left posterior temporal gyrus) and for visuospatial processing (left and right parietal regions) (see also Fischler et al., ; Reichle et al., ; Herbert & Kübler, ). Not only does negation make sentence verification difficult, it also makes logical reasoning difficult. One famous negation-related reasoning bias is the ‘matching bias’ (Evans, ,
; Evans et al., ; and similarly Wason & Johnson-Laird, ). Participants evaluate conditional rules such as ‘if there is not an “A” then there is not a “”’ against pairs of letters and numbers. Participants tend to look for elements that have been mentioned in the rule (‘A’ cards and ‘’ cards), even when those elements are negated. They have difficulty detecting that the pair ‘P’ ‘’ falsifies the rule ‘if there is not an “A” then there is a ’. Evans () suggests that the ‘matching bias’ can be explained by interpretation heuristics related to the function of negation, which he called the ‘not-heuristic’. In natural language, a constituent is relevant even when negated. The function of negation is to ‘deny presuppositions rather than to assert new information’ (Evans, : ). When participants read a negative construction they do not automatically construct the complement set. Rather, the positive argument remains the focus of attention. Oaksford and colleagues (Oaksford & Stenning, ; Oaksford & Moussakowski, ) argue that negation does automatically trigger the construction of a contrast set, though the construction can be difficult when a large number of possible states of affairs is involved. Prado & Noveck () tested these two accounts using the conditional rule task. Participants verified sentences about letters and shapes such as ‘if there is an H then there is not a square’ against letters presented in geometric shapes (e.g. an H inside a circle). Participants were faster and more accurate to decide that H-in-a-square contravenes the rule than to decide that H-in-a-circle complies with the rule, suggesting that negation does not automatically prime a search for the alternatives.
. T : W ’
.................................................................................................................................. In addition to the difficulty of negation, another prevalent finding concerns the role of the positive argument in negation processing. Studies have found that, in the early processing stage, negation seems to be ignored and negative sentences seem to be processed as if they were positive (e.g. Fischler et al., ; Hasson & Glucksberg, ; Kaup, Yaxley, et al., ; Lüdtke et al., ; Dale & Duran, ). This effect has been found in probe recognition tasks, in sentence verification, and in ERP studies. A series of studies used a probe recognition task, which tests the activation level of the representation of certain targets. Kaup and colleagues (Kaup, Zwaan, & Lüdtke, ; Kaup, Yaxley, et al., ) found that, shortly after reading a negative sentence, participants simulate the situation consistent with the sentence’s positive argument. For example, milliseconds (ms) after reading ‘the door isn’t open’, participants were faster to respond to an image of an open door than a closed door. This effect was reversed later on (at ,ms latency). Hasson & Glucksberg () studied the comprehension of negated metaphors. They tested the activation level of probe words consistent with the metaphor’s vehicle (e.g. ‘strong’) after reading positive or negative sentences (e.g. my brother is (not) a rock). They found that for positive metaphor, response times to vehicle-consistent words were facilitated at all latencies. For negative metaphors, there was similar facilitation of the positive metaphor meaning at short latencies (at ms and ms delays), suggesting that participants initially represent the positive argument of the negative metaphor, and later suppress it.
In sentence verification, studies found that negation not only makes verification more costly, but can also sometimes interact with sentence truth value: while verifying true positives is always faster than verifying false positives, verifying true negatives (TNs) can be less, equally or more difficult than verifying false negatives (FNs) (Gough, ; Trabasso et al., ; Young & Chase, ; Clark & Chase, ; Carpenter & Just, ; Valle Arroyo, ; Mayo et al., ; Kaup et al., ; Lüdtke et al., ). For example, in Clark & Chase (), sentences such as () to () were shown alongside a picture that makes the sentence true or false, such as a plus above a star. Verifying TNs takes longer than verifying FNs, and this result was interpreted in terms of a strategy based on the truth-functional property of negation: negation reverses the truth value of a proposition. When verifying a negative sentence against a picture, participants first represent and evaluate the truth value of its positive argument, and then reverse the truth value. The initial focus on the positive argument was also found in a mouse-tracking study by Dale & Duran (: Experiment ) and Mckinstry et al. (). ()
The plus is above the star.
(true affirmative (TA))
()
The star is above the plus.
(false affirmative (FA))
()
The star isn’t above the plus.
(TN)
()
The plus isn’t above the star.
(FN)
ERP studies (Fischler et al., ; Lüdtke et al., ) found that certain kinds of TN sentences (e.g. ‘a robin is not a tree’) gave rise to greater N effects than corresponding FN sentences (‘a robin is not a bird’), while the corresponding FA (‘a robin is a tree’) predictably gave rise to greater N effects relative to TA (‘a robin is a bird’). Lüdtke et al. () presented participants with positive and negative sentences such as ‘in front of the tower there is a ghost / there is no ghost’, and then a picture that is either consistent or inconsistent with the sentence meaning at either ms delay or ,ms delay. At ms delay, they found no negation effect. That is, a picture of a ghost in front of a tower is primed equally by a positive and a negative sentence, both eliciting a smaller N than a picture with no ghost. At ,ms delay, pictures that are consistent with the sentence meaning are primed more than inconsistent pictures. These results have been interpreted as showing that processing the positive argument is the first stage of negation processing. While it seems that participants often represent the positive argument during negation processing, some studies show that they do not always do so (e.g. Nieuwland & Kuperberg, ; Tian et al., ; Dale & Duran, ; Orenes et al., ). In the ERP literature, Nieuwland & Kuperberg () found that contextually felicitous TNs do not give rise to a greater N effect compared to either TAs or FNs. When participants read pragmatically licensed negative sentences, such as ‘with proper equipment, scuba-diving isn’t very safe/ dangerous’, words that make the sentence false (‘safe’) elicited a bigger N than words that make the sentence true (‘dangerous’). However, when participants read pragmatically unlicensed negative sentences such as ‘bulletproof vests aren’t very safe/dangerous’, ‘safe’ elicited a smaller N than ‘dangerous’. They concluded that there is no principled obstacle for participants to incorporate negation immediately. With appropriate context, participants do not have to first process the positive argument.
Dale & Duran (: Experiments and ) showed that, the more contextual support that the items have, the less the tendency to first consider the positive argument. Using a mousetracking paradigm, participants read short passages such as ‘ “you want to lift an elephant?” the mother said to her child, “but elephants are not small” ’, and judged the sensibility of the adults’ statement. They found no polarity by truth value interaction: that is to say, true statements were easier to verify than false statements regardless of the sentence polarity. fMRI studies (Tettamanti et al., ) show that, when reading sentences with action verbs (e.g. ‘grip’, ‘clasp’), the motor regions (left-hemispheric action-representation system) were activated during positive sentences but not negative sentences (see also Tomasino et al., ). Tian et al. () used a picture recognition paradigm similar to Kaup, Yaxley, et al. (), but compared the processing of simple versus cleft negation sentences. They found that ms after the presentation of a simple negative sentence (e.g. ‘John hasn’t ironed his shirt’), participants responded faster to a picture that is consistent with the positive argument of negation (an ironed shirt), but that ms after the presentation of a cleft negative sentence (e.g. ‘It is John who hasn’t ironed his shirt’), participants responded faster to a picture that is consistent with the negative sentence meaning (a crumpled shirt). This shows that sentence form can somehow affect how a negative sentence is processed. When they changed the negative sentence form but not the propositional content, participants no longer showed a response advantage for the picture that is consistent with the positive counterpart. Tian et al. () argued that the change of linguistic form to a cleft sentence causes a change of accommodated context. A sentence-picture verification study by Tian & Breheny (under review) tested how picture context affects the verification strategy. Participants verified negative sentences against pictures of one or two objects: for example, the sentence ‘the banana isn’t peeled’ against a picture of a peeled banana (one-object) or a picture of a peeled banana and a whole orange (two-object). The different picture contexts were set up to manipulate the prominent Question Under Discussion (QUD). With one-object pictures, the prominent QUD is positive (whether the banana is peeled ); with two-object pictures, the prominent QUD is negative (which one isn’t peeled ). They found that with pictures of one object, the verification of TN was faster than FN in the early phase of the experiment, but this difference diminished in later trials, suggesting that a strategy in which the positive argument was first evaluated was developed during the course of the experiment; with pictures of two objects, TN was faster than FN throughout the experiment, suggesting that participants did not first consider the positive argument, but inferred the state of affairs consistent with the negative sentence, and verified this inferred state against the picture.
. N
.................................................................................................................................. By now we have seen many different findings suggesting that negation is difficult to process. However, the majority of these studies tested negation processing without context. Many studies show that, with appropriate contextual support, negative sentences are not
difficult to process (e.g. Giora, ; Nieuwland & Kuperberg, ; Dale & Duran, experiment ; Orenes et al., ). Wason () suggested that negation requires a context of ‘plausible denial’. Negation is supported by a context in which a salient expectation can be denied. Specifically, when describing the properties of an object, it is easier to say that an item from a minority set does not have the property of the majority set than the other way round. This was tested in Wason (), where participants were shown an array of eight circles, seven of which were in one colour and one in another. Participants completed sentences in the ‘exceptionality condition’ such as ‘circle No. is (not) . . . ’, or in the ‘ratio condition’ such as ‘exactly one/ seven circle(s) is/are (not) . . . ’. Wason found that sentence completion is faster when the negative sentence describes the exception item in the ‘exceptionality’ condition than in the ‘ratio’ condition. He attributed this difference to the coding stage. When participants coded the array in the ‘exceptionality’ condition, the properties of the special circle are viewed against the similar circles. Therefore the two are reciprocally related. However, in the ‘ratio’ group, participants coded the array as two independent groups. Therefore the special item is not evaluated against the group of the complement set. The fact that negation processing is faster in the ‘exceptionality’ condition than in the ‘ratio’ condition suggests that being a minority or having a unique property doesn’t reduce the processing cost of negation about the item. Rather, it depends on the type of relevance of the corresponding majority set in the context. Glenberg et al. () studied the effect of explicit context for negative sentence processing. They propose that negation often involves extra processing cost due to underinformativeness. Standalone negative sentences are often more ambiguous than their positive counterparts, therefore they are not effective for introducing new information (e.g. being discourse-initial). When negation is supported by context, it can introduce new information by eliminating prior expectations or uncertainties. In this case, the processing time of a negative sentence should be comparable to that of a positive sentence. In their experiment, participants read passages such as (). Results showed that, with non-supporting context, negative critical sentences took significantly longer to read than positives. With supporting context, there was no difference between negatives and positives, suggesting that negative sentences need not be difficult to process when there is contextual support. ()
Introduction: Marcy needed a new couch for her family room. Context: Supporting context: She wasn’t sure if a darkly colored couch would look best or a lighter color. Non-supporting context: She wasn’t sure what kind of material she wanted the couch to be made of. Continuation: She finally picked one out and had it delivered to her home. Critical sentence: Critical positive: The couch was black. It looked very nice in her family room. Critical negative: The couch wasn’t black. That probably would have been too dark.
A similar study was conducted by Lüdtke & Kaup (). They propose that contexts which reduce or eliminate negation cost should either explicitly mention the positive counterpart, or strongly imply a positive expectation. In the experiment, participants
read passages such as () containing a positive or negative sentence preceded by different context sentences. They found that the reading times of the last sentence were in general longer if the sentence was negative. Within the negative condition, reading times were faster when the positive counterpart was explicitly mentioned, as in contexts (a) and (b), than when it was not mentioned, as in context (c). In a second experiment, they found that, when the context strongly implies a positive expectation, the negation cost is completely eliminated. () On her way to the pool, Danielle wondered . . . Context (a): whether the water would be warm. Context (b): whether the water would be warm or cold. Context (c): what the water would be like. The water was (not) warm.
. N
.................................................................................................................................. The pragmatic significance of negation is reflected in at least two areas: () negative utterances have more specific contextual requirements than their positive counterparts. Out-of-context negative sentences warrant more background inferences. () Negation’s contribution to utterance meaning is more than reversing the truth value of a proposition. A doubly negated sentence is not synonymous with its positive version; a negated sentence often doesn’t convey the opposite of its positive counterpart. First, negative utterances have more specific contextual requirements than their positive counterparts; they are rarely used ‘out-of-the-blue’ (i.e. they rarely appear in discourse-initial position). Russell () said that ‘perception only gives rise to a negative judgment when the correlative positive judgment has already been made or considered’. A corpus study by Tottie () shows that the two main functions of negation are rejection and denial. Rejection can be seen in negative responses to an offer; denials are used to reverse the truth of a proposition. This means that, for negative utterances to be felicitous, the target of rejection or denial has to be in the context. When such information is not in the context, negative utterances warrant rich inferences about the background. Consider this example. There are streets in London with many small hotels. When you walk along such a street, you will see many hotel signs. If among these houses, you see a sign which says ‘This is not a hotel’, you do not just understand that the house is not a hotel, you also infer that many people have asked if it was a hotel or have mistaken it for one. The sign is for people with such a query or assumption. This suggests that accessing specific contextual information and/or drawing inferences about the context is integral to negative utterance interpretation. Second, negation’s contribution to utterance meaning is often not (or not just) reversing the truth value of a proposition. Doubly-negated constructions tend to communicate a weaker scale than their positive counterparts. Horn (: chapter ) discussed several examples: a door mat saying ‘NOT UNWELCOME’ is less welcoming than ‘WELCOME’;
commenting that someone was ‘certainly not unjust’ is less positive than saying she was ‘just’ (Horn, : ). Also, the meaning of negation depends on the frequency/markedness of the predicate under its scope. ‘Not happy’ often communicates the opposite of ‘happy’, that is, ‘sad’, while ‘not sad’ tend to communicate a medial state between happy and sad (see more discussion in Horn, : chapter ). Taking stock, we have surveyed studies on negation processing and pragmatic effects of negation. Processing-wise, many psycholinguistic studies have shown that negative sentences are more costly to process than their positive counterparts. The extra cost manifests in longer reading times, longer response times in sentence verification tasks, higher error rates, and worse performance in logical reasoning when negation is involved. When there is contextual support, the extra costs of negation are reduced or eliminated. In addition, studies show that in the early stages of negation processing, its positive argument is often, but not always represented. Specifically, this happens when simple negative sentences are processed without context, but does not happen when linguistic context makes a negative sentence ‘natural’, or when external (e.g. picture) context changes the relevance of a negative sentence, or when sentence form focuses on a constituent other than negation (e.g. a focus of the subject in cleft sentences). Pragmatically, negation does more than reversing the truth value of a proposition. Negation has stricter context requirements than its positive counterpart, and out-of-the-blue negative constructions warrant rich background inferences. The meaning contribution of negation also depends on the frequency/ markedness of the predicate under its scope. These effects call for a theory of negative sentence processing that can answer these three questions: why are negative sentences difficult to process (at least in psycholinguistic experiments)? Why is the representation of the positive counterpart often but not always involved? How does negation allow us to infer such background information?
. A
..................................................................................................................................
.. Rejection accounts The most popular perspective suggests that negation is difficult to process because we have to represent the positive counterpart first. The meaning of negation is only incorporated in a second step. This idea draws on the function of negation as a truth-functional operator. A sentence such as ‘the door is not open’ has the structure ‘not (the door is open)’. To be able to process the negative sentence, we must first process the positive counterpart. We will call these accounts rejection accounts. Rejection accounts can be roughly divided into propositional models and simulation accounts, differing in the nature of representation that they posit. Propositional models were proposed to account for the polarity by truth-value interaction in sentence verification (Clark & Chase, ; Carpenter & Just, ). These models say that negative sentences are represented in a propositional format where a positive argument is embedded under a
negative operator. For example, the sentence ‘A is not above B’ is represented as ‘Not (above [A, B])’. The content of the positive argument is evaluated first, only after which can the meaning of negation be incorporated. The extra embedding step explains why negative sentences take longer to process than positive ones. A similar model was proposed by Khemlani, Onenes, & Johnson-Laird (), according to which negative sentences are represented by a negative operator that takes the extension of the positive argument, and returns its complement set. Negation-related cost comes not only from the embedding step in the first process, but also from drawing inferences (calculating a complement set) in the second process. This approach thus predicts less processing cost when the complement set contains fewer members. For example, ‘the cat is not dead’ should be processed faster than ‘the cat is not black’. In comparison, the two-step simulation account (Kaup, Zwaan, & Lüdtke, ) agrees with the idea of embedding the positive counterpart, but states that mental representations are perceptual in nature. It follows the embodied language processing view that language comprehension is achieved through the construction of mental simulations (Glenberg, ; Barsalou, ; Glenberg et al., ; Zwaan, ). These simulations of both concrete and abstract concepts are perceptual in nature, and hence they are similar to representations constructed in nonlinguistic cognition, using sensorimotor systems. Kaup and colleagues (Kaup et al., ; Kaup, Yaxley, et al., ; Kaup, Zwaan, & Lüdtke, ) propose that negation is processed by way of two-step simulations. First, comprehenders construct a mental simulation of the positive counterpart; then this simulation is rejected and replaced with a simulation consistent with the sentence meaning. The meaning of negation is encoded by the deviation of the two simulations. When a negative sentence does not specify the actual situation (for example, ‘Susan’s dress is not red’), the second stage simulation will simply contain unspecified properties (in this case representing Susan with a dress of an unspecified colour). Like propositional models, the two-step simulation account explains the processing cost of negation in terms of the extra step of simulation. The representation of the positive counterpart is an obligatory first step. If the representation of the positive counterpart is available prior to encountering negation, all that remains to be done is rejecting the first simulation and simulating the actual situation. The theory thus predicts that, when the prior context contains the positive argument, negative sentences take less time to process. Rejection accounts can explain both the extra cost of negation and why the positive counterpart is represented. However, they are faced with at least two challenges. First, many studies find that representing the positive counterpart is not mandatory, sometimes even when the sentences are presented without context (Tian et al., ). These findings are incompatible with rejection accounts. Second, rejection accounts suggest that the meaning of negation cannot be incorporated incrementally. This implication seems at odds with abundant evidence for incremental language processing. Research has found that comprehenders activate linguistic and even pragmatic information as soon as cues are encountered, and use such information to form predictions incrementally. For example, as soon as hearing ‘the boy will eat . . . ’, but not ‘the boy will move . . . ’, participants predict words for food (Altmann & Kamide, ). Recently, research into the on-line integration of pragmatic information has also grown. We can integrate common ground and speaker’s epistemic state at the earliest moment and use such information to predict upcoming referents (Heller et al., ; Breheny et al., ); we can access scalar implicatures on-line
with little or no delay, especially with contextual support (Breheny et al., ; Grodner et al., ; but see Huang & Snedeker, b); we infer information about the speaker, using accents and cultural heuristics, and anticipate upcoming words in a sentence (Van Berkum et al., ). Pragmatic information interacts with linguistic information during on-line sentence processing. In light of the findings on incremental processing, rejection accounts must explain why negation is an exception.
.. Contextual approach The contextual approach focuses on the importance of context for negation processing (Wason, ; Givón, ; Leech, ; Horn, ; Nieuwland & Kuperberg, ). The main idea of this approach is that, without context, negative utterances have two pragmatic drawbacks: infelicity and under-informativeness. The infelicity of negation out of context was discussed by Wason (), who argues that negative utterances are often used to deny or contradict a positive proposition. Therefore, negative sentences out of appropriate context are often infelicitous, and therefore hard to process. A similar idea was voiced by Horn (), appealing to Gricean maxims. Philosopher Paul Grice introduced four maxims that govern our conversational behaviour: Quality, Quantity, Relevance and Manner (Grice, ). Horn () suggests that a negative sentence is relevant to the consideration of its positive counterpart. When the positive counterpart is not in the context, uttering a negative sentence violates the maxim of Relevance. A second pragmatic effect discussed by contextualists (Givón, ; Leech, ; Horn, ) is that standalone negative sentences are often less informative than their positive counterparts. The sentence ‘the girl’s dress is not red’ has indeterminate verification conditions. Assuming that the speaker is being cooperative when uttering a negative sentence, the comprehender must draw an inference which justifies the apparent violation of the Gricean maxim ‘be informative’ (Grice, ). However, if being under-informative is a cause for the extra processing cost, it shouldn’t apply to negative sentences with a binary predicate. The sentence ‘the fish is not dead’ is just as informative as ‘the fish is alive’. Current contextual approaches explain the difficulty of negation in terms of its requirement for specific context. While this line of thought seems plausible, as well as being empirically supported, it raises a number of important questions that are yet to be answered. First, negation is not alone in requiring special contextual conditions for its appropriate use. It is widely agreed that virtually every utterance contains elements that require some kind of contextual completion for their full interpretation—for instance, anaphoric or pronominal elements, tense, quantifiers and so forth. Moreover, some positive sentences require a ‘context of plausible assertion’ just as much as negative sentences require a ‘context of plausible denial’. That is, ‘the door is open’, just like ‘the door is not closed’, is typically produced only when the state of the door is at issue. So the first question is, what is it about the contextual demands of negative sentences that make them particularly difficult to process when presented out of the blue, compared to their positive counterparts? Second, as mentioned earlier, it is found that, when we process negative sentences, the positive argument of a negative sentence is often represented during the early processing stages. No contextual account so far explains why, when lacking contextual support, the positive argument is represented in the first place.
.. Dynamic pragmatic account The dynamic pragmatic account (Tian et al., , ) is a contextual theory that specifies exactly how negation interacts with context, whether an explicit context was provided or not. Following on from the observations from the contextual approach, it seems clear that there are differences in how negative sentences interact with context compared to positive sentences. The dynamic pragmatic account argues that the processing effects have a pragmatic cause, even though, in the majority of the relevant psycholinguistic studies, no context was explicitly provided. The dynamic pragmatic account brings together two well-established assumptions in language research. From formal semantics/pragmatics, we assume that language use functions to update a shared information state, called a ‘context’. Thus meaning, in a more extended sense, is a function from input states to output states (Stalnaker, ; Kamp, a; Sperber & Wilson, /). Furthermore, the states that are updated contain not only information that is taken for granted (common ground), but also information about the interaction itself—the current goals of the interlocutors, what issues are under discussion and so forth. Interlocutors aim to achieve relevance and this is typically modelled in terms of a notion of QUD (Ginzburg, ; Roberts, a). Thus, it is useful to describe a speaker’s aims in terms of addressing a prominent QUD. A second assumption of the dynamic pragmatic account of negation comes from language processing research—that interpretation is incremental (Altmann & Steedman, ; Kamide et al., ; Van Berkum et al., ). This means that tentative decisions about utterance interpretation are made as the utterance itself proceeds. Putting dynamic pragmatics and incrementality together, the dynamic pragmatic approach assumes that incremental language processing systems make decisions not only about the semantic content of an utterance, but also about the likely source of relevance, or QUD, for an utterance. Clifton & Frazier () demonstrate that QUD mismatch interferes with on-line reading processes (see also Clifton & Frazier, ). Decisions about the intended update often involve accommodation of context—both of background assumptions and of QUDs (Stalnaker, ; Ginzburg, ; Roberts, a). Thus participants presented with a context-less sentence will make some inferences about the likely intended source of relevance (or QUD) as well as likely presuppositions or background information that such a sentence normally would carry. The dynamic pragmatic approach argues that negation is a cue for retrieving a prominent positive QUD. Without other cues, the most prominent QUD for a negative sentence ‘not p’ is the positive polar question whether p. This is due to the most frequent uses of negation (rejection and denial), but it is sensitive to other factors (e.g. frequency of the predicate and context) and other QUD cues (e.g. prosodic focus and cleft construction). Processing a negative sentence without context triggers the accommodation of a positive QUD, which happens automatically and incrementally. That is, in the absence of other cues, upon hearing a negative sentence such as ‘John hasn’t . . . ’, we tend to activate the prominent QUD is ‘whether John has . . . ’. This form of on-line QUD accommodation contributes to the extra processing cost associated with negation; it explains why the positive counterpart is often represented; and it also explains why contextual support reduces or eliminates both the extra cost and the positive representation. A welcome
consequence of this explanation is that the meaning of negation can be incorporated incrementally, rather than after the positive argument is first processed. This prediction fits nicely with the growing evidence for incremental sentence processing.
.. Evaluating theories of negation processing against empirical findings Before presenting theories of negation processing, we summarized three key empirical findings: (i) negative sentences are often reported to be more difficult to process; (ii) in early processing stages, the positive argument is often but not always represented; (iii) contextual support can eliminate the above two effects, and out-of-context negative utterances warrant rich pragmatic inferences. Rejection accounts can fully explain (i) and partially explain (ii) and (iii), but they cannot explain why the positive argument is not always represented. General contextual accounts can explain (i) and (iii), but do not explicitly explain why the positive argument is sometimes represented. The QUD accommodation account can explain these effects: without context support, sentential negation ¬p triggers the accommodation of a positive QUD whether p. However, what QUD gets accommodated is influenced by top-down contextual and linguistic cues (such as visual context, the presence of a cleft construction, and the prosodic focus of the constituent other than negation). The conflicting representation of the positive QUD and the negative sentence meaning contributes to the extra processing cost. The accommodated QUD gives rise to contextual inferences. What about the time course of negation processing? Rejection accounts predict that the activation of the positive argument is short-lived, and negation cannot be incorporated until the positive argument is processed. In comparison, the QUD accommodation account predicts incremental incorporation of negation, and is noncommittal about how long the accommodated positive QUD can be active. Autry & Levine () tested how contexts affect the activation of the negated concept (the positive argument). They found that, when there is no contextual support, the negated concept (the positive argument) is activated for a prolonged period of time. This result is inconsistent with the rejection accounts, according to which the representation of the positive argument is a first and short-lived step. However, the QUD accommodation account can explain this result. When negation is not licensed, participants accommodate a positive QUD, and they can reasonably expect this QUD to be the topic of the upcoming discourse. Tian et al. () tested when the positive argument is represented, and when the meaning of negation is integrated, during out-of-context negative sentence processing. In a visual world paradigm, participants listened to positive and negative sentences in simple or cleft forms (e.g. [It is] Matt [who] hasn’t shut his dad’s window), while looking at scenes containing a target and a competitor (where the target matches the implied shape of the noun, while the competitor mismatches it). They found that, in the simple but not the cleft condition, there is a difference between negatives and positives: shortly after the verb, there are more looks to the competitor in the simple negatives than the positives. They concluded that the representation of the positive is not a mandatory first step of negation processing. Rather, both sentence content and the contextual source of relevance (QUD) are processed incrementally.
.. General implications and future directions Research on negative sentence processing has both methodological and theoretical implications for the study of sentence/utterance comprehension in general. Methodologically, when we study sentence processing using psycholinguistic experiments, we must not overlook the pragmatic effects associated with these sentences. Commonly, psycholinguistics research studies the effect discourse context has on language processing by explicitly providing discourse contexts. If no context is provided, contextual effects are often ignored. We have seen that, when negative sentences are presented without context, participants automatically retrieve and accommodate relevant contextual information. Sentence comprehension is never independent from contextual effects. This finding suggests that, when we design psycholinguistic experiments and interpret their results, we must consider the likely pragmatic processes, and how they interact with the processing of the sentences’ linguistically coded meaning. Theoretically, the research on negation processing and the effect of context relates to the broader question of how pragmatic information is incrementally updated during sentence processing. There is growing research in the on-line integration of pragmatic information, such as the speaker’s epistemic state (Heller et al., ; Breheny et al., ) and scalar implicatures (Grodner et al., ; Breheny et al., ; Degen & Tanenhaus, Chapter in this volume). This chapter covers a small sample of research on negative sentence processing. There are many open questions remaining: what (different) effects will we observe in the processing of implicit negation, embedded negation and negative quantifiers? How can different theories of sentential negation deal with these phenomena? What are the pragmatic effects of negation in polar questions? How does negation influence memory? How does it interact with reasoning? How are negative utterances processed by young children? Or by comprehenders with Autism Spectrum Disorder? We hope more researchers join our quest, as we believe understanding negation is an important albeit small piece of understanding the part of cognition that is uniquely human.
. C
.................................................................................................................................. This chapter surveyed studies of negation processing. Out-of-context negative sentences are more difficult to process than positives, and their positive counterparts are often represented in the early stage of processing. With contextual support, these effects diminish or disappear. In addition, negative utterances often lead hearers to infer richer background information than positive utterances. We reviewed three theories of negation: rejection accounts, a general contextual approach, and a dynamic pragmatic account. The dynamic account can explain the processing and pragmatic effects of negation. More broadly, these findings are consistent with the idea that pragmatic information is incrementally updated during sentence processing, and that there is no pragmatics-free comprehension.
......................................................................................................................
......................................................................................................................
. I
.................................................................................................................................. P morphology in English is associated with a ‘more than one’ meaning. For example, a sentence like (a), in contrast to its singular counterpart in (b), is generally interpreted as conveying (). This data point in itself is unproblematic: on the basis of our intuitions about (a), we might simply conclude that the literal meaning contributed by plural morphology precisely excludes singularities among the possible denotations of the predicate, as proposed for instance by Chierchia (). ()
a. Emily fed giraffes. b. Emily fed a giraffe.
() Emily fed more than one giraffe. The well-known puzzle associated with these data, however, is the following: if (a) and () share the same meaning, we should expect the negation of the former to be equivalent to the negation of the latter. But this is not the case; the negation of (a), in (a), is clearly not equivalent to the negation of (), in (). Rather, (a) is typically interpreted along the lines of (b), which corresponds to the negation of the singular counterpart in (b).1 ()
a. Emily didn’t feed giraffes. b. Emily didn’t feed a (single) giraffe.
() Emily didn’t feed more than one giraffe.
1 There is a marked reading of (3a) that makes it equivalent to (4), as supported by the continuation in (i). This generally requires stress on the syllable containing the -s morpheme. We will return to this reading.
(i) Emily didn’t feed giraffes, she fed only one!
The puzzle extends beyond negation, and includes questions such as (), to which we would intuitively respond affirmatively if Emily fed a single giraffe, in contrast to how we would respond to the question in (). () Did Emily feed giraffes? () Did Emily feed more than one giraffe? There are two main kinds of approaches to this puzzle in the literature. One approach argues for an ambiguity in the meaning of the plural, and invokes a principle for selecting between the two possible meanings in different contexts (Farkas & De Swart, ; Grimm, ). An alternative approach posits that the ‘more than one’ meaning of plural morphology comes about as a pragmatic inference; specifically, the multiplicity inference is proposed to arise as an implicature (Sauerland, ; Sauerland et al., ; Spector, ; Zweig, ; Ivlieva, ; Magri, ; Mayr, ). Most experimental work on plurality to date has focused on testing some version of the implicature-based approach. These studies have typically focused on four aspects of the multiplicity inference: (i) its sensitivity to logical polarity, (ii) its context-dependence, (iii) the possibility of its suspension, and (iv) its development in young children. As we will see, while experimental data from adults regarding (i) and (ii) suggest that the multiplicity inference is indeed a scalar implicature, the existing evidence pertaining to (iii), at least prima facie, provides an argument in the opposite direction. We will argue that the developmental data may provide the critical evidence for adjudicating between the two theoretical approaches. In particular, data obtained from preschool-aged children are generally consistent with the predictions of the implicature-based approach, while they require auxiliary assumptions on the ambiguity-based approach. The rest of this chapter is organized as follows. In section ., we briefly outline the implicature approach and its predictions, and review the existing experimental work that has investigated the implicature approach. In section ., we briefly consider two implementations of the alternative ambiguity-based approach and the predictions that they make. In section ., we discuss how the two theoretical approaches contend with the experimental findings to date, in particular focusing on the findings related to monotonicity, suspension and context-dependence, and acquisition. Section . concludes with an outlook for future research.
. T
..................................................................................................................................
.. The implicature approach and its predictions We have already seen that a sentence like (a), repeated from earlier, gives rise to a multiplicity inference that Emily fed more than one giraffe. We can alternatively refer to the strengthened meaning of (a) as that in (b), once the multiplicity inference has been factored in.
()
a. Emily fed giraffes. b. Emily fed more than one giraffe.
We have also seen that this strengthened meaning disappears under negation, as (a) does not seem to convey (b). This behaviour extends beyond the distinction between affirmative and negative declarative sentences to include questions () and antecedents of conditionals (), suggesting sensitivity to a more general distinction between upward-entailing and downward-entailing contexts (though we will discuss some apparent exceptions in section ..). ()
a. Emily didn’t feed giraffes. b. Emily didn’t feed more than one giraffe.
()
Did Emily feed giraffes?
() If Emily fed giraffes, she will get a prize. The implicature approach to the data above starts from the observation that this pattern is not unique to plurals, and is in fact parallel to the pattern of scalar implicatures more generally. For instance, a disjunctive sentence like (a) gives rise to the strengthened exclusive meaning in (b), but the latter disappears under negation, as shown in (ab); under the reading in (b), (a) would be compatible with Emily feeding both an elephant and a giraffe. () a. Emily fed a giraffe or an elephant. b. Emily didn’t feed both a giraffe and an elephant. () a. Emily didn’t feed a giraffe or an elephant. b. It’s not the case that Emily fed a giraffe or an elephant but not both. In response to these data, the most prominent proposal in the literature is to treat the multiplicity inference as a form of implicature. Under this approach, multiplicity inferences are not part of the literal meaning of plural morphology, but arise from the comparison of (a) with (b). That is, (a) simply means that Emily fed one or more giraffes, and it is through the negation of the stronger (b) that we derive the ‘more than one’ meaning. () a. Emily fed giraffes. b. Emily fed exactly one giraffe. When (b) is placed under negation, however, as in (b), its meaning becomes weaker relative to the assertion in (a), and therefore the implicature is not derived. () a. Emily didn’t feed giraffes. b. It’s not the case that Emily fed exactly one giraffe. The story above can be implemented in various ways, and existing proposals in the literature range from treating the multiplicity meaning as a conversational implicature,
arising from a Maximize Presupposition-based comparison with the presuppositional singular (something like (b)) (Sauerland, ; Sauerland et al., ), to treating it as a recursive scalar implicature (Spector, ; Magri, ) or an embedded scalar implicature (Zweig, ; Ivlieva, ; Mayr, ).2 All of these different implementations of the implicature approach predict some sort of uniformity across multiplicity inferences and scalar implicatures, along the lines of (). () Uniformity Prediction: If multiplicity inferences are implicatures, then everything being equal, multiplicity inferences and implicatures should exhibit similar behavioural profiles. Now, given this prediction, it is important to note that adults’ rates of implicature computation have been reported to vary quite widely across different lexical scales (Van Tiel et al., ). For this reason, some of the studies that we will turn to next involve comparisons across different populations, namely children and adults. This is important because, even if adults do not compute multiplicity inferences and classical scalar implicatures at precisely the same rates, we might expect that children initially exhibit similar performance on the two kinds of inferences, or that they will differ from adults in the same manner for the two kinds of inferences. By comparing across two populations, these studies are informative about the relationship between the two inferences, even if adults do not compute them at exactly the same rates.
.. Experimental work We will now turn to a series of experiments that have tested the Uniformity Prediction in (), both with adults and children.
... Sauerland et al. () Sauerland et al. () present one of the first studies comparing adults’ and children’s performance on multiplicity inferences, with the aim of testing whether the latter can be analysed as a form of implicature. If multiplicity inferences are derived as implicatures, one might expect adults and children to treat the two kinds of inferences in a parallel fashion. While there is considerable variability in the results reported across developmental studies of implicature, a fairly robust finding has been that children compute fewer classical scalar implicatures of the kind in (a,c) than adults do (Gualmini et al., , Chierchia et al., , Noveck, , Papafragou & Musolino, , among others; see also Papafragou & Skordos, , and Skordos & Barner, Chapter in this volume, for relevant discussion and overviews). That is, children have been shown to accept sentences such as (a) in situations where the stronger alternative (b) is also true. For the most part, this observation has been taken to indicate that children have failed to compute the implicature in (c). 2
See Stateva et al. () for experiments in Italian designed to test diverging predictions of a Maximize Presupposition-based theory (Sauerland et al., , Sauerland, ) and a higher order implicature theory (Spector, ), concerning the interpretation of plural nominals in the scope of universal and non-monotonic quantifiers.
() a. Emily fed some of the giraffes. b. Emily fed all of the giraffes. c. Emily fed some but not all of the giraffes. More recent developmental research indicates that children can succeed on certain kinds of implicatures, in certain experimental contexts (see, for example, Miller et al., ; Barner et al., ; Katsos & Bishop, ; Stiller et al., ; Tieu et al., ; Singh et al., ; Skordos & Papafragou, ; Tieu, Yatsushiro, et al., ). There is currently debate in the developmental literature as to the underlying source of children’s apparently selective success across different implicatures; there is also disagreement as to whether nonimplicature responses in children reflect a true absence of implicature. These debates are not necessarily important for the comparison we will make; we will simply rely on the generalization that without facilitatory experimental manipulations or specially selected scales, children behave systematically unlike adults in many apparently scalar implicaturebased tasks. If this is so, and if we further assume that multiplicity inferences are implicatures, we are led via Uniformity to the prediction that children should also exhibit difficulty computing multiplicity inferences. In particular, they should compute fewer multiplicity inferences than adults, under the same kinds of conditions in which they compute fewer standard scalar implicatures than adults. Sauerland et al. () designed an experiment that involved a puppet playing the role of an alien asking questions about life on Earth. In particular, the puppet would ask participants polar questions such as (). The expectation was that a reading involving the multiplicity inference (e.g. Does a dog have more than one tail?) would lead to a negative answer, given common knowledge, while a literal interpretation of the question (e.g. Does a dog have one or more tails?) would lead to an affirmative response. () Does a dog have tails? The authors compared () to control questions like (), in which the literal meaning of the plural is incompatible with world knowledge, and with (), in which both the literal meaning of the plural and the multiplicity inference are compatible with world knowledge. () Does a boy have beaks? () Does a cat have feet? The authors tested fourteen children aged years, and found the following pattern. Children were adult-like on the control questions, responding negatively to () and affirmatively to (). However, they overwhelmingly tended to produce affirmative responses to target questions like (), answering ‘yes’ to such questions per cent of the time. This would indicate that children accessed the literal meaning of the question in (), Does a dog have one or more tails? The observed non-adult-like behaviour exhibited by the child participants mirrors exactly the differences that have been reported in previous studies of implicatures, with children tending to accept implicature targets more than adults. Sauerland et al.’s () results therefore provide a first piece of evidence in favour of the implicature approach to plurality.
However, as Sauerland et al. () point out (see also Pearson et al., ; Tieu et al., ), their study has some potential limitations. First, multiplicity inferences are not always computed in polar questions, even for adults. Sauerland et al. () discuss examples like (), drawn from Krifka (), in which a response assuming a multiplicity inference from (a) seems to be infelicitous. () a. Does your office have windows? b. Yes (only one though). c. #No, only one. Sauerland et al. () attribute the difference between a question like (a) and one like their experimental () to whether the person asking the question already knows the answer to the question, with the relevant distinction between true information-seeking questions and exam-type questions. According to the authors, only exam-type questions systematically give rise to multiplicity inferences. As Pearson et al. () point out, however, the experimental setting in Sauerland et al. () was precisely one in which the puppet was asking true information-seeking questions; the puppet was ignorant about life on Earth and wanted to acquire information by asking the participants questions about life on Earth. It is therefore unclear that this distinction is relevant for explaining how the participants behaved in Sauerland et al.’s () study. Moreover, Sauerland et al.’s () experimental stimuli involved generic interpretations, which could have been misinterpreted by participants as containing dependent plurals. For example, children might have interpreted () along the lines of Do dogs have tails?, in which case yes-responses would be entirely reasonable, independently of the children’s ability to compute multiplicity inferences. While the findings of Sauerland et al.’s () study can be taken to provide some initial evidence in favour of an implicature approach, the results are not entirely conclusive for the reasons just discussed.
... Pearson et al. () Pearson et al. () tackle the multiplicity-as-implicature hypothesis from a different angle, designing an experiment that aimed to test whether such inferences are cancellable in upward-entailing contexts. The authors used a Covered Box task, modelled on Huang et al. (). Huang et al. () presented participants with three boxes; the contents of two of the boxes were clearly visible to both the experimenter and the participant, while the contents of the third box were hidden. On each trial, participants were asked to give the experimenter the box that matched the experimenter’s description. As an example, on one target trial, participants would see one box containing one cookie, one box containing three cookies, and a covered box. The experimenter would direct participants to ‘Give me the box with two cookies’. The expectation was that if the exactly-n meaning of the numeral was a cancellable implicature, participants would select the box containing three cookies, thereby satisfying the experimenter’s request. If, on the other hand, the exact meaning was not a cancellable implicature, but rather part of the semantics of the numeral, participants would necessarily have to select the covered box. Huang et al. () found that adults chose the covered box on the numeral target trials, suggesting, at least prima facie, an exact semantics for the numeral. In contrast to the numeral trials, Huang et al. found that adults were more
willing to cancel the scalar implicature of ‘some’, for example selecting the box where Cookie Monster had all of the cookies when asked to give the box where Cookie Monster had ‘some of the cookies’. If the multiplicity inference is an implicature, just like the one arising from the scalar quantifier some, one should expect that in exactly the same experimental setup adults would also cancel the multiplicity inference in order to give the experimenter one of the visible boxes, just as was observed for some in Huang et al.’s () study. This is precisely what Pearson et al. () set out to investigate. They presented participants with three cards at a time, two visible and one face down. On one critical target, for example, the two visible cards showed Big Bird with no kite and Big Bird with just one kite, respectively. Participants then heard the instruction in (). () Point to the card where Big Bird has kites. The expectation was that if participants could cancel the multiplicity inference, interpreting () as Point to the card where Big Bird has one or more kites, they would be happy to select the visible picture where Big Bird had only one kite. On the other hand, if participants interpreted () with the multiplicity inference (i.e. Point to the card where Big Bird has more than one kite), they would have to select the face-down picture. Pearson et al.’s () study consisted of two experiments. In the first experiment, sixteen native speakers of English were presented with sentences like () in target conditions like the one described earlier. Participants displayed an overwhelming preference for the facedown card on target trials, in parallel with Huang et al.’s () results for the numeral targets but not the some targets. Prima facie, such a finding goes against the implicature approach, which would predict that the multiplicity inference should be cancellable in the same contexts in which the some-not-all implicature is cancelled. In a follow-up experiment, Pearson et al. () introduced the term ‘only’ into the test sentences, as in (), and found that participants were much happier to select the visible picture in which Big Bird had exactly one kite. () Point to the card where Big Bird only has kites. The authors speculate that this result may be due to the fact that on a standard semantics of ‘only’, the multiplicity inference is merely presupposed, and not asserted. Even if this is so, we are nevertheless left with a difference that is puzzling on the implicature approach: when presented in the same experimental setup, the implicature of some, but not the multiplicity inference, is easily cancellable—even in the absence of ‘only’. In sum, Pearson et al.’s () results offer somewhat mixed evidence, and do not argue conclusively in favour of the implicature account, due to the observed difference between multiplicity inferences and scalar implicatures with respect to the ease of cancellability.
... Tieu et al. (, , ) Building on Sauerland et al. (), Tieu et al. (, , ) conducted another pair of acquisition experiments aimed at detecting multiplicity inferences in preschool-aged children. These authors set out to test the developmental Uniformity Prediction that arises from
the scalar implicature approach to multiplicity inferences, namely that the pattern of children’s multiplicity inferences should mirror their performance on standard scalar implicatures. On the implicature account, any observed differences in the acquisition profiles of the two inferences would minimally have to be explained via auxiliary assumptions. In order to address some of the potential limitations of the Sauerland et al. () study, Tieu et al. () avoided the use of polar questions and instead tested the interpretation of the plural in affirmative and negative declarative sentences. By manipulating monotonicity, the authors could further test the implicature approach advanced by Spector (), which would predict the disappearance of multiplicity inferences in downward-entailing environments.3 The authors used a Truth-Value Judgement Task (Crain & Thornton, ), in which participants watched a series of short stories and then had to judge a puppet’s descriptions of the stories. A sample target story is provided in (); in this story, the protagonist fed exactly one pig. Participants were expected to accept the accompanying upward-entailing target sentence (a) if they accessed the literal existential meaning, but they were expected to reject the sentence if they accessed the multiplicity inference. The downward-entailing counterpart of (a) can be found in (b) (though the actual stories and characters were different across the upward- and downward-entailing conditions, to keep the task interesting for the child participants). Participants were expected to reject the downward-entailing target in (b) if they accessed the literal meaning of the test sentence, since it was false that Emily didn’t feed any pigs (she fed one). They could, however, accept the sentence if they computed the multiplicity inference locally under negation, giving rise to the meaning Emily didn’t feed more than one pig. () Story: Emily is visiting the pig farm today. It’s lunchtime for the pigs. Emily has an apple, and that’s just enough to feed the first pig! Oh no! What about the other pigs? The farmer says, ‘That’s okay, Emily! I’ll feed the others later!’ So in the end, Emily only fed this pig! E: Hey Ellie, what happened in the story? a. P: Emily fed pigs! Plural UE target b. P: Emily didn’t feed pigs! Plural DE target In their first experiment, Tieu et al. () tested twenty-eight English-speaking - and -year-olds and forty-three adult native speakers of English. The adult participants rejected the upward-entailing targets per cent of the time, and accepted the downward-entailing targets per cent of the time. Given that the computation of the multiplicity inference corresponded to rejection in the upward-entailing condition and to acceptance in the downward-entailing condition, adults computed more multiplicity inferences in upwardentailing than in downward-entailing environments, but also computed multiplicity inferences under negation to some degree.4 Children, on the other hand, computed multiplicity
3
An exception to this pattern could arise through the presumably less preferred computation of the multiplicity inference in the scope of negation, a point we return to later. 4 A less plausible source of the yes-responses to the negative sentences would be a wide scope reading of the bare plural above negation, although bare plurals are generally assumed to resist wide scope (cf. Carlson, ).
inferences only per cent of the time in the upward-entailing condition, and per cent of the time in the downward-entailing condition. In a second experiment, Tieu et al. () conducted a more direct comparison of the plurality inference and classical scalar implicatures by including a some implicature condition. The plurality inference condition was similar in structure to that of the first experiment, with the materials modified slightly to ensure maximal parallelism between the plural and some condition. A plural target can be found in () and a scalar implicature target in (). For both kinds of targets, participants were expected to accept the sentence if they accessed the literal meaning, and to reject the sentence if they computed the relevant inferences in (b) and (b). () Scenario: Zebra is at the orchard. There are four bananas in one tree and four oranges in another tree. Zebra picks one of the four bananas. a. Zebra picked bananas. b. ⇝ Zebra picked more than one banana () Scenario: Lion is at the orchard. There are four oranges and four apples on the ground. Lion carries four of the four apples. a. Lion carried some of the apples. b. ⇝ Lion carried some but not all of the apples Tieu et al. () tested seventeen English-speaking - and -year-olds and twenty-seven adults. Similarly to the first experiment, adults computed more multiplicity inferences in the upward- than in the downward-entailing environment (% vs. %), and children computed far fewer multiplicity inferences than adults regardless of monotonicity (% in the upward-entailing condition vs. % in the downward-entailing condition). The child participants also computed fewer implicatures from some than adults (adults % vs. children %), consistent with the pattern from multiplicity inferences. Children’s responses to the two kinds of inference targets were also significantly correlated. On the whole, the results from the two experiments appear to be consistent with the implicature approach to plurality.
... Patson () Patson (), building on Patson et al. (), conducted two picture-matching task experiments with adult participants, aimed at probing the interpretations of sentences containing plural definites like ‘the leaves’. Participants read sentences that implied either a particular spatial distribution of the objects denoted by the definite description (Experiment ), or different set sizes (Experiment ). For example, in Experiment , ‘the leaves’ could be implied to be spatially gathered, as in (), spatially distributed, as in (), or neutral with respect to spatial distribution (): () The gardener raked up the leaves. () The breeze scattered the leaves. () Thomas was confused about the shape of the leaves.
After reading the test sentence, participants were presented with a picture of the denoted object, and they had to decide if the picture was of an object that had been mentioned in the sentence. Crucially, the picture either matched in spatial distribution, mismatched in spatial distribution, or was neutral with respect to spatial distribution (e.g. a pile of leaves; a group of scattered leaves; a single leaf). Patson observed high accuracy in the picture judgement task, with participants matching the ‘collective’ pictures with the ‘collective’ sentences, and the spatially distributed pictures with the ‘scattered’ sentences. Patson also collected reaction times for participants’ responses, and found that participants were quicker to accept pictures that matched the sentences in terms of spatial distribution than pictures that mismatched the sentences. Crucially, participants were as fast to accept the matching pictures as they were to accept the singular pictures, suggesting the participants had activated the ‘at least one’ interpretation of the plural definite descriptions. In the second experiment, Patson manipulated the implied set size of the objects rather than spatial distribution. The test sentences either implied a small or large set size (e.g. The scientist examined the tray of leaves vs. During autumn the yard is full of leaves) or were neutral with respect to set size (e.g. I like the smell of leaves). The pictures accordingly depicted, for example, a small set of leaves, a large (uncountable) set of leaves, or a single leaf. Patson again observed high accuracy rates across conditions. Crucially, the reaction time results replicated those of the first experiment: participants were as fast to respond to pictures that matched the implied set size as they were to accept the singular picture. These results replicate and extend those in Patson et al. (). The findings suggest that when participants represent plural definite descriptions they also activate the corresponding singular meaning. As Patson argues, such findings are in line with the hypothesis that the plural is number-neutral in its literal meaning, and is associated with a multiplicity inference via an implicature. By contrast, the results are much harder to reconcile with a semantic proposal on which the multiplicity meaning is encoded in the literal meaning of the plural definite description. Had the plural definite descriptions only activated representations of multiple objects, one might have expected participants to be slower to respond to the singular pictures than to the pictures containing multiple objects.
. A-
.................................................................................................................................. Having examined a set of experimental studies that set out to investigate the implicature approach to the ‘more than one’ meaning of plural morphology, we turn next to ambiguity approaches. These approaches essentially posit two possible meanings for the plural, along with some additional principle for selecting between two possible meanings. Compared to the implicature approach, these approaches have been relatively less investigated from an experimental perspective (see Tieu, Bill et al., , for detailed discussion). In the following subsections we present a brief overview of the existing theoretical approaches and their accompanying experimental investigations where applicable.
.. Farkas & De Swart () Farkas & De Swart () propose that plural morphology is polysemous between a weak inclusive meaning and a strong exclusive meaning. A plural sentence like (), repeated from earlier, can thus be associated with two possible readings: a weak reading along the lines of (a) and a strong reading along the lines of (b). () Emily fed giraffes. () a. Emily fed one or more giraffes. b. Emily fed more than one giraffe.
Under negation, the relative strength of the two readings is reversed; the negative () can be associated with the strong (a) or the weak (b). () Emily didn’t feed giraffes. () a. Emily didn’t feed one or more giraffes. b. Emily didn’t feed more than one giraffe.
As we have seen from our previous discussion, people have rather strong tendencies to interpret () as (b) and () as (a). To explain these interpretive preferences, Farkas & De Swart () assume the Strongest Meaning Hypothesis in () (see Dalrymple et al., , Winter, , among others). () The Strongest Meaning Hypothesis for Plurals: for a sentence involving a plural nominal, prefer that interpretation of the plural which leads to the stronger overall interpretation for the sentence as a whole, unless this interpretation conflicts with the context of utterance. With (), the ambiguity approach can capture the monotonicity effect we observe: the principle favours the interpretation (b) for the positive () and (a) for the negative (), as these correspond to the strongest possible interpretations of the respective sentences. Note that the principle in () will also allow for the weaker interpretations to emerge if the stronger ones are in conflict with the context of utterance. For instance, a weak interpretation can be forced by adding a continuation that is in conflict with the strong reading, thereby blocking it, as illustrated in (); here the negated plural receives the weak interpretation along the lines of (b), rather than the strong interpretation in (a). () Emily didn’t feed giraffes, because she fed only one! Like the implicature approach, the ambiguity approach can therefore capture the monotonicity effect and the presence of readings like that in (). That is, the two approaches make the same predictions when it comes to the interpretation of plural morphology in positive and negative sentences, and they likewise make the same prediction regarding the availability of the reading in ().
The two theories diverge, however, in what they predict about the relationship between multiplicity and implicatures. Unlike the implicature approach, the ambiguity approach makes no particular predictions regarding this relationship, since the two phenomena are not seen as related. This relationship between multiplicity and implicatures therefore in principle provides a fruitful avenue for potentially adjudicating between the two approaches. One complication to note, however, is the observation that there is considerable variation in implicature computation rates across scales (Van Tiel et al., ); even if the multiplicity meaning arises as an implicature, we might not expect to observe it occurring at similar rates as standard cases of implicature. A further avenue for investigating the predictions of the theories lies in child language acquisition. As Tieu, Bill, et al. () discuss, the ambiguity approach makes certain predictions about how plurality should be acquired. On this approach, children must not only acquire the two proposed meanings of the plural, they must also be able to make use of the Strongest Meaning principle in an adult-like manner, in order to resolve the ambiguity. Tieu, Bill, et al. () present three possible learning stages during which children might be observed to be non-adult-like in their interpretation of the plural. First, they might only have the strong meaning of the plural, in which case we should expect them to be adult-like on the plural in upward-entailing environments but not in downward-entailing environments. Second, they might only have the weak meaning of the plural, in which case they should be adult-like only in downward-entailing environments. Finally, they might have both meanings for the plural, but fail to deploy the Strongest Meaning principle as adults do. They might instead, for example, be guided by charity, selecting the interpretation that is made true in the context. Summarizing, the ambiguity approach of Farkas & De Swart () and the implicature approach both predict more multiplicity inferences in upward-entailing than in downward-entailing environments, but they differ in how they view the relationship between multiplicity and implicatures. Furthermore, they differ in the predictions that they make for how children and adults should perform on multiplicity inferences (and scalar implicatures). In the next subsection, we turn to a different incarnation of the ambiguity approach, based not on monotonicity but rather on the distinction between particularized and general reference.
.. Grimm () While many of the pre-existing accounts of the plural meaning place importance on monotonicity, Grimm () presents an ambiguity analysis based on the distinction between particularized and general reference. Grimm points out that the inclusive plural meaning can arise in environments like (), which can be shown to be neither strictly downward-entailing (), nor to block scalar implicatures (). () Sherlock Holmes should question local residents to find the thief. ↛ Sherlock Holmes should question groups of two local residents or more () Sherlock Holmes should question local residents. ↛ Sherlock Holmes should question local residents with red hair
() Sherlock Holmes should question three local residents. ⇝ Sherlock Holmes should question exactly three local residents Note that the fact that the multiplicity inference appears to be context-dependent is not per se a challenge to the implicature approach. Indeed, the latter can readily account for context-dependence by appealing to relevance. That is, the data above are challenging only if it can be shown that the exactly-one alternative is relevant in the context against which () is evaluated. Grimm takes a different approach altogether and argues that, rather than downwardentailingness, the relevant property to account for the distribution of the inclusive plural reading is the weakened referential status of the nominal. That is, the inclusive plural reading arises in contexts where the nominal doesn’t presuppose the existence of any particular referents. Specifically, Grimm considers the relevant distinction to be between object-level and concept-level entities. Grimm sets out to experimentally probe the inclusive meaning of plurals in interrogative contexts, by conducting two experiments involving plural nouns in two kinds of contexts. In the first experiment, participants were presented with pictures which clearly instantiated the object under question. For example, participants would see a photo of a woman holding either one mug or two mugs, and would have to answer the yes/no question Is the woman in this picture holding mugs? In the second experiment, participants were presented with ‘rules-and-regulations’ contexts meant to foster generic or kind-level construals (Carlson, ). For example, participants saw questions meant to evoke a corporate environment, such as (). () Did your team terminate projects this fiscal quarter? E : employee’s team has terminated exactly one project this fiscal quarter What answer should the employee give? Yes/No Grimm reports that in the first experiment, participants provided yes-answers per cent of the time when the picture contained multiple objects, but gave yes-answers only per cent of the time when presented with a single object. In contrast, in the second experiment, adults answered yes per cent of the time when the ‘employee facts’ line contained a plural entity like three projects, and crucially still preferred the inclusive interpretation on the critical targets (), responding yes per cent of the time. Grimm’s experimental data reveal that participants disfavour inclusive readings of the plural when considering object-referring uses of plural nouns, but favour inclusive readings in contexts that encourage general reference. To capture the data, Grimm proposes an analysis that integrates Krifka’s (b) analysis of common nouns and Roberts’s () framework for discourse structure involving the Question Under Discussion (QUD). First, Grimm draws on Krifka (b) to provide the two representations needed for the two possible interpretations for a given plural noun: a quantified object reading and an instance of a concept reading (b). () a. Quantified Object: λnλiλx½Ri ðx;DOGÞ ^ OUi ðDOGÞðxÞ ¼ n b. Instances of a Concept: λiλx 9 n½Ri ðx;DOGÞ ^ OUi ðDOGÞðxÞ ¼ n
These two representations then provide two possible interpretations for a given plural noun. The plural marker can be interpreted as a number marker (which Grimm diverges from Krifka in assuming is interpreted as more than one), as in (a); in this case the interpretation is true when the noun’s referent is the type of thing named and has cardinality greater than one. Alternatively, the plural noun can enter the derivation as a kind or concept, whereupon the interpretation succeeds when at least one entity realizes the kind dog, as in (b); in this case, the reading is number-neutral, being true as soon as the noun’s referent is the type of thing named. () a.⟦dogs⟧:= λiλx[Ri (x, DOG) ^ OUi (DOG) (x) 2] b.⟦dogs⟧:= λiλx9n[Ri (x, DOG) ^ OUi (DOG) (x) = n] Next, Grimm situates his analysis within the discourse framework of Roberts (). In essence, distinct QUDs can be derived that query either the existence or the number of objects, respectively: () a. Did Ed see dogs? (Existential QUD) ?(λi9x9n [Ed see x in i ^ Ri (x, DOG) ^ OUi(DOG)(x) = n]) b. Did Ed see dogs? (Quantitative QUD) ?(λi9x [Ed see x in i ^ Ri(x, DOG) ^ OUi(DOG)(x) ]) According to Grimm, the interpretation of the plural noun is then guided by which of (a,b) corresponds to the immediate QUD that the conversational participants are attempting to answer. When reference to particular entities is not yet determined, the Existential QUD is appropriate, licensing an inclusive reading; when reference to particular entities is determined, it is the Quantitative QUD that is the immediate QUD, and the exclusive reading is licensed. This would explain why, in Grimm’s first experiment, where participants were presented with particular objects (i.e. one or more mugs), participants preferred the Quantitative QUD and therefore the exclusive reading of the plural, whereas in the second experiment, where there was no presumption that any projects were completed, participants preferred the Existential QUD and therefore the inclusive reading of the plural.
. G
.................................................................................................................................. The existing experimental investigations into the semantics and pragmatics of plurality yield a number of insights, all the while raising new questions for future research. We now turn to these insights and remaining puzzles.
.. Sensitivity to monotonicity The multiplicity inference appears to be sensitive to monotonicity, as most clearly evidenced by the data in Tieu et al. (, ). The adult participants in these experiments
generally preferred to reject both positive and negative plural descriptions like (a,b) in singular contexts, such as those in which Emily fed exactly one giraffe. Such responses indicate that adults interpreted the plural exclusively in the upward-entailing (a), but inclusively in the downward-entailing (b). () a. Emily fed giraffes. b. Emily didn’t feed giraffes. As we have seen, both the implicature approach and the ambiguity approach of Farkas & De Swart () predict precisely this pattern, the former through implicatures that arise in upward-entailing but not downward-entailing environments, and the latter through a Strongest Meaning principle. What about the ambiguity approach of Grimm ()? On Grimm’s proposal, participants’ interpretation of the plural nouns would have been guided by whether they accessed the Existential QUD or the Quantitative QUD. However, Tieu et al. (, ) used the same kinds of stories for their upward-entailing and downwardentailing targets. Given the stories were the same, presumably the salient QUDs were also equivalent across the two conditions. It is therefore somewhat puzzling that adults, particularly in the second experiment, computed far fewer multiplicity inferences under negation (% vs. %). One would have to say that the positive and negative sentences themselves invoked different QUDs, and therefore different interpretations of the plural, since only the sentences differed between conditions.5 Future research might further investigate the respective roles that the QUD and monotonicity may play in adults’ interpretation of plural morphology, and the interaction between the two factors (if any). For the moment, it is clear that adults distinguish between upward- and downward-entailing contexts when interpreting plural morphology. This sensitivity to polarity is expected under all of the various incarnations of the implicature approach in the literature and on certain implementations of the ambiguity approach.
.. Suspension and context-dependence The theoretical approaches we have discussed are not quite so straightforward to reconcile with the experimental evidence pertaining to suspension and cancellability. First, we have discussed the observation that multiplicity inferences are harder to suspend than regular scalar implicatures, as can be seen by the contrast between () and (). () Some of the students came. In fact, all of them did. () Emily fed giraffes. #In fact, she fed exactly one. This difference is reflected in Pearson et al.’s () results, especially those of their first experiment. While their second experiment reveals that multiplicity inferences are easier to suspend in the scope of a presupposition trigger, the exact nature of the interaction
See Tian & Breheny (Chapter in this volume) for an overview of experimental studies investigating negation processing and the pragmatic effects of negation. 5
between presuppositions and implicatures in such cases is not very clear, rendering it difficult to draw conclusions regarding suspension of multiplicity inferences from this particular data point. The challenge for the implicature approach is to account for the difficulty of suspending multiplicity inferences in cases like () above, while at the same time accounting for their apparent context-dependency in cases like Grimm’s (). More generally, it remains an open question how to account for the observed difference in cancellability between scalar implicatures of terms like some and multiplicity inferences, as well as the role played by presuppositions in apparent cases of suspension. As Pearson et al. () discuss, the observation that multiplicity inferences are difficult to cancel has already been made in the literature. Chierchia et al. (), in particular, observe the oddness of a case like (), and propose that the implicature triggered by () is in fact obligatory. That is, the implicature arising from some in () is context-dependent and easily suspended, while the multiplicity inference arising from the plural in () is not. An outstanding question is how exactly to reconcile this with Grimm’s () data, which suggest context-dependency of the multiplicity inference. As mentioned, Grimm’s () observation that multiplicity inferences are contextdependent is not in and of itself a problem for the implicature approach. The multiplicity inference could be treated as an optional implicature, the presence of which is dependent on the context. In particular, the presence of the implicature would depend on whether the exactly-one alternative is made relevant in the context. In Grimm’s first experiment, this alternative was made relevant in the context, while in the second experiment it was not. In this way, the implicature approach can explain both the sensitivity to polarity and the context-dependency displayed by the multiplicity inference. One would, however, have to work through the details of how to reconcile cases where context-dependence is evident (like Grimm’s ()) with the apparent lack of suspension in cases like (). Notice that the same issue arises on the ambiguity approach of Farkas & De Swart (), which would need to explain why the Strongest Meaning principle appears to be violable in cases like () but not in cases like ().
.. Children’s development of plural meanings The data pertaining to monotonicity, suspension, and context-dependence do not unequivocally argue for one approach to multiplicity over the other. However, considerations of language acquisition and learnability may be able to shed light on the debate. Let us consider how each approach might explain the experimental results we have discussed. We have observed that children do not compute as many multiplicity inferences as adults do (e.g. Sauerland et al., ; Tieu et al., , ). What could be missing, according to the implicature approach? On the implicature analysis, the child must learn the relevant alternatives, contrasting the plural form with some version of the singular; otherwise no inference can arise from the plural. Children who cannot yet compute multiplicity inferences may lack knowledge of or access to the required alternatives. In fact, access to alternatives has been proposed to be a problem for children’s generation of implicatures more generally; in cases where children do have access to the required alternatives, research indicates they are fully capable of generating implicatures and implicature-like meanings
(see, for example, Barner et al., ; Stiller et al., ; Tieu et al., ; and Singh et al., ). On the implicature approach, the multiplicity inference may be particularly challenging for the child learner of English, given that neither the plural morpheme -s nor the (enriched) singular alternative correspond to standalone lexical items that the child can easily contrast with each other (unlike the cases of, for example, ‘some’/‘all’, ‘or’/‘and’, ‘might’/‘must’, and numerals). How might the ambiguity approach of Farkas & De Swart () contend with the developmental data? On the whole, the child data appear to pose a challenge to this approach. While the ambiguity approach is not incompatible with similar developmental profiles for multiplicity inferences and standard implicatures, it would not appear to readily account for children’s behaviour across the polarities, that is, in positive and negative sentences. Recall that on the ambiguity approach, the positive (a) is ambiguous between the readings in (b) and (c), while the negative (a) is ambiguous between the readings in (b) and (c). The Strongest Meaning hypothesis is expected to favour (b) for (a) and (c) for (a). ()
a. Emily fed giraffes. b. Emily fed more than one giraffe. c. Emily fed one or more giraffes.
S
() a. Emily didn’t feed giraffes. b. Emily didn’t feed more than one giraffe. c. Emily didn’t feed one or more giraffes.
W
As we discussed in section .. (following discussion in Tieu, Bill, et al., ), there are three main developmental scenarios that might give rise to non-adult-like behaviour on the ambiguity approach. Children might access only the weak meaning of the plural; they might access only the strong meaning of the plural; or they might have both meanings but fail to apply the Strongest Meaning hypothesis in an adult-like way. Yet what the child data in Tieu et al. (, ) reveal is that children in fact favour the weak reading of the positive (a), but the strong reading of the negative (a). It is unclear how the ambiguity approach might capture this observed pattern across the two polarities. As suggested by Tieu, Bill, et al. (), for example, the ambiguity approach might posit that children engage with the Strongest Meaning principle differently from adults, obeying the principle only in downward-entailing contexts. But while it is plausible that children at a certain age might differ from adults in the deployment of a pragmatic principle like the Strongest Meaning hypothesis, it’s not clear why their application of such a principle would vary systematically with the polarity of the context. Alternatively, the ambiguity approach might explain the data by positing that children in this age range have only mastered the weak inclusive meaning of plural morphology, hence accessing a globally weak reading of the positive (a) but a globally strong reading of the negative (a). But as Tieu, Bill, et al. () discuss, such a scenario raises a potential learnability problem, in the form of a subset problem (Berwick, ; Crain et al., ; Gualmini & Schwarz, ). In particular, if a child has acquired the weak meaning of the plural morpheme, any evidence they encounter for the strong meaning will also be compatible with the weak meaning, given that the strong meaning entails the weak
meaning. What evidence would then prompt the child to move from a weak to a strong meaning for the plural morpheme? Finally, let us consider what the child would have to learn on an account like Grimm’s. One possible explanation for children’s difficulty with multiplicity inferences, under Grimm’s proposal, might lie in a difficulty with selecting the correct QUD, and thereby accessing the target meaning of the plural. In this scenario, what the learner needs to develop is something more general than a particular kind of inference; the child has yet to become adult-like in her management of the discourse structure.6 In the Tieu et al. (, ) experiments, for instance, one might posit that children favoured the Existential QUD, while adults favoured the Quantitative QUD; what would develop over time is an alignment with the adult-like selection of the target QUD.7 Alternatively, as on the ambiguity approach of Farkas & De Swart (), Grimm’s approach might also posit that children simply haven’t yet mastered both meanings or representations of the plural. Given their preference for the inclusive meaning of the plural, one might posit that children acquire the kind reading earlier than the quantified object reading. Future work might more closely investigate the developmental connection between plurality and genericity.8 Summarizing, the existing experimental data from children appear to be more straightforwardly accounted for under the implicature approach. However, as we have discussed, both approaches to multiplicity inferences provide possible avenues for further investigating the developmental path that children may take to an adult-like understanding of plural morphology. Before closing this section, it is worth noting that the acquisition studies by Sauerland et al. () and Tieu et al. (, ) pertained specifically to preschoolers’ knowledge of multiplicity inferences. This age range (– years) is apt for investigating connections between multiplicity inferences and scalar implicatures, as previous studies of implicature have likewise targeted this age range, and shown that preschoolers typically compute fewer standard implicatures than adults do. On the face of it, the findings of Sauerland et al. () and Tieu et al. (, ) appear to support the hypothesis that multiplicity inferences are a form of implicature, with similar developmental profiles. However, the observation that -year-old children may not be fully adult-like in their interpretation of plural morphology is prima facie at odds with the findings of studies conducted with much younger children. In particular, it has been reported that children begin producing plural morphemes at months of age
6 See Zondervan (, , ) for discussion of the role of the QUD in implicature computation, and Gualmini et al. (), Conroy et al. (), and Lewis () for relevant discussion of the role of the QUD in child language experiments. 7 It is worth pointing out, however, that the critical stories in the Tieu et al. experiments were such that the protagonist engaged with a particular entity that was explicitly mentioned; according to Grimm (), this therefore should have activated the Quantitative QUD as the immediate QUD. The justifications that children provided for accepting (the inclusive reading of) the plural in sentences like Emily fed pigs (where Emily fed exactly one pig), do not seem consistent with this (e.g. Because she feed a pig / Because she said the pig has been feeded, and that happened). 8 See Sneed (, ) and Lazaridou-Chatzigoga et al. () for discussion of children’s acquisition of generic noun phrases.
(e.g. Brown, ; Mervis & Johnson, ; Fenson et al., ; Barner et al., ),9 and that -month-olds are sensitive to plural marking in comprehension (Wood et al., ; Davies et al., ). For example, Wood et al. () used a manual search paradigm in which -year-old children were prompted to search for objects in a box. The authors found that the -month-old participants searched longer in the plural condition, suggesting they were expecting to see ‘more than one’ object in such cases. More recent work by Davies et al. () also suggests that English-acquiring -month-olds have an understanding of plural morphology (in particular, of the voiceless /s/ allomorph). Tieu, Bill, et al. () attempt to explain the difference in the findings from -year-olds and -year-olds by identifying differences in the experimental paradigms used to test the two age groups. Studies like Wood et al. () (and the more recent Davies et al., ) allow toddlers to derive a meaning for the linguistic description, and require the toddler to execute an action (or look to a particular object) based on that interpretation. By contrast, the Truth-Value Judgement Tasks in Tieu et al. (, ) presented -year-olds with situations in which only a single object was depicted. The acceptance of a plural description in such a situation doesn’t necessarily mean that children didn’t expect there to be more than one object. The -year-olds may simply have been willing to override their expectation, and accept the use of plural morphology in a context in which they themselves would not have produced the plural. As Tieu, Bill, et al. () point out, this acquisition scenario is quite consistent with a recent proposal regarding children’s so-called pragmatic tolerance in the domain of scalar implicatures (Katsos & Bishop, ). According to Katsos & Bishop (), preschoolers are simply more tolerant and charitable than adults when forced to accept or reject under-informative scalar descriptions.
. C
.................................................................................................................................. In this chapter, we have presented an overview of recent experimental work investigating the semantics and pragmatics of plurality. These studies have examined the ‘more than one’ meaning that is seemingly contributed by plural morphology, referred to in the literature as a plurality or a multiplicity inference. As we have discussed, theoretical treatments of the multiplicity inference include deriving the ‘more than one’ meaning as a kind of implicature (Sauerland, ; Sauerland et al., ; Spector, ; Zweig, ; Ivlieva, ; Magri, ; Mayr, ), or treating the plural as ambiguous between two possible meanings (Farkas & De Swart, ; Grimm, ). Recent findings from experiments conducted with children and adults reveal that the plural meaning is sensitive to monotonicity, context-dependent, suspendable under certain conditions, and computed by preschoolaged children on a par with standard implicatures (Sauerland et al., ; Pearson et al., ; Tieu et al., , ; Grimm, ). As we have discussed, the implicature approach to multiplicity inferences, in all of its different incarnations, is largely consistent with the
9 There are also reports, however, that children do not always produce the plural in required contexts (for discussion, see Berko, ; Cazden, ; Brown, ; Winitz et al., ; Mervis & Johnson, ; Marcus et al., ; Marchman et al., ; Zapf & Smith, ).
experimental evidence accumulated in the literature in recent years, in particular that arising from the comparison between preschool-aged children and adults. This body of experimental work, however, has also introduced certain challenges, arising in particular from the way these inferences can be suspended, their apparent context-dependency, and the relationship between multiplicity inferences and presuppositions. In addition, while some initial attempts to compare multiplicity inferences and classical scalar implicatures have suggested that the two exhibit similar behavioural profiles, more work is needed in order to understand the exact relationship between multiplicity inferences and standard cases of implicature, in both the developing and adult grammar.
......................................................................................................................
......................................................................................................................
č
. I: W ?
.................................................................................................................................. Sentences with multiple quantifiers (expressions like every cake, each man, a boy) are ambiguous in ways in which sentences with one quantifier are not. Consider (), for example: (a) makes a statement about one individual boy, introduced by the quantifier a boy; in contrast, (b) makes a statement about potentially many boys, one for each girl, because the quantifier a boy now interacts with the other quantifier in the clause, each girl. ()
a. Mary kissed a boy. b. Each girl kissed a boy.
We say that (b) receives two different interpretations depending on which quantifier takes scope over the other. If each girl takes scope over a boy, the resulting interpretation is: each girl is such that she kissed some boy or other; the boys are possibly different from girl to girl. If a boy takes scope over each girl, the resulting interpretation is: there is one boy such that each girl kissed him; all the girls kissed the same boy. Sentences like () with multiple quantifiers were difficult to formalize: they did not receive a proper account in mathematical logic until Frege ().1 Multiple quantifiers are similarly challenging for theories of human reasoning: it is not obvious how the human mind represents them and their multiplicity of scope (see, for example, JohnsonLaird et al., ).
1
For an introduction to the treatment of multiple quantifiers in mathematical logic and semantics, see Gamut () and Heim & Kratzer () among others.
. F
.................................................................................................................................. In addition to the intrinsic difficulty of formalizing quantifier scope ambiguities and identifying cognitively plausible ways of representing them, explicitly mapping the cognitive consequences of scopal interactions between multiple quantifiers is non-trivial because of the variety of factors affecting scope. The factors influencing quantifier scope can be grouped into three classes: • syntactic information—Ioup (); Vanlehn (); Micham et al. (); Gil (); Gillen (); Kurtzman & MacDonald (); Villalta (); Anderson (); Radó & Bott () • linear information—Fodor (); Gillen (); Kurtzman & MacDonald (); AnderBois et al. () • lexical information—Ioup (); Micham et al. (); Gillen (); Kurtzman & MacDonald (); Tunstall (); AnderBois et al. (); Dwivedi (); Brasoveanu & Dotlačil (b). These factors, or closely related ones, have also been identified as crucial scope disambiguation predictors in the computational linguistics literature on quantifier scope (Poesio, ; Saba & Corriveau, ; Higgins & Sadock, ; Srinivasan & Yates, ; Evang & Bos, ; Manshadi, ; among others).2 Let us show how the factors influencing quantifier scope are deployed with an example. Consider the sentence in (), which includes two quantifiers, a caregiver and every night. ()
A caregiver comforted Mary every night.
How can the human processor/interpreter identify the scope of these quantifiers? One option is that the interpreter consults syntactic information. Apart from the references discussed earlier, this position has often been defended in theoretical linguistics work (Reinhart, ; May, ; Heim & Kratzer, ). A common assumption shared by this family of approaches is that scope can be read off the syntactic positions in which the quantifiers appear: for example, Reinhart argues that a quantifier A scopes over a quantifier B iff A c-commands B. For ‘canonical’ sentences—that is, sentences without transformations—we can equivalently make use of the grammatical role hierarchy shown in (): a quantifier taking a particular role scopes over every argument which appears to the right of that role. 2
The same factors have also been used to explain interactions of scope-bearing elements other than quantifiers, for example, negation and modal verbs. In this chapter, we only focus on quantifier scope. For a discussion of negation, see (Tian & Breheny and Xiang, Chapters 12 and 26, respectively, this volume).
()
č Grammatical-Role/Syntactic hierarchy: Subject ≫ Indirect Object ≫ Object
The hierarchy in () says that an indirect-object quantifier has an object quantifier—but not a subject quantifier—in its scope, while a subject quantifier has every argument in its scope. The scope position of a temporal modifier like every night in () depends on what one thinks about the syntax of such modifiers. In some accounts (e.g. Larson, ; following McConnell-Ginet, ), they are below objects in the syntactic structure; in others (e.g. Nilsen, ), they are higher than objects but below subjects. Either way, it follows that the subject quantifier a caregiver has the temporal modifier every night in its scope, which yields the following reading: there is one caregiver such that, every night, Mary was comforted by that caregiver. Another possibility is that linear order resolves scope. In that case, the scope of a quantifier is equivalent to the material that appears to the right of the quantifier. Finally, it has been argued that the lexical identity of the quantifier influences its scope. Commonly, only the role of determiners is investigated in this approach (but see Dwivedi, ). For instance, Ioup () postulates a scope hierarchy among quantifiers according to which the distributive quantifiers each and every should take quantifiers like most or indefinites in their scope but not vice versa. Using only this hierarchy, we would conclude that, in (), the temporal modifier has the subject in its scope, yielding the interpretation that every night, there is a possibly different caregiver who comforted Mary. The lexical role of determiners in determining scope has also been studied in theoretical linguistics: Szabolcsi () contains very detailed and careful discussions and is the classical reference on this topic. As has been known at least since Montague (), considering only one hierarchy is not sufficient to derive all the available interpretations for sentences with multiple quantifiers. For example, in (), every night is an adjunct following the rest of the clause. Thus, from the perspective of several hierarchies, it should not be able to take scope over the subject or the object. But although this reading is dispreferred, it is possible to understand () as introducing multiple caregivers into discourse, one per night—a reading in which the indefinite a caregiver is interpreted in the scope of the distributive quantifier every night (Ioup, ; Gil, ; Tunstall, ; Anderson, ; AnderBois et al., ; Dotlačil & Brasoveanu, ; among others). The presence of two readings for () could be explained by assuming that one hierarchy, for example, linear order, establishes the preferred interpretation, and another hierarchy, for example, the quantifier hierarchy, establishes a dispreferred but possible interpretation. However, it is common to explain both readings using just one hierarchy type. Then, two readings are derived by postulating that any hierarchy must have a certain amount of flexibility in it. For instance, much of the linguistic literature on the syntax and semantics of quantifier scope postulates that, in a range of cases, Adjuncts can be promoted to the first position in the hierarchy, thereby taking scope over Subjects. This promotion is labelled inverse scope because the covert scope representation is the inverse of the overt/ ‘surface’ one. ()
Marked hierarchy for a particular instantiation of inverse scope: Adjunct ≫ Subject ≫ Object ≫ . . .
A similar point has been made about direct object quantifiers: they can take inverse scope over subject quantifiers but, as for adjuncts, this is a marked option. For example, the sentence in () below has an unmarked surface-scope reading according to which a single boy climbed every tree, and a marked inverse-scope reading in which the direct object quantifier takes wide scope to the effect that for every tree, some boy or other (possibly different from tree to tree) climbed it. ()
A boy climbed every tree.
() Marked hierarchy for another instantiation of inverse scope: Object ≫ Subject ≫ Adjunct ≫ . . .
. T : E
.................................................................................................................................. Suppose that () above is followed by another sentence that starts either as shown in (a), or as in (b). Previous psycholinguistic research has established that (b) leads to processing difficulties compared to (a). The same point can be made about the sentence in () above and the two possible continuations in (a) and (b) below: according to the same kind of processing difficulty measures, (b) is hard relative to (a). ()
a. The caregiver . . . b. The caregivers . . .
[easy] [hard]
()
a. The boy . . . b. The boys . . .
[easy] [hard]
The behavioural measures showing that plural continuations yield more processing difficulty than singular continuations include: preference for singular continuations in forced choice or Likert-scale judgement tasks; increased reading times for plural continuations in self-paced reading studies (Tunstall, ; Anderson, ); and more regressions and increased reading times for plural continuations in eye-tracking studies (Filik et al., ; Paterson et al., ; Dotlačil & Brasoveanu, ). We turn now to reviewing these experimental findings in some detail. Tunstall (: chapter ) reports the results of a word-by-word self-paced reading task coupled with a stops-making-sense judgement task with a design based on Kurtzman & MacDonald (). The items consisted of a dative sentence with a universal and an indefinite in direct object and indirect object positions, followed by a continuation sentence intended to be consistent only with one possible scoping of the previous sentence. An example item is provided in () below (see Tunstall, : , example ()). () a. Kelly showed a photo to every critic last month. i. The photo was of a run-down building. ii. The photos were of a run-down building.
č b. Kelly showed every photo to a critic last month. i. The critic was from a major gallery. ii. The critics were from a major gallery.
The condition in (a) has an indefinite in direct object position and a universal in indirect object position. The condition in (b) switches the two determiners (universal in direct object position, indefinite in indirect object position) while keeping everything else in the sentence the same. The continuations contain a definite in subject position that is always anaphoric to the indefinite Noun Phrase (NP) in the previous sentence. The plural number morphology on this anaphoric definite in (a-ii) and (b-ii) disambiguates the quantifier scoping in the preceding sentence: it is compatible only with a narrow-scope indefinite, which is the inverse-scope reading of (a) and the surface-scope reading of (b). In the critical three-word region consisting of the definite and the verb in the continuation sentence, Tunstall (: et seqq.) found that number had a reliable effect in the a-every quantifier order ((a) above), with plural taking longer than singular. This was as predicted by Tunstall’s Principle of Scope Interpretation (PSI; Tunstall, : , ()), which hypothesizes that quantifier scope is read off the Logical Form (LF) of a sentence, and the default scoping is based on the LF structure that is minimally different from its S(urface)structure (for reasons of general processing economy). In contrast, number did not have a significant effect in the every-a quantifier order ((b) above), which Tunstall attributes to the fact that the singular continuation is in fact compatible with the universal taking scope over the indefinite—we just happen to choose the same critic over and over again for every photo (her Vagueness Principle: Tunstall, : , ()). Anderson (: chapter ) reports the results of several off-line (questionnaire) and realtime (self-paced reading) studies showing that inverse scope is costly, both for sentences in isolation and for sentences occurring in a context that supports the inverse-scope interpretation. An example item used in a questionnaire study is the following (from Anderson, : , ()): participants were shown a sentence like () (the subject was always an a-indefinite and the object an every-universal) and were asked to choose the paraphrase that corresponded to their initial interpretation out of two possible paraphrases (a) and (b). () A cashier greeted every customer. a. One cashier greeted customers. b. Several cashiers greeted customers. Anderson found that in such out-of-the-blue cases, the paraphrase that is compatible with surface scope (a) is selected per cent of the time. She then proceeded to embed these doubly-quantified sentences in contexts that would favour one reading or the other, for example (see Anderson, : et seqq.), for example: () a. Context supporting surface scope: The members of the gourmet club decided to put out a cookbook of their favourite recipes. They wanted the recipes to be easy enough for an inexperienced cook. The president of the club requested that a volunteer test the recipes to make sure that the instructions were correct. After a short discussion, a member of the club tested every recipe.
b. Context supporting inverse scope: The members of the gourmet club decided to put out a cookbook of their favourite recipes. They wanted the recipes to be easy enough for an inexperienced cook. Members who nominated recipes were required to test the recipes to make sure that the instructions were correct. A member of the club tested every recipe. Each paragraph was followed by a how many comprehension question targeting the indefinite in subject position and asking participants to choose whether one vs. several entities are contributed by that indefinite, for example: How many members tested recipes? Participants were asked to identify the answer that matched their first interpretation of the sentence. The surface-scope interpretation was selected per cent of the time in surfacescope biased contexts—not significantly different from the percentage observed for sentences in isolation—but the inverse-scope interpretation was selected per cent of the time in the inverse-scope biased contexts. Thus, we see that the surface-scope biased context does not really boost the already large default preference for surface scope, but the inverse-scope biased contexts do manage to raise the preference for inverse scope from about per cent to about per cent. Anderson also investigated sentences with a every-universal in subject position and an a-indefinite in object position, exemplified in () below. The paraphrase in (a) was only compatible with the surface-scope reading, while the paraphrase in (b) was taken by Anderson to be compatible with the inverse-scope reading.3 Anderson observed that the interpretation compatible with the inverse-scope reading, (b), was dispreferred; this was true even in inverse-scope biased contexts (Anderson, : et seqq.; judgements on a five-point Likert scale). ()
Every player rubbed a lucky charm. a. Each player rubbed a different charm. b. All the players rubbed the same charm.
The same kind of two-quantifier sentences (in both the a-every and the every-a order) were investigated with self-paced reading tasks involving two-sentence items of the (by now) familiar kind (Anderson, : et seqq.): the scopally ambiguous sentence was presented first, followed by a continuation with a definite in subject position anaphoric to the indefinite in the previous sentence. Just as in Tunstall (), the plural number marking on the definite disambiguated the scope of the previous indefinite towards narrow scope. For the a-every quantifier order, the plural continuation was read more slowly than the singular continuation, but the effect did not reach significance in any of the individual regions of interest (words in the second sentence). A statistically significant difference between plural and singular was observable only when reading times for the entire continuation sentence were considered: as expected, the plural continuation was slower. For the every-a quantifier order, the singular continuation (compatible with inverse scope)
3
Technically speaking, the paraphrase in (12b) is also compatible with the surface-scope reading— see Tunstall’s Vagueness Principle mentioned in connection with (9b-ii) above.
č
was slower than the plural continuation, but the difference was not significant in any region or when the continuation sentence was considered as a whole. Anderson also let participants read a-every sentences without any disambiguating information, as in (). Each sentence was followed by a comprehension question (a) with two possible answers (b). Interestingly, the target sentence was read more slowly if participants in their next step selected the ‘Several’ answer indicating the inverse-scope interpretation. ()
A paratrooper jumped from every plane. a. How many paratroopers jumped from planes? b. One vs. Several
The same basic pattern—higher reading times for plural relative to singular continuations in a-every sentences—can be observed when sentences and continuations are placed in contexts that support either the surface or the inverse scope reading (Anderson, : et seqq.). Importantly, the inverse-scope reading is harder (takes longer) even in contexts that support it (Anderson, : et seqq.). Anderson takes this to support her Processing Scope Economy principle (p. , ()), which conjectures that the human sentence processing mechanism prefers to compute a scope configuration with the simplest syntactic representation (or derivation). As an alternative explanation, Anderson discusses the principle of parsimony (Crain & Steedman, ; Altmann & Steedman, ), which states that the interpreter favours a reading that carries fewer unsupported presuppositions. A modified version of that principle, considered by Anderson, would say that the inverse scope of () is dispreferred because the interpreter has to accommodate multiple entities (multiple paratroopers) into the discourse model, while the surface scope interpretation has to accommodate only one such paratrooper. This explanation would crucially not work, as it would not explain why the inverse-scope reading is harder even in supporting contexts that introduce multiple entities before the target quantified sentence. Note that Processing Scope Economy is similar to Tunstall’s PSI, except it explicitly states that computing a more complex configuration incurs a processing cost. Turning now to eye-tracking studies, Filik et al. () manipulated quantifier order (a-every vs. every-a) and grammatical-role order (direct-object vs. indirect-object first) in two-quantifier sentences of roughly the same kind as the ones studied by Tunstall (). Following the same pattern, these sentences were followed by continuations containing a singular vs. plural definite in subject position that was anaphoric to the indefinite in the previous sentence. An example item with numbered regions of interest is provided below (see Filik et al., : , table ); the numbered boundaries indicate where the regions end. ()
a. Direct object first, a-every: The celebrity gave| an in depth interview to every reporter from the newspaper, but| the interview(s) was/were| not very| interesting.| b. Direct object first, every-a: The celebrity gave| every in depth interview to a reporter from the newspaper, but| the reporter(s) was/were| not very| interested.|
c. Indirect object first, a-every: The celebrity gave| a reporter from the newspaper every in depth interview, but| the reporter(s) was/were| not very| interested.| d. Indirect object first, every-a: The celebrity gave| every reporter from the newspaper an in depth interview, but| the interview(s) was/were| not very| interesting.| In region , which contains the two interacting quantifiers, there was an interaction of quantifier order and grammatical-role order, with longer total reading times for every-a than for a-every when the direct object was first, and for a-every than for every-a when the indirect object was first. Upon closer inspection, the interaction for direct-object-first stimuli was due to first-pass times (reading time until exiting the region for the first time), while the interaction for indirect-object-first sentences was due to second-pass times (defined as total times minus first-pass times). It is important to note that these findings do not indicate any processing cost of scope since they did not interact with the scope-disambiguating continuation. In region (the anaphoric definite), plurals took longer than singulars overall but, unexpectedly, there were no effects related to quantifier scope. Filik et al. (: ) conjecture that this might be due to the ‘normal reading’ task, which does not encourage participants to evaluate the two scopal readings as deeply as the tasks in Tunstall ()—or Anderson (), we might add (Anderson, , and Filik et al., , were completed at about the same time, and they do not seem to be aware of each other). Paterson et al. () investigate the same kind of discourses as Filik et al. () except they use the universal quantifier each rather than every. They conjecture that the milder lexical bias for wide scope and distributive interpretations associated with every—as opposed to each—was the reason for the lack of quantifier-scope effects in region of Filik et al.’s items. The results for region in Paterson et al. () were similar to the ones in Filik et al. (): total reading times were higher for each-a than for a-each when the direct object was first, and for a-each than for each-a when the indirect object was first. And just like in Filik et al. (), there were no quantifier-scope effects in region . To sum up, the self-paced reading studies of Tunstall () and Anderson () see that plural continuations increase reading measures when they disambiguate towards inverse scope (of a-every sentences). In contrast, the eye-tracking studies of Filik et al. () and Paterson et al. () observe that plural continuations always increase reading measures, irrespective of scope disambiguation. A different way of disambiguating scope is used in Bott & Schlotterbeck (), who report an eye-tracking study investigating scope ambiguities in German. They consider sentences with the marked Accusative-Nominative word order, exemplified in () below. The sentence-initial Accusative argument consists of a quantifier that either includes the bound possessive pronoun seiner (a, c) or does not (b, d).4 The Nominative
Bott & Schlotterbeck establish in one of their pretests that the pronoun seiner is interpreted as bound in (15a). 4
č
argument is either a quantifier (exactly one teacher, (a, b)) or a definite NP (the new teacher, (c, d)). () a. Jeden seiner Schüler| lobte| genau ein Lehrer| voller| Each of-his pupils-acc| praised| exactly one teacher-nom| full-of| Wohlwollen.| goodwill| ‘Exactly one teacher praised each of his pupils full of goodwill.’ b. Jeden dieser Schüler| lobte| genau ein Lehrer| voller| Each of-these pupils-acc| praised| exactly one teacher-nom| full-of| Wohlwollen.| goodwill| ‘Exactly one teacher praised each of these pupils full of goodwill.’ c. Jeden seiner Schüler| lobte| der neue Lehrer| voller| Each of-his pupils-acc| praised| the new teacher-nom| full-of| Wohlwollen.| goodwill| ‘The new teacher praised each of his pupils full of goodwill.’ d. Jeden dieser Schüler| lobte| der neue Lehrer| voller| Each of-these pupils-acc| praised| the new teacher-nom| full-of| Wohlwollen.| goodwill| ‘The new teacher praised each of these pupils full of goodwill.’ Since binding is parasitic on scope, the presence of the bound pronoun seiner in (a) requires the Accusative quantifier to be interpreted in the scope of the Nominative quantifier. That is, (a) is only compatible with the inverse-scope reading.5 But the inverse-scope configuration is not required in the other three conditions: it is not needed in (b,d) because these sentences have no pronouns, and it is not needed in (c) because coreference is sufficient in this case. In line with previous observations that inverse scope is associated with processing cost, Bott & Schlotterbeck () observe an interaction between the presence/absence of the pronoun and the type of Nominative NP (quantificational vs. referential) in the early reading measures for regions and , and the late reading measures for regions –. As expected, the pronoun seiner is associated with increased reading times only when it has to be bound by the Nominative quantifier.6 Notably, Bott & Schlotterbeck () provide evidence that the parser has difficulties with inverse-scope readings even in every-a sentences. The fact that previous studies did not uncover this cost is probably due to the way scope disambiguation was performed in those studies. While Bott & Schlotterbeck () use bound readings to force inverse-scope interpretations, previous studies use singular vs. plural continuations, which is less illuminating since the singular continuation is technically speaking compatible with both the 5 6
This is confirmed by the results of another pretest reported in the paper. The same effect is also observed in region 4 in a self-paced reading study reported in the same paper.
surface-scope and the inverse-scope reading. In other words, there is no way to unambiguously force the inverse-scope reading for every-a sentences when using singular vs. plural continuations. From the studies discussed so far, it might seem that what the processor finds costly is to interpret the scope of quantifiers in a way that goes against their linear order. There is at least one construction that clearly argues against this position. When one quantifier is embedded inside another one, as in (), people prefer to interpret the sentence as if the quantifier introduced later, every city, scopes above the earlier one, someone. Such cases are known as ‘inverse linking’. Inverse linking has been shown (for German) to be associated with processing difficulties precisely when quantifier scope matches linear order (Bott & Radó ). ()
Someone from every city despises it.
All the studies discussed so far focused on the interaction of two quantifiers. There is at least one study, that of Dotlačil & Brasoveanu (), which investigated the processing of quantifier scope in sentences with three quantifiers and four possible continuations using eye tracking and self-paced reading. An example is given below: ()
A caregiver comforted a child every night. a. The caregiver [] wanted the child [] to get some rest. b. The caregivers [] wanted the child [] to get some rest. c. The caregiver [] wanted the children [] to get some rest. d. The caregivers [] wanted the children [] to get some rest.
There were four regions of interest in the continuation sentence, delimited as follows: ()
The caregiver(s) wanted| the child(ren)| to get| some rest.|
Dotlačil & Brasoveanu report that, in their eye-tracking study, plural subjects were associated with increased reading-time measures, as expected given the results in Filik et al. () and Paterson et al. (). The higher cost of plural subjects was observable in region (the subject definite) and spilled over into region (the object definite) and region (the wrap-up region). The processing cost of plural objects was observable in regions and . Interestingly, the results reveal that in several regions and measures, the cost of plural morphology is not additive: the condition with two plurals is at most as costly as the conditions with a single plural argument. This is visible as a negative interaction term when both the subject and the object were plural. The negative interaction term is statistically significant in region (probability of regression and rereading), region (first-pass times), and region (probability of regression and rereading). The eye-tracking results were further confirmed by a self-paced reading study, which also added another manipulation: letting the sentences in (a d) play a scope disambiguating role—if the sentences followed a multiple-quantifier sentence like ()—or not—if the sentences appeared out of the blue. The study revealed the same negative interaction as the eye-tracking experiment, indicating facilitation when both subject and object were plural, but crucially only when the plural morphology played a scope-disambiguating role.
č
This completes our overview of the empirical results about processing quantifier scope. With this summary in place, we now turn to a brief discussion of the theories put forth to explain these results.
. P
.................................................................................................................................. This section discusses and evluates the major processing theories of scope that we already informally introduced when we discussed the empirical results in the literature.
.. Underspecification theories vs. hierarchy-based accounts As we already mentioned, it has been repeatedly established that in a sentence with two or more scopally interacting quantifiers, the inverse-scope interpretation is dispreferred and harder to process (Ioup, ; Tunstall, ; Anderson, ; Reinhart, ; AnderBois et al., ; among many others). The asymmetry between the processing of surface and inverse scope indicate that underspecification theories of scope on their own are not sufficient to capture the processing behaviour exhibited by the human interpreter. In such theories (see Ebert, , for a summary and references), one assumes no scopal ordering unless specifically required. If we abbreviate underspecified scope as ‘’ (which stands for ‘≫ or ≪’), we can represent the default scope order for a sentence like () as shown below: ()
Subject Object
Thus, by default, any argument can be in the scope of or take scope over any other argument. Underspecification theories (on their own) predict that processing difficulty for surface and inverse scope should be the same: either they are both easy—no scope is processed until needed—or they should both be equally difficult—when scope finally needs to be processed, one scopal relation needs to be added for both the surface and the inverse scope reading. One way in which underspecification theories of scope might account for the contrast between singular and plural continuations is by attributing the difficulty to ‘shallower’ processing levels. For example, as Filik et al. () note, the processing cost of plural continuations after a-every and every-a sentences might arise simply because readers/listeners prefer to match morphological features of the definite to its antecedent NP, which, however, appears in the singular (see also Kemtes & Kemper, ). The mismatch in number for plural definites adds an extra processing cost visible in increased reaction times. There are, however, arguments against the hypothesis that inverse-scope difficulties can be reduced to morphological number mismatches. First, it would not explain the findings of several studies discussed in the preceding section that plural continuations are only costly in a-every continuations.
Furthermore, a surface-form number matching account makes specific predictions about items like () above: we should see independent processing costs when we anaphorically pick up these indefinites with plural definites in a subsequent sentence. For example, (b) and (c) should be equally difficult to process, and more difficult than (a), and (d) should be more difficult than either (b) or (c)—roughly by the same amount as these ones are relative to (a). However, this pattern has not been observed. In particular, while (b) and (c) exhibited an additional processing cost relative to (a), the combination of two plural arguments in (d) yielded a negative interaction: if one argument is plural and disambiguates towards inverse scope, it facilitates the inverse-scope interpretation of a subsequent plural argument. In contrast to underspecification theories, we can straightforwardly make sense of this data in hierarchy-based accounts. These accounts take scope representations to be total orders over the quantifiers in a sentence, which are computed based on some aspect of the (surface) form of the sentence (linear order, or c-command, or thematic-roles; see also section .). In particular, a total order requires totality (for any quantifiers Q and Q’, either Q scopes over Q’ or vice versa) and transitivity of scope (for any quantifiers Q, Q’ and Q”, if Q scopes over Q’ and Q’ scopes over Q”, then Q scopes over Q”). Consequently, in a sentence like (), the subject indefinite a caregiver is by default assumed to take scope over the object indefinite a child. When this sentence is followed up by the double-plural continuation in (d), we know that the universal every night takes scope over the subject indefinite as soon as we read the plural definite the caregivers in the continuation. By totality and transitivity of scope, it follows that the universal every night has to also take scope over the object indefinite. We therefore expect the plural subject to make the plural object easier to process, which matches well with the observed negative interaction (see Sayeed for a formal analysis of this data in event semantics). Morphological number matching accounts built on top of underspecification theories make the opposite prediction: the two plurals in (d) should take longer than the single plural in (b). And ‘bare’ underspecification accounts (with no surface number-matching requirements) do not distinguish between continuations like (d) and (b) since readers should do the same amount of work in both cases: they need to independently specify the two scopal relations between the universal and each of the two indefinites. Cases like () also provide an argument against accounts in which no scopal representations are retained. For example, let’s assume scopal relations are determined by the linear order of the quantifiers (Johnson-Laird et al., ), but are not stored as such. Instead, readers directly modify the mental model they construct as they interpret a sentence, and that mental model is the only representation they maintain. Concretely, let’s take the interpretation of a sentence to consist of a mental model along the lines of Johnson-Laird (). Extending this work, Johnson-Laird et al. () propose that distributive quantifiers add a loop to the mental model, which can be summarized as follows: i. introduce one element from the domain of the quantifier; ii. interpret the scope of the quantifier, introducing a single new element for every indefinite; iii. go back to step (i); iv. repeat until all elements in the domain of the distributive quantifier are introduced.
č
Assuming that readers start with the default model created by the linear order of quantifiers and directly modify their mental model only if necessary, they should consider the subject indefinite and the object indefinite in () outside the loop of the distributive quantifier. Then, when they read the continuation with a plural subject like (b), they minimally revise the mental model so that the (discourse) referent contributed by the subject indefinite is inside the loop contributed by the universal quantifier. There is no reason to make additional revisions, so we expect the object indefinite to still outscope the universal quantifier because that was the default scope induced by linear order (importantly, linear order is used to construct the original mental model, but it is not invoked later on). The double-plural continuation in (d) requires two revisions of the default mental model, so if anything, it should be slower / more difficult to process than (b). This is contrary to the facts. Thus, the results in Dotlačil & Brasoveanu () are most compatible with the position that a total hierarchy is stored and manipulated when interpreting and revising quantifier scope. Importantly, this finding does not automatically provide an answer to the separate question of processing cost. That is, we now have a good idea about the kind of scopal representations that are built and maintained/updated in discourse, but we have not yet identified the cause for the processing cost associated with inverse scope. This issue is addressed in the following subsection.
.. Cost of inverse scope as scope revision The main idea behind most processing accounts of inverse scope is that the cost associated with it is due to the revision of a mental representation that is constructed ‘by default’ when a sentence is interpreted. The accounts differ mainly with respect to the nature of this mental representation. Differences can also arise depending on the specification of the revision procedure, but we focus here on the nature of the representation that is manipulated. The two major choices are (i) LF revision and (ii) discourse model revision. We can think of these two classes of mental representations as different stops on the way from the surface form of a sentence to its truth-conditional interpretation. LFs are the first stop on the way to truth conditions, which disambiguates between various syntactic structures that are left unspecified by the surface structures. Basically, this is surface structure + quantifying-in / Quantifier Raising (QR; Montague, ; May, ; among many others). LFs are very rich representations: they maintain the surface morphosyntactic information and add to it additional information necessary to identify a specific set of truth conditions. Discourse models—specifically, Discourse Representation Structures (DRSs; Kamp, a; Kamp & Reyle, ; Kamp, )—are further removed from surface syntactic structures. They might retain some of the surface morphosyntactic information—for example, person, number, and gender information needed for pronoun resolution, or other kinds of form-related features needed for presupposition resolution (e.g. the specific location in the surface syntactic structure where the presupposition was triggered)—but many morphosyntactic features that have no semantic import are not encoded. Discourse models also contain scopal information, but the way that information is encoded might abstract away from many features of the overt syntax that would still be present in LFs.
In this subsection, we will briefly introduce these two types of processing theories of scope and attempt to empirically distinguish between them.
... Logical form revision In one of the two classes of accounts we mentioned, processing difficulties associated with inverse scope are the result of reanalysing a logical form representation that encodes scope (Kurtzman & MacDonald, ; Tunstall, ; Anderson, ; Pylkkänen & McElree, ; Hackl et al., ). Abstracting away from the specifics of the linguistic representations, which vary somewhat from account to account, the strategy employed by this family of approaches goes as follows. Recall the example in (), namely A boy climbed every tree. The processor considers the basic hierarchy ‘Subject ≫ Object’ in () by default and builds an initial representation (‘LF’) for () in which the subject a boy scopes over the object every tree. When the processor encounters the continuation The boys . . . in (b), it has to switch from that default scope representation to the marked hierarchy ‘Object ≫ Subject’ in (), which allows the subject indefinite to introduce multiple individuals. This reanalysis, as is often the case, incurs processing cost. One common way to specify this general strategy is to say that inverse scope requires an extra QR operation (Tunstall, ; Anderson, ; Pylkkänen & McElree, ; Reinhart, ; among others) to derive the requisite logical form representation. This is shown in (), modelled after Fox (), where quantifiers always adjoin to Verb Phrase (VP) in their original order of c-command, and the inverse-scope interpretation requires an extra movement and adjunction of a quantifier. ()
A boy climbed every tree.
Surface scope:
NPi a boy
Inverse scope: NPj
VP
ti
climbed
tj′
VP ti
V′ V
VP
NPi a boy
VP
NPj every tree
every tree
tj
V′ V
tj
climbed
Another version of this approach appeals to type-shifting instead of QR: an optional typeshifting operator has to be inserted to derive inverse-scope readings (Hendriks, ). The operator can be reified (that is, actually inserted) in the logical form representation, or can have a more ethereal nature and be introduced only during semantic composition. Either way, an extra reanalysing operation is necessary, which can explain the processing cost of inverse scope.
č
... Discourse model revision Alternatively, we could explain inverse-scope processing difficulties in terms of changes to the discourse model structure: inverse scope is harder because it requires revising the already built discourse model structure (Fodor, ; see also Crain & Steedman, , Altmann & Steedman, ). To see this, consider how sentence () is interpreted on-line. We first hear or read A boy climbed . . . , at which point we add a new entity to our discourse model that is a boy and that stands in the climbing relation to whatever direct object we are about to interpret. Then we hear or read the direct object . . . every tree. If we want the direct object quantifier to take wide scope, we need to revise the current discourse structure and introduce a set of boys, each of which is associated with a possibly distinct tree. Before moving on, we note that in this presentation, we tacitly assume that discourse models are built eagerly and incrementally: that is, they are basically updated after every word. The presented account would also be compatible with the position that the discourse model structure is constrained and specified only when certain semantically coherent ‘chunks’ (domains) are processed.7 We can make the proposal more specific in various ways. Following Johnson-Laird et al. (), we can take the object quantifier to add a loop to the mental model. By default, the subject indefinite is interpreted outside this loop, and this assumption has to be revised when the universal in object position takes inverse scope over the subject. Alternatively, we can take discourse / mental models to be nothing other than the DRSs that are building blocks of meaning in Discourse Representation Theory (DRT; Kamp, a; Kamp & Reyle, ). Using DRT as the basic framework is perhaps the best way to conceptualize this line of approaches since, as a representational theory of meaning and interpretation, DRT far outstrips any other mental model-based approach in formal explicitness and sophistication, as well as empirical success. In DRT, quantifiers are taken to introduce conditions with a specific structure (duplex conditions) to the current DRS. A universal in object position introduces a duplex condition in the current / main DRS—crucially, after the subject indefinite already introduced a discourse referent in the same DRS. The discourse referent introduced by the subject is therefore outside the duplex condition contributed by the universal, hence outside its scope. The inverse-scope reading requires a revision of the main / current DRS to the effect that the discourse referent contributed by the subject indefinite (and
7 This is what Bott & Schlotterbeck () argue for. They base their argument on sentences in which a lexical verb either appears between two quantifiers (QVQ) or follows them (QQV). If inverse scope is costly and calculated right away, we should observe processing cost on the second quantifier irrespective of the verb. If only particular chunks receive scope interpretations—more concretely, propositions— inverse scope would be calculated on the second quantifier in the QVQ condition, but delayed until the lexical verb appears in the QQV condition, delaying the cost in the latter case. The evidence is equivocal. In their self-paced reading study, Bott & Schlotterbeck report that inverse scope is costly on (the spillover of) the second quantifier only in the QVQ condition. However, the interaction term showing this is only marginally significant in the ANOVA analysis by items. The follow-up eye-tracking study does not replicate the effect, not even marginally so. Rather, readers slow down on the second quantifier in both QVQ and QQV orders.
whatever other conditions the indefinite contributed to the main DRS) be moved inside the duplex condition contributed by the universal. This change requires additional cognitive operations, yielding the additional processing cost associated with inverse scope.
.. Distinguishing between logical form and discourse-model theories The empirical results discussed in section . are compatible with approaches in which the cost is driven by changes in LF. In fact, several accounts explicitly adopt that position, albeit with slight modifications (Tunstall, ; Anderson, ; Bott & Schlotterbeck, ; among others). They are also compatible with discourse-model revisions. For example, the most robust result, namely the processing cost tied to the inverse scope of a-every sentences, can be derived because it requires a change in the constructed DRT, as we just discussed. Inverse-scope also requires a change in the number of valuations associated with the discourse referent contributed by the indefinite (one vs. many). Brasoveanu & Dotlačil (a) consider a new type of evidence that can distinguish between logical form and discourse-model approaches. This evidence is provided by the interaction between quantifiers and adjectives of comparison (AOCs) like same and different. Consider the sentence in () below. ()
The same student saw every movie.
The AOC same has the so-called sentence-internal reading here, on which one and the same student saw all the movies. This reading is only possible if same is interpreted in the scope of a semantically plural NP that distributes over it (Carlson, ; see also Syrett, Chapter in this volume, for a discussion on distributivity). Semantically plural NPs include morphologically singular and distributive universal quantifiers like every movie, but also morphologically plural NPs like all the movies. Sentence-internal same has the potential to distinguish between logical form and discourse-model theories. This is because same in the sentence-internal reading requires inverse scope in sentences like () but no revision of the discourse model structure is necessary when a quantifier takes inverse scope over same. The universal every movie scopes and distributes over same to license its sentence-internal reading, but the model will contain only one student both before and after the processing of every movie.8 The self-paced reading study in Brasoveanu & Dotlačil (a) considered four different quantifiers that either preceded same, as in (a) below, or followed it, as in (b). The quantifiers had to license the sentence-internal reading, otherwise the sentence would be infelicitous.
Other experimental studies examining scope and AOCs like same and different are Anderson () and Dwivedi et al. (). The former only looked at the interpretation of different, which unlike same cannot distinguish between different families of approaches to scope processing. The latter examined event-related brain potentials in the processing of sentence-internal same and different, but considered only surface-scope structures. 8
č
9 8 all the > > > > = < each researcher(s) learned the same language spoken in the eastern a. . . . every > > > > ; Indonesian islands. : the 9 8 all the > > > > = < each language(s) spoken in the eastern b. . . . the same researcher learned every > > > > ; : Indonesian islands. the
()
The main regions of interest are the four words that follow the direct object, which is the word same in (a) or the quantificational licensor in (b). These regions were matched across all conditions and are boldfaced in (a,b) below. Several other measures were considered, including reading times for full sentences (since Anderson, , observed processing cost in these measures) and answers to comprehension questions. Interestingly, no effect of inverse scope was detected. This is compatible with discoursemodel theories but problematic for the theories that tie cost to logical form revision. The sentence-internal reading of same requires logical form revision as much as the inversescope interpretation of an indefinite does. However, these revisions do not change some parameters of the discourse model, unlike indefinites. Specifically, the same researcher introduces the same number of discourse referents—or the same number of valuations for the discourse referent it contributes—under both the surface-scope and the inversescope reading. One concern about this argument is that it is based on a null effect, so it is in principle possible that the finding is simply due to the lack of power of the reported experiment. This, however, does not seem to be the case—see Brasoveanu & Dotlačil (a) for arguments using Monte Carlo simulations that this is very unlikely.
. C: B
.................................................................................................................................. In section . and subsection .., we saw that scope hierarchies—potentially identical to LF structures—are built and maintained/updated when discourses containing sentences with multiple quantifiers are interpreted. In subsections .. and .., we saw that only changes in discourse models (or semantic valuations associated with discourse models) yield processing costs. It might seem that these findings contradict each other, but in fact it is fully coherent to accept both of these experimental conclusions. Here is one possible way to integrate them into a unified processing theory of scope. Suppose for simplicity that the processor does indeed construct and maintain LF structures as part of its interpretation procedure. LF structures, however, are not the end result of interpretation; they cannot be—we need an actual semantics for them to capture intuitions about entailment, logical equivalence, etc. Thus, LFs have to be interpreted relative to a model. Furthermore, suppose that LF structures are not forgotten as soon as they are interpreted. Instead, they—or similar representations, for example, DRSs possibly enriched with
scope hierarchies—are carried over intra- and cross-sententially and can be modified in light of new discourse information. This makes sense: if scope hierarchies were used to encode quantifier scope in the first place, it is natural that they would be consulted again and updated when revising that quantifier scope. Now LF modifications/revisions triggered by new discourse information might in turn trigger new, updated discourse model structures (or valuations thereof)—or they might not. The evidence in subsection .. indicates that LF modifications themselves are not costly. The cost only comes about if the LF modifications have consequences for discourse models: for example, if new interpretations are incompatible with previous assumptions about (the number of) discourse referents and/or valuations for these referents. In sum, what is modified and what is costly for the processor (at least in the range of tasks we examined) might only partially overlap. Some LF-level representation revisions might be cost-free as long as they do not have specific kinds of effects on discourse models. This seems to suggest that the representations built by the human processor/interpreter have multiple ‘layers’ or facets—an LF-like one that is closer to the surface syntactic structure, and a discourse-model one that abstracts away from various surface features. This ‘layered’ view of semantic processing raises many questions about the exact nature of these representations, as well as about which interpretive subprocesses target which of these ‘layers’/facets, and what the temporal unfolding of these subprocesses is.
A We would like to thank our editor Chris Cummins, as well as Alexandre Cremers, Donka Farkas, and Thom van Gessel for comments, questions, and suggestions. Jakub Dotlačil was supported by the NWO VENI grant --.
......................................................................................................................
......................................................................................................................
.
I is well-established that young children have difficulties restricting the scope of a universal quantifier (e.g. all, each, or every) to a specific noun phrase. To cite a famous example (Philip, ), children often reject the sentence Every farmer is feeding a donkey as a description of a scene depicting a mismatch between the number of farmers and donkeys (see Figure .), presumably on account of the extra donkey. The associated comprehension error is commonly referred to as quantifier spreading, which implies that the scope of the universal quantifier has spread beyond the noun it modifies. This chapter reviews the literature on quantifier spreading errors and related issues involving comprehension of universal quantification, defined as ‘a complete binding exercised over all members of a set which falls within its scope’ (Freeman, : ). We first summarize the seminal work of Inhelder & Piaget (, a) who first described young children’s errors with universal quantification in relation to the development of logical reasoning (i.e. children’s grasp of class inclusion relations). We then describe early work on quantifier spreading that utilized the truth-value judgement task (Donaldson & Lloyd, ; Freeman, ), a methodology that would remain popular in studies of quantifier spreading in subsequent decades, complemented by work utilizing picturechoice and act-out tasks. The next section describes early work on deductive reasoning, which tested children on categorical syllogisms and identified conversion errors as a related phenomenon (Bucci, ; Johnson-Laird et al., ). Following this introduction, we summarize accounts of quantifier spreading that explained errors in terms of structureneutral representations (Bucci, ), incomplete mastery of the mappings between linguistic cues and semantic representations (Brooks & Braine, ), quantification over events (Philip, ), ‘floating’ quantifiers (Roeper et al., ), and shallow processing of syntactic structure (Brooks & Sekerina, /). We include in our review studies that extended the phenomenon of quantifier spreading to adults (Berent et al., ; Street & Dąbrowska, ), suggesting that individual differences in language background and educational attainment are predictive of error rates. The accounts attributing errors to syntactic processing are contrasted with the full competence hypothesis that children have adult-like knowledge of universal quantifiers (Crain et al., ). Under this view and related pragmatic accounts of quantifier spreading, children make errors due to the demand characteristics of the truth-value judgement task. We conclude by discussing efforts to identify signature patterns of attention associated with
.. Every farmer is feeding a donkey (Philip, )
quantifier spreading errors using eye-tracking methodology (e.g. Minai et al., ; Sekerina & Sauermann, ) and future directions.
. I
.................................................................................................................................. Inhelder & Piaget (, a) noted errors in children’s comprehension of the universal quantifier all in the context of the class inclusion task—an assessment of children’s ability to classify objects on the basis of common features as well as their understanding that sets of objects may be subsets of larger classes; see also Skordos & Barner (Chapter in this volume). In a typical class inclusion task, the child might be shown a set of counters comprising, for example, five blue round discs, two blue squares, and two red squares. When asked Are all the circles blue? young children would typically answer ‘no’ while pointing out the blue or red squares. As an explanation of their misinterpretation of the question, Inhelder & Piaget (a: ) suggested that It looks as if the child’s thinking is conditioned by a need of symmetry: the extension of the predicate blue must be the same as that of the subject round. < . . . > What our subjects are doing is to assimilate the expression All As are B to the expression All Bs are A. [Our subjects] substitute equivalence (A=B) for class inclusion (A>B or B>A).
.
In other words, children interpret All the circles are blue as implying also that All the blue ones are circles. Although Inhelder & Piaget never used the term quantifier spreading, it appears that the concept they described is essentially the same phenomenon, as exemplified in their statement ‘It therefore looks very much as if the true explanation is that at stage II children extend the quantifier all to the logical predicate of the sentence as well as to its logical subject’ (a: ). Inhelder & Piaget (, a) suggested that young ‘preoperational’ children do not have an understanding of universal quantification until the age of or , when concrete operational reasoning emerges. At younger ages, children rely on spatial relations and physical arrangements when dividing objects into sets. Not long after Inhelder & Piaget (, a) first described children’s errors in restricting the scope of the universal quantifier all, other researchers launched studies to delineate the phenomenon. These early studies used variants of the Piagetian truth-value judgement task in which children answered a yes/no question about a situation. In the modified tasks, children were instructed to reward (or punish) a puppet for making true (or false) statements about the situation. (Other variants used in later work on quantifier spreading involved sentence-picture verification, with participants asked to compare a sentence with a picture to determine whether they match.) The truth-value judgement task is thought to assess what the participant understands about the situation (or picture) and its relationship to the sentence that describes it, without requiring metalinguistic awareness about the structure of the sentence (Clark & Chase, ; Gordon, ). Donaldson & Lloyd () introduced the truth-value judgement task as a methodology for exploring children’s comprehension of the universal quantifiers all and each. Fourteen preschool children (- to -year-olds) were shown an array of four garages and a set of either three or five cars, with the cars arranged in partial one-to-one correspondence with the garages. The children were instructed to tell the toy panda whether he was right (by pressing a bell) or wrong (by pressing a buzzer) after the panda produced a statement as a description of the scene. When there were four garages and three cars, children tended to evaluate the statement All the cars are in the garages as wrong, often justifying their answer by noting the emptiness of the fourth garage. Similarly, where they saw four garages and five cars, they rejected The garages have all got cars in them—stating, for example, ‘Cos not all in them. That police car’s out’ (: ). Notably, although all children made errors on the task, Donaldson & Lloyd () emphasized that the pattern of errors varied across children, with some children’s responses suggesting an insistence on one-to-one correspondence between cars and garages and other children’s responses seeming more random. A decade later, Freeman () would distinguish two different error types, attributing the errors to overexhaustive and underexhaustive search, respectively, although only the former error was described in Donaldson & Lloyd (). These terms continue to be used in literature to the present day, with overexhaustive search recognized as the classic quantifier spreading error, involving failure to restrict the scope of the universal quantifier to its subject (e.g. incorrectly rejecting the statement All the cars are in the garages as a description of a scene with an extra garage). Underexhaustive search resulted in a much less common error, involving failure to extend the search for referents within the scope of the universal quantifier beyond the subset in one-to-one correspondence (e.g. incorrectly accepting the statement All the cars are in the garages as a description of a scene with an extra car). In a second study contrasting two sets of cars and garages (Donaldson & McGarrigle, ), forty - to -year-olds made yes/no judgements about whether one set had more cars
than the other, and whether one set had all of the cars. In this study, the cars were arranged on two shelves, one under another, with five cars on one shelf and four cars on the other. Children were first asked whether one of the shelves had more cars on it than the other—a question that the majority could answer quickly and accurately. After that, the row of four cars was enclosed by a row of four garages (i.e. with one car per garage), and the row of five cars was enclosed by a row of six garages (i.e. with one empty garage). When the same question was repeated, about a third of the children changed their answer, indicating that the shelf with four cars (in garages) had more cars on it than the shelf with five cars (and an extra garage). Similarly, when the garages were present and the children were asked Are all of the cars on this shelf? they often agreed that the smaller set (four cars in garages) had all the cars while denying it for the larger set (five cars in garages, with an empty garage). In summarizing the results across studies, Donaldson () emphasized that the salience of the empty garage influenced children’s decisions: ‘The statement which they actually judged, irrespective of the variations in linguistic form, was All the garages are full’ (: ). As a direct test of the hypothesis that children’s errors were due to the salience of the empty garage, Freeman () modified the design of Donaldson & Lloyd () by positioning a boat in the last (empty) garage, with the boat condition compared with the original empty-garage condition in a between-subjects design. Having a boat in one of the garages was predicted to inhibit overexhaustive errors in response to the question Are all the cars in garages? when there were three cars in four garages. Likewise, when there were five cars and four garages, having a boat in one of the garages was expected to inhibit underexhaustive errors in response to the question Are all the cars in garages? by providing a clue that ‘the set of garages is not be taken as a sure guide to the number of cars to be considered’ (: ). In addition to manipulating the presence/absence of a boat in the fourth garage, Freeman () also examined whether provision of background information would eliminate errors. Half of the children were explicitly told that not all of the garages would be used for housing cars (i.e. not everybody has got a car); this background information was expected to reduce children’s tendency to make overexhaustive errors by giving them an explanation for the empty garage (i.e. there is no car that ought to be there). Four groups of twenty-two children (- to -year-olds) were tested, with the hypothesis that when both extra variables were applied (i.e. boat present in garage, background information provided), children would respond to the questions with higher levels of accuracy. (Note that for each set of cars and garages, children were asked two questions Are all the cars in the garages? and Have all the garages got cars in?) The experiment produced two major findings: first, the background information was irrelevant to children’s responses, implying that, in line with Inhelder & Piaget (a), the physical arrangement of the cars and garages was more salient to the children than the verbal information. Second, although the presence of the boat reduced errors (.% of children in the boat condition made no errors in comparison to only .% in the empty-garage condition), the fact that the majority of children in the boat condition still made overexhaustive and/or underexhaustive errors indicated that the salience of the empty garage was only one of several factors contributing to task difficulty. Freeman & Stedmon () built on these findings by considering further how the presence of a salient boundary influences children’s response patterns. In line with earlier work by Karmiloff-Smith (), they elicited underexhaustive errors by splitting the garages into two sets (i.e. two distinct ‘parking lots’). Children often ignored the cars in
.
the more distant lot when asked if all the cars are in a garage, yet when asked to collect all the cars, they readily took the cars from both locations—indicating their grasp of the universal extension of all. Hence, children’s errors in comprehending all were only evident when tasked with judging the alignment of the two sets (cars and garages). Freeman & Stedmon (: ) concluded that ‘children are simply so eager to find a boundary which the word “all” cues them for, that they stop as soon as they reach any marked boundary since that does indeed fulfill the condition of search-up-to-a-boundary.’ In other experiments, Freeman & Stedmon observed error rates that varied as a function of the specific universal quantifier used (every, all, absolutely all, or every single), highlighting the importance of examining semantic distinctions among universal quantifiers in relation to quantifier spreading—a topic addressed in work by Brooks & Braine () and described later in this chapter.
. L
.................................................................................................................................. Universal quantifiers play a critical role in deductive reasoning (i.e. predicate logic) due to the rich set of inferences they license (Braine & O’Brien, ). Hence, research on the development of logical reasoning has centred on children’s acquisition of logical vocabulary (e.g. universal and existential quantifiers, such as all and some; logical connectives, such as if, and, or, not) and children’s ability to use these terms to draw deductive, category-based inferences from sets of premises (Braine & Rumain, ). Early studies of syllogistic reasoning identified conversion errors in which a premise, for example All As are Bs, appeared to be interpreted as equivalent to its converse, All Bs are As (Chapman & Chapman, ). For example, when adults were given categorical syllogisms of the form shown in (), conversion of the second premise to All Bs are Cs often resulted in acceptance of the erroneous conclusion All As are Cs. ()
All As are Bs. All Cs are Bs.
Newstead et al. () investigated when the adult pattern of responding to syllogistic setinclusion relations emerged in development. They presented the passage shown in () to a total of children, ranging in age from to years: ()
Some children were deciding what candies to buy. All the round candies were chocolates. (All As are B) All the chocolates were large. (All Bs are C) All the large candies were hard. (All Cs are D) They all went to the shop and bought what they wanted.
Children were then asked a series of questions that included valid inferences (e.g. Were all of the round candies large? Correct answer: yes) as well as their converse (e.g. Were all of the
large candies round? Correct answer: can’t tell). From the age of years, children’s responses matched the pattern observed with adults in previous research. When the inference was valid, accuracy was highest if the relevant information straddled adjacent premises (e.g. All the round candies are large), and declined when the relevant premises were non-adjacent (e.g. All the round candies are hard). Conversely, if the inference was invalid, performance was lowest when relevant information was contained in adjacent premises (e.g. All the large candies are round), presumably due to higher probability of making conversion errors. Newstead and colleagues concluded that children as well as adults interpret these syllogistic forms as involving transitive and convertible similarity relations; that is, if All As are Bs and All Bs are Cs, then All As are Cs and All Cs are As. Other research focusing on developmental changes in deductive reasoning examined the role of problem content in constraining children’s ability to follow logical rules. In an early study, Bucci () tested thirty-two children in each of two age groups (- to -year-olds; - to -year-olds) as well as adults on various syllogistic forms; see Table . for examples of problem types. Children at both ages demonstrated high accuracy on Modus ponens and Modus tollens, but showed significantly lower accuracy on syllogistic forms requiring a ‘not sure’ response. For the latter problem types, children were much more accurate on problems where they could rely on broad-based factual knowledge than on problems with abstract content (see ‘block’ problems in Table .), which suggests that they prioritized real-world knowledge (e.g. their ability to think of other animals with whiskers besides cats) over logical form when making inferences. Hawkins et al. () also found young children’s syllogistic reasoning to be influenced by the relationship between problem content and their real-world knowledge. In this study, forty young children (- to -year-olds) were presented with a variety of syllogistic
Table .. Four forms of categorical syllogisms with example problems Modus ponens
Affirming the Consequent
Denying the Antecedent
Modus tollens
Syllogistic Form
All F are G. X is an F. Is X G?
All F are G. X is G. Is X F?
All F are G. X is not an F. Is X G?
All F are G. X is not G. Is X F?
Correct Response
Yes
Not Sure
Not sure
No
Content Type
Broad predicate
Abstract
Broad predicate
Abstract
Example Problem
All football players are strong. This man is a football player. Is he strong?
All the square blocks are purple. The block in my hand is purple. Is it square?
Every cat has whiskers. This animal is not a cat. Does it have whiskers?
Every white block is small. The block in my hand is not small. Is it white?
Source: Adapted from Bucci ().
.
problems, a third of which had incongruent content that violated real-world knowledge, as in example (). ()
Birds can fly. Everything than can fly has wheels. Do birds have wheels?
Children provided very few correct responses to the problems with incongruent content (averaging only .% correct), and often justified their erroneous responses by referring to their existing knowledge, as in ‘No, birds have wings!’ In contrast, their performance approached ceiling (.% correct) on problems with congruent content that matched their real-world expectations. The children also showed considerable success in drawing inferences from problems with fantasy content (e.g. Every banga is purple. Purple animals always sneeze at people. Do bangas sneeze at people?), averaging . per cent correct overall, with their performance approaching ceiling (.%) when the fantasy problems were presented first (i.e. prior to the problems with incongruent content). The authors concluded that under some circumstances preschool-age children could draw deductive inferences from the logical form of premises; however, pragmatic forms of reasoning relying on pre-existing knowledge were more pervasive. Johnson-Laird et al. () went further to fully reject the notion that children were using formal sets of logical rules to perform syllogistic reasoning. They offered the alternative proposal that children, as well as adults, were constructing mental models of the information provided in the premises, drawing information from these mental representations to construct hypotheses, and scanning the models for alternatives prior to drawing their conclusions. Under this view, the development of deductive reasoning reflected children’s growing skills in constructing and manipulating mental models of problem content. Across two experiments, they presented children ranging in age from to years with syllogisms comprising all of the possible pairs of premises involving the terms all, some, and none (where only twenty-seven of these sixty-four pairs generate valid conclusions). The children were given problems with sensible content (e.g. Some of the footballers are musicians. All of the footballers are runners.) and instructed to state what followed from the premises with the assistance of cards that summarized the possibilities (including no valid conclusion). The authors found problem difficulty to be related to the complexity of creating mental models of each premise as well as the number of distinct models that had to be considered when searching for counterexamples to tentative conclusions. They identified a specific problem type associated with conversion errors, shown in (), for which children often drew the erroneous conclusion All A are C. ()
All B are A. All B are C.
In this and other presumed cases of conversion of a premise to its converse (e.g. All B are A is taken to imply All A are B), Johnson-Laird and colleagues argued that the errors were due to the difficulties of aligning the mental models constructed for each premise. Their account challenged the view that children applied a general rule or logical operation that converted all premises of the form All A are B to entail their converses.
. S-
.................................................................................................................................. In addition to testing children’s ability to solve logical syllogisms, Bucci () introduced an act-out task in combination with probe questions to assess children’s grasp of universal quantification. Children (thirty-seven at ages to years; thirty-eight at ages to years) and adults were asked to assemble buildings using multi-coloured blocks, with varied instructions as illustrated in (): ()
Make a building where all the yellow blocks are square. Make a building where all the square blocks are yellow.
After participants completed each building, they were asked questions in relation to different blocks (e.g. blue squares, blue rectangles), such as Could we use any of these in a building in which all the yellow blocks are square? The majority of children in both age groups gave responses that suggested they had adopted a structure-neutral interpretation of the instructions: that is, one in which they failed to make a distinction between the subject and the predicate of the sentence. In other words, it appeared as if they encoded sentences as a string of words without hierarchical relationships; for example, All the yellow blocks are square was encoded as ‘all, yellow, square’. When constructing the buildings, children used only the blocks with both properties (e.g. yellow and square) and often expressed frustration that they were limited to using just a few blocks, as in ‘Why do you keep telling me to make these stupid little buildings?’ (Bucci, : ). When responding to the probe questions, they rejected blocks that were square but not yellow (e.g. blue squares) as well as blocks that were neither yellow nor square (e.g. blue rectangles). (Note that Inhelder & Piaget (a) reported both of these error types on class inclusion problems, for example, with some children justifying their answers to the question Are all the circles blue? by referring to objects that were neither circles nor blue, as in ‘No, there are red squares’.) Bucci () suggested that the lack of information provided by the structure-neutral representation would lead children to rely on the discourse context, real-world knowledge, or guessing, in order to answer the questions. She referred to this strategy as a pragmatic mode of processing that reflected children’s lack of expertise in using linguistic cues to delimit the scope of the universal quantifier.
. D
.................................................................................................................................. Starting from the hypothesis that children’s errors might reflect a lack of knowledge of the linguistic cues to universal quantification, Brooks & Braine () extended the line
.
.. Examples of collective (left) and distributive (right) pictures, adapted from Brooks & Braine (: Experiment )
of research by focusing on children’s acquisition of the lexical semantics of the universal quantifiers all and each. Building on the work of Vendler () and Ioup (), they explored linguistic factors that might influence whether children would assign collective or distributive interpretations to sentences containing universal quantifiers. One hundred children, ranging in age from to years, were asked to choose between a collective picture depicting three people acting together in relation to a single object and a distributive picture depicting three people acting individually on three different objects (bears in the example shown): see Figure .. Children were instructed to point to the picture that went best with a sentence spoken aloud, with the sentences varying over trials in terms of the universal quantifier (all vs. each) and the sentence structure (active vs. passive voice), as illustrated in (). Note that different sets of pictures were used for each trial, with materials counterbalanced across sentence structures. ()
All of the men are washing a bear. A bear is being washed by all of the men. Each man is washing a bear. A bear is being washed by each man.
From the age of years, children’s picture choices varied as a function of both the universal quantifier (with more collective pictures selected when the quantifier was all) and the sentence structure (with more collective pictures selected when the sentence was in the passive voice), with response patterns becoming more stable with age. The authors interpreted the findings as suggesting that children acquire sets of linguistic cues that jointly determine sentence meaning, with sensitivity to these cues emerging by years of age. A subsequent picture-choice experiment focused on quantifier spreading. Sixty children (ages to years) were presented with sets of collective and distributive pictures depicting extra, unattended objects (bears in the example shown) along with a third ‘exhaustive’ picture that showed a set of three people acting on five objects: see Figure .. Brooks & Braine () hypothesized that children who adopted a structure-neutral representation of
.. Examples of collective (left), distributive (centre), and exhaustive (right) pictures, adapted from Brooks & Braine (: Experiment ) and Brooks et al. ()
the sentence All of the men are washing a bear (or Each man is washing a bear) would opt for the exhaustive picture over a collective or distributive picture with extra objects. Contrary to the structure-neutral hypothesis, children from the age of years distinguished the meanings of all and each and associated sentences containing these quantifiers with the collective and distributive pictures, rather than the exhaustive one. These findings were replicated and extended in a cross-linguistic study using translational equivalents of all and each in Mandarin Chinese and Brazilian Portuguese (Brooks et al., ). Across samples, children showed sensitivity to lexical semantic cues to universal quantification from about years of age, with responses becoming more consistent with age. Brooks & Braine () also explored whether children’s errors in restricting the scope of a universal quantifier would replicate in a picture-choice task, and whether errors would be observed outside of the classic context involving sets in partial one-to-one correspondence. Four sentence structures were used, as illustrated in (): ()
All of the men are washing a bear. There is a man washing all of the bears. Each man is washing a bear. There is a man washing each of the bears.
For sentences with all, children were asked to choose between collective pictures with two extra objects (bears in the example shown) or two extra actors (see Figure .), whereas for sentences with each, children chose between distributive pictures with two extra objects or actors (see Figure .). Sixty children were tested, ranging in age from to years. At all ages, children were highly accurate (> % correct) in choosing the correct collective picture for the sentences with all, which suggested that they could use the combined set of cues (i.e. the position of the universal quantifier, and singular vs. plural forms of the nouns in subject and object position) to match the sentences with the appropriate pictures. In contrast, children made frequent errors with the distributive sets in partial one-to-one correspondence, with only - and -year-olds showing success (> % correct) in varying their picture choices as a function of the sentence structure (i.e. whether each modified the subject or the object of the sentence). These findings, as well as later work with additional control conditions (Brooks & Sekerina, /), highlighted the specific challenges associated with using linguistic cues to match sentences with distributive depictions of sets in partial one-toone correspondence.
.
.. Examples of collective pictures showing sets with extra objects (left) or extra actors (right), adapted from Brooks & Braine (: Experiment )
.. Examples of distributive pictures showing sets with extra objects (left) or actors (right), adapted from Brooks & Braine (: Experiment )
. T
.................................................................................................................................. In a widely cited dissertation, Philip () used the truth-value judgement task to explore preschool children’s errors with the universal quantifier every in contexts such as the one described at the beginning of this chapter and depicted in Figure .. Based on findings that
Quantifier
Restrictor
EVERY (x) farmer (x)
Nuclear Scope
x is feeding a donkey
‘For every x, x a farmer, x is feeding a donkey’
.. Standard semantic representation of the sentence ‘Every farmer is feeding a donkey’, adapted from Drozd & Philip ()
Quantifier
Restrictor
EVERY (e) !x [farmer’ (x,e)] or !y [donkey’ (y,e)]
Nuclear Scope
x is feeding (e) y
‘For every e, x is a (unique) farmer in e or y is a (unique) donkey in e or both, e consists of x feeding y’
.. Event quantification representation of the sentence ‘Every farmer is feeding a donkey’, adapted from Drozd & Philip ()
over half of preschool children rejected Every farmer is feeding a donkey as the description of a picture where there was one extra donkey, Philip proposed the event quantification hypothesis that errors stemmed from children misinterpreting the universal quantifier as an adverbial modifier of events, as opposed to a determiner modifier of individuals. Under this hypothesis, preschool-age children fail to construct the standard semantic representation of the sentence Every farmer is feeding a donkey, shown in Figure .. Instead, they assign an event quantification representation, shown in Figure ., wherein every event (within a specified discourse context) that involves a farmer and a donkey consists of a farmer feeding a donkey. In support of the event quantification hypothesis, Philip found that substituting a state description (e.g. Every farmer is a donkey-feeder) for an event description (e.g. Every farmer is feeding a donkey) led to a significant decrease in the number of children rejecting the sentence as a description of the picture with the extra donkey. Roeper et al. () proposed a three-stage model of the development of syntactic competence with universal quantification. According to this model, at the initial stage universal quantifiers are interpreted as adverbial modifiers, leading children to extend the scope of a universal quantifier broadly to encompass all events within a specified discourse context. This results in errors such as answering ‘no’ to the question Is every bunny eating a carrot? in relation to a picture of three rabbits, each eating a carrot, and a dog eating a bone. Roeper and colleagues referred to these errors as bunny spreading, where the scope of the
.
universal quantifier extends to all events within the discourse context, including events that involve neither a bunny nor a carrot. At the next stage, children acquire the syntactic operation of quantifier ‘floating’ wherein the universal quantifier may be separated from the material within its scope, as in the sentence The children all had pizza. At the ‘floating’ stage, children no longer make the bunny-spreading error, but are prone to make the classic quantifier spreading error. This error occurs because the universal quantifier has been lifted out of its noun phrase (i.e. the quantifier is c-commanding the whole sentence), which leaves its interpretation ambiguous. Finally, in middle childhood ( to years of age), children enter the adult stage wherein universal quantifiers are treated as determiners. Roeper et al. () further developed this account, while contrasting children’s acquisition of every and each. In an adaptation of the picture-choice task of Brooks et al. (), thirty-eight English-speaking children (- to -year-olds) and forty adults were shown three pictures of flowers in vases: a collective picture of three flowers in single vase along with two extra vases, a distributive picture with three flowers each in its own vase along with two extra vases, and an exhaustive picture with five flowers arranged in three vases (with no flowers or vases left over). Participants were asked which picture or pictures matched the sentences Every flower is in a vase and Each flower is in a vase, and to explain why. Children showed a strong preference to match the exhaustive picture with the sentence with every, whereas per cent of adults indicated that all three pictures were okay. For the sentence with each, over per cent of the children rejected the distributive picture that per cent of adults had indicated was the preferred interpretation. Taken together, the results suggested that exhaustivity (inclusion of all of the items in both sets) was a more salient aspect of the meanings of these universal quantifiers than distributivity (one-to-one correspondence); see Syrett (Chapter in this volume) for more on distributivity. Roeper and colleagues interpreted children’s justifications of their preferred interpretations as evidence for quantifier floating, as children often paraphrased the sentences by shifting the position of the quantifier, such as pointing to the exhaustive picture and saying ‘it’s the only one with flowers in every vase’. Moreover, unlike adults, per cent of the children did not distinguish their interpretations of every and each, which suggested incomplete mastery of lexical information. Street & Dąbrowska () examined individual differences in susceptibility to errors in interpreting universal quantifiers among adults, comparing performance of individuals with high versus low educational attainment on a picture-choice task. The authors observed poor performance in the group with low educational attainment, which they attributed to their lack of knowledge of the grammatical constructions (e.g. Every dog is in a basket vs. Every basket has a dog in it). The participants’ performance on the task improved dramatically after a brief training was used to provide an explanation of the sentence structures; note that a control group trained on passive sentences did not show improvements in understanding universal quantification. Using a novel version of a sentencepicture verification task to examine acquisition of universal quantification, Berent et al. () compared performance of three groups of college students: native speakers of English, deaf students, and second language (L) learners of English. Deaf speakers and L learners performed similarly and, like children, exhibited quantifier spreading errors (referred to as symmetric errors in this paper). Berent and colleagues attributed the errors to participants’ underdeveloped pragmatic knowledge of English, which was a likely consequence of restricted access to English language input.
Based on observations of quantifier spreading in college students in studies utilizing picture-choice (Brooks & Sekerina, /) and sentence-picture verification tasks (Brooks & Sekerina, ), Brooks & Sekerina suggested that errors might stem from shallow syntactic analysis, as opposed to faulty grammar. Shallow processing is a term used to describe a quick-and-dirty sentence-processing mode that generates representations ‘good enough’ for current purposes but lacking in details (Sanford & Sturt, ; Sanford & Graesser, ). When presented with an ambiguous sentence like Every flower is in a vase, the listener will often select a candidate interpretation based on contextual factors, rather than generate multiple interpretations (e.g. three flowers in one vase, three flowers in individual vases, five flowers in three vases, as in Roeper et al., ). Shallow processing resonates with Bucci’s () pragmatic processing mode in which children arrive at a semantic interpretation based on pragmatic factors, or even guessing, while ignoring syntactic information.
. C’ ,
.................................................................................................................................. In contrast to the abovementioned accounts that treat quantifier spreading as involving the misinterpretation of a universal quantifier, Crain et al. () argued that such errors are a manifestation of inappropriate testing conditions—that is, the failure to meet pragmatic felicity requirements for yes–no questions in studies utilizing the truth-value judgement task. For a yes–no question to make sense, the child should have an opportunity to consider an alternative outcome that differs from the one presented—in other words, the condition of plausible dissent should be satisfied. Crain and colleagues showed that preschool-age children who exhibited quantifier spreading (in a replication of Philip’s study) accepted the correct interpretation per cent of the time when the condition of plausible dissent was satisfied. To set up this condition, Crain et al. embedded the truth-value judgement task in a scenario in which story characters were engaged in decision-making (i.e. between drinking soda or hot apple cider after skiing). Based on preschoolers’ correct acceptance of the statement Every skier drank a cup of hot apple cider as a description of a scenario where there were leftover cups of cider, Crain and colleagues concluded that young children have full grammatical competence with universal quantification and that interpretive errors result from flaws in testing procedures. More recent work has highlighted other features of the truth-value judgement task that may inadvertently elicit errors. O’Grady et al. () observed a sharp decrease in quantifier spreading among Japanese preschool children when they were asked to act out the sentences (e.g. Every cat is biting a fish) rather than make yes‒no truth-value judgements in the typical context of a picture showing an extra object (e.g. fish). O’Grady and colleagues’ argument that the critical factor is the salience of the extra object is reminiscent of Donaldson’s () emphasis on the salience of the empty garage in her studies. Similarly, Kiss & Zétényi () have argued that quantifier spreading is an ostensive effect elicited by the features of the pictures. A line drawing is an abstraction of the depicted situation and may be perceived by the viewer as providing relevant clues for interpreting
.
the sentence. When naturalistic photographs were used instead of line drawings, children exhibited markedly lower rates of quantifier spreading, presumably because the extra objects were less salient in the photographs. Given the proliferation of findings indicating that even preschool children are highly sensitive to contextual manipulations of the truth-value judgement task, researchers have offered various pragmatic accounts of quantifier spreading (e.g. Kang, ; Rakhlin, ); see Brasoveanu & Dotlačil (Chapter in this volume) for discussion of other factors influencing the processing of quantifier scope. It is now well-established that familiarizing children with the intended domain of quantification through a storyline or other contextual support drastically decreases error rates (e.g. Crain et al., ; Philip, ). Drozd (; Drozd & Van Loosbroek, , see also Geurts, ) proposed that without such contextual support, children will fail to consider presuppositions constraining the use of universal quantifiers, and thus are prone to interpret them as weak rather than strong quantifiers (see Schwarz, Chapter in this volume). Weak quantifiers, which include vague quantifiers, such as few, some, or most, are interpreted in a context-dependent way, by factoring in expectations about what would be considered normative in a given situation (Moxey & Sanford, ). For example, the expression some would likely denote a larger amount in the context of a school purchasing some new books than in the context of an individual purchasing some new books. Unlike weak quantifiers, strong quantifiers, such as the universal quantifiers all, each, and every, do not specify a range of possible values, and must be interpreted in relation to a set under discussion. It may be too cognitively demanding for a child to think about two sets (e.g. cars and garages) to compare them or, alternatively, ‘It may never occur to a child who is presented with an exhaustive pairing task that she should be distinguishing and comparing sets’ (Drozd, : ). Rather, the child may treat the universal quantifier as if it were a weak quantifier, allowing a sentence such as Every car is in a garage to be interpreted in accordance with situational expectations (e.g. how many cars should there be given the presence of five garages).
. E-
.................................................................................................................................. In recent years, research on quantifier spreading has supplemented behavioural designs focusing on response accuracy with eye-tracking measures of visual attention. In the first such study, Minai & colleagues () tracked the eye movements of - to -year-old Japanese children and adults in the context of a sentence-picture verification task. They ran two versions of the task, varying the number of extra objects in the scene: for example, three turtles each holding an umbrella along with either three extra umbrellas or one extra umbrella. In line with previous work (e.g. Freeman & Stedmon, ), quantifier spreading occurred more often when there was only a single extra object, highlighting its salience as a distractor. Analysis of eye movements revealed that children tended to fixate more on the extra object than a control group of adults did. Note, however, that the increased fixations to the distracting object occurred prior to the onset of the sentence rather than simultaneously with sentence processing. Following Rakhlin () and O’Grady et al. (),
1.0
quan
Proportion of Fixations
0.9
npo
v p
npc
silence
0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0
500
1000 1500 Time (ms) acc 0
2000
2500
1
.. Fixations to the extra containers as a function of accuracy in sentence-picture verification for sentences of the form ‘Every apple is in a bowl’; quan = quantifier, npo = noun phrase for object, v = verb, p = preposition, npc = noun phrase for container
Minai and colleagues attributed this finding to children’s inability to ignore the salience of the extra object due to undeveloped skills in executive control of visual attention. Sekerina & Sauermann (, ) identified a signature pattern of visual attention associated with quantifier spreading in adult heritage speakers of Russian and Russian children performing the sentence-picture verification task. Sekerina et al. () extended these findings to school-age children (forty-one native English speakers, ages to years). On trials designed to elicit quantifier spreading (i.e. with containers and objects in partial one-to-one correspondence, such as apples in bowls), children performed at chance (.% correct) in verifying Every apple is in a bowl as a description of a picture showing three apples each in a bowl with two extra bowls. As shown in Figure ., quantifier spreading was associated with increased fixations (visual attention) to the extra bowls in the picture relative to correct responses. Note that the increased fixations appeared after the children had processed the quantified noun phrase (every apple): that is, the fixations were timelocked to the occurrence of the verb phrase (is in a bowl). Moreover, incorrect responses were associated with faster reaction times than correct responses—consistent with the view that quantifier spreading occurs when individuals engage in shallow processing of sentence structure. Sekerina and colleagues () suggested that quantifier spreading is associated with cognitive overload in real-time sentence processing, such that participants whose processing resources are taxed exhibit lesser control of visual attention, concurrently with exhibiting the error.
. C
.................................................................................................................................. For half a century, researchers in the fields of developmental psychology and linguistics have puzzled over young children’s difficulties in grasping universal quantification, focusing on their tendency to extend the domain of a universal quantifier beyond the noun
.
phrase it modifies. Using a broad range of methodologies, researchers have emphasized children’s reliance on contextual and pragmatic factors to produce what appear to be sensible, if not logical, responses to questions about set relationships that constitute violations of one-to-one correspondence. Despite a great deal of progress in exploring this topic, the extant studies have tended to rely on similar cross-sectional designs that fail to provide a comprehensive account of individual differences in sentence processing skills over the course of language development. Given evidence that adults as well as children are susceptible to quantifier spreading, future studies should place greater emphasis on online measures that assess the dynamics of sentence processing in real time. Future work using longitudinal designs is needed to evaluate how the development of pragmatic competence influences sentence processing and interpretation of universal quantification.
A The authors thank Elizabeth Che for creating the original artwork for this chapter.
......................................................................................................................
......................................................................................................................
. I
.................................................................................................................................. G adjectives and scalar expressions more generally have been one of the most active topics of research in formal semantics over the past twenty or so years. This period has been characterized not only by important theoretical progress but also by an increasing use of experimental approaches in service of theoretical goals, and as such this area offers a wealth of case studies illustrating the benefits of experimental research in semantics. This chapter is intended to offer a survey of recent work in experimental semantics relating to the topics of gradability, scalarity, and adjective meaning, with a view to informing future research in this area. There is a long tradition of research on adjective meaning and its acquisition from the perspective of cognitive and developmental psychology as well as psycholinguistics (see e.g. Eilers et al., ; Bartlett, ; Rips & Turnbull, ; Ebeling & Gelman, ; Sedivy et al., ). I do not attempt to do justice to this tradition here, but rather focus on work that is more explicitly oriented towards issues in formal semantics. In order to achieve more general relevance, the material that follows is organized not primarily by the specific research questions investigated, but rather according to the role that experimentation has played in the process of theory development and evaluation. Section . considers research that has served to strengthen and reinforce intuition-based data on which theoretical accounts are based. Section . moves on to cases where experimental approaches have yielded data beyond what is accessible to intuition and introspection, thus offering new power to test and refine formal theories. Finally, section . considers research that is not guided by a specific theoretical proposal or hypothesis, but rather serves to map out the empirical landscape with respect to a particular phenomenon, potentially raising new questions of theoretical importance. Under each of these broad headings, one or more specific research issues have been selected as case studies of the potential application of experimental techniques. Along the way, the specific methodologies used are described in some detail; this discussion is expanded in section ., which offers some more general observations on appropriate methodologies for research on adjective meaning, and suggests points that the researcher should keep in mind.
. R -
.................................................................................................................................. Traditionally, the semanticist’s data are sourced through introspection and perhaps informal elicitation methods, and are based on intuitions regarding the acceptability of linguistic expressions and their meaning, the latter diagnosed via judgements on allowable contexts of use, entailments, and suitable paraphrases. Such approaches still form the backbone of work in formal semantics, but it goes almost without saying that data collected in these ways are subtle, open to dispute, and subject to possible influence from the scholar’s own theoretically-driven perspective. Thus, as experimental methods have become more accessible and widely accepted in formal linguistics, it has become increasingly common for researchers in semantics to look to the results of structured experiments with ‘ordinary’ speakers to provide additional substantiation of the data on which theories are based. In what follows I discuss one case study where a formal theory of adjective meaning has been strengthened in this way.
.. Scale structure and the absolute/relative distinction Probably the most influential recent theory of gradable adjective meaning is the scalar theory of Rotstein & Winter (), Kennedy & McNally (), and Kennedy (), according to whuch gradable adjectives lexicalize measure functions that relate individuals to degrees on scales. This is exemplified by the following lexical entry for the adjective tall:1 ()
⟦tallhe;di ⟧ ¼ λx:μHEIGHT ðxÞ
On this approach, all gradable adjectives have the same basic semantic form. Adjectives differ, however, in the structures of the scales they lexicalize, specifically in whether or not these have maximum and/or minimum points. This gives rise to four possible scale types, on the basis of which gradable adjectives can be divided into two broad classes: absolute gradable adjectives, which lexicalize totally or partially closed scales, and relative gradable adjectives, whose corresponding scales are open on both ends. () ●―――――――――● CLOSED (e.g. transparent, opaque, full, empty) ●――――――――― > LOWER CLOSED (e.g. wet, dirty, rough) every, ?every > a
If this is the case with QR in ACD, we would not be able to access the intended interpretation of (), and would only be left with the embedded clause reading of (). However, observations by Wilder (, ) and Fox (), and further theoretical proposals about successive cyclic movement of QR through the syntactic structure to allow such readings by Cecchetto (), have strongly suggested that this is not, in fact, the case. This empiricallybased disagreement with earlier theoretical assumptions thus leaves room for experimental work to provide relevant informative evidence about constraints on QR, and therefore possible limits on available interpretations of ACD sentences. ()
Lady Cora expects that everyone will fancy Lady Rose that Lord Grantham does. 〈expect that those people will fancy Lady Rose〉
()
Spratt said he noticed every imperfection that Danker did. Embedded: 〈noticed〉 Matrix: 〈said she noticed〉
4
For extensive discussion of the linguistic properties of NPIs and the environments that license them, see Xiang (Chapter in this volume).
-
This review has uncovered a host of characteristics and puzzles connected to the phenomenon of ACD. First, ACD is a form of VPE in which the site of ellipsis is contained in its antecedent. To resolve this situation, the QNP must undergo semantically-motivated QR to a position outside of the antecedent VP (most likely adjoined to vP). Second, this landing site is not only above the VP, but below the subject position, a fact that we can diagnose with binding relations and licensing of NPIs. Moreover, the landing site is not restricted to one location: ambiguous ACD sentences are such because there are multiple landing sites, correlated with different interpretations. Finally, there are cases where the accessibility of a landing site has been questioned, based on a priori assumptions about QR, but ACD sentences have the potential to provide data bearing on such claims. In the sections . and ., I will summarize a growing number of experimental studies designed to tap into precisely these aspects of ACD. In section ., we will evaluate what we know about children’s interpretation of ACD sentences, and what the experimental results say about the mechanisms that are active in the child grammar. In section ., we will evaluate evidence from studies investigating adults’ interpretation of ACD sentences, which have direct bearing on linguistic theory. In the first case, evidence from ACD sentences provides us with a basis for questioning the apparent clause-boundedness of QR. In the second case, evidence from reaction times and acceptability ratings is taken to bear on theoretical assumptions regarding the nature of ACD resolution and our linguistic representations. The combined results across all tasks illustrate how experimental research on ACD sentences provides us with a unique opportunity to investigate the composition of our grammar, and push the implications and limits of linguistic theory.
. I ACD
.................................................................................................................................. One of the main goals that researchers of language child acquisition have is to determine the level of continuity in language acquisition and development. To that end, we seek to ascertain the interpretations that children assign to specific linguistic constructions as a means of crystallizing statements about the nature of their grammar, identifying the mechanisms they deploy for sentence interpretation, and determining the division of labour between what is learned and what is not. For reasons outlined in the Introduction, ACD provides us with ideal testing grounds.
.. Evidence for QR in child grammar Given that QR is a key component of the adult grammar (under the theoretical framework being assumed here), one might ask whether children have QR as a part of their grammar. While a host of studies have examined children’s ability to interpret quantificational sentences (e.g. Lidz & Musolino, ; Miller & Schmitt, ; Musolino & Lidz, ), few (if any) of these studies have provided unambiguous support for the presence of QR in child grammar. For example, the ability to interpret a scopally ambiguous sentence such as (), which could mean either that every horse does not have the property of jumping over
the fence or that it is not the case that every horse has this property, could arise from reconstruction of the QNP into the VP-internal base position. The ability to access the inverse scope reading of () might also arise via a choice function, rather than scope shifting (e.g. Reinhart, ; Winter, ). ()
Every horse didn’t jump over the fence.
∀ > ¬, ¬ > ∀
()
The boy didn’t blow out a candle.
¬ > ∃, ∃ > ¬
ACD, however, requires QR for resolution. So if children are able to successfully interpret sentences with ACD, this provides us with more reliable evidence that QR is part of the child grammar. This was the starting point for Syrett & Lidz’s () investigation of ACD in language acquisition. Four-year-olds and adult controls participated in a Truth-Value Judgement Task (TVJT; Crain & McKee, ; Crain & Thornton, ).5 Because all of the acquisition experiments reported in this section—and some of the adult experiments reported in section ..—make use of these age groups and this methodology, I will take some space here to lay out the details of such a paradigm. In a TVJT, there are typically two experimenters: one tells the child a series of short stories using toys and props or computer images, as a puppet (played by the second experimenter) watches alongside. The child is told that the puppet is learning, and that the child’s job is to watch and listen to the story carefully along with the puppet, so that he can help the puppet learn. At the end of each story, the puppet makes a statement about what she thinks happened in the story. This statement includes the target construction—in this case, a sentence with ACD. The child assesses whether the puppet’s statement was accurate or not, given the context presented in the story. That is, the child must judge the truth-value of the puppet’s sentence, but instead of responding ‘true’/‘false’, the participant responds by saying ‘right’/‘wrong’ or ‘yes’/‘no’. If the puppet is right, she gets a special treat (e.g. a cupcake or cookie), and if she is wrong, she gets something lesser (e.g. milk or a banana); either way, she gets something. Participants are encouraged to provide a justification for the response, but with children, this is not mandatory. In Syrett & Lidz (), participants were randomly assigned to one of four experimental conditions, depending on the context and the linguistic construction they heard. The construction was either an ACD sentence or a sentence with coordinated conjunction. This contrast was chosen based on previous evidence that when children appear to misinterpret a relative clause construction (which ACD contains), they seem to default to conjunction (Tavakolian, ).6 There were multiple trials for each condition, as there were in all TVJT experiments reviewed here.
5
See also Gordon () for further details about this methodology. Based on these misinterpretations by children, as exhibited in act-out tasks, early researchers concluded that children could not actually represent relative clauses in their grammar (Tavakolian, 1981). However, this conclusion based on performance data was unwarranted, since children have no difficulty comprehending relative clauses when the felicity conditions governing the use of a relative clause are satisfied, and as control items from the TVJT experiments reported here indicate (see, in particular, Hamburger & Crain, 1982). 6
- ()
Miss Red jumped over every frog that Miss Black did.
()
Miss Red jumped over every frog, and Miss Black did, too. Conjunction)
(ACD)
(Coordinated
Each of these sentence types was presented in one of two contexts. In one, there was one set of objects (e.g. frogs that two characters had to jump over), and in another, there were two sets (e.g. two sets of frogs, one for each character to jump over). In the one-set condition, each character attempts to jump over all of the frogs in succession. However, in the end, neither one does so, and consequently, neither wins a prize. In the second condition, each character jumps over all of his frogs (but none of the other character’s frogs), and each therefore receives a prize. In the one-set condition, the ACD sentence is true, and the conjunction sentence false, while in the two-set condition, the ACD sentence is false, and the conjunction sentence true. Accordingly, both child and adult participants consistently accepted the ACD sentence in the one-set condition, and rejected the conjunction sentences, and demonstrated the reverse pattern in the two-set condition. The only mild surprise was that some adult controls were not as willing as the children were to restrict the domain of quantification in the two-set condition, and rejected the sentence, because the characters had not jumped over every frog. The results of this experiment thus demonstrate that -year-old children, like adults, correctly interpret ACD sentences, and therefore that they must have the QR mechanism in their grammar. Presented with evidence that children can perform QR and interpret ACD sentences, one might further probe just how adult-like their QR and ACD resolution are. That is, do they target the same landing sites as adults? The next two experiments reported in this section were designed to address precisely this question.
.. Evidence for the landing site of QR in child grammar Earlier, we observed that the landing site of QR in ACD sentences must be below the subject position. There are a few consequences to targeting this position. First, the QNP lands below negation, thereby allowing for NPIs in the ACD construction to be licensed. Second, ACD bleeds Principle C violations when there is a pronoun in the object position and an R-expression in the relative clause, since the R-expression raises to a position outside of the VP and therefore higher than the object, allowing for co-construal. Finally, ACD feeds Principle C violations when there is a pronoun in subject position and an R-expression in the relative clause, since even after movement, the QNP remains below the subject, and the pronoun therefore remains in a position to c-command the R-expression, preventing co-construal. Kiguchi & Thornton () took these observations from Fox and Merchant, and sought to determine how children interpreted ACD sentences in which binding principles were invoked. Their test sentences across a set of experiments had structures as indicated in ()‒().
()
a. Dora gave himj the same color paint the Smurfj’s father did. b. Dora [the same color paint the Smurfj’s father did ] [VP gave himjt]
()
a. *The Mermaid baked himi the same food that Cookie Monsteri did. b. *The Mermaid [the same food that Cookie Monsteri did ] [VP baked himi t]
()
a. *Hei jumped over every fence that Kermiti tried to. b. *Hei [every fence that Kermiti tried to ] [VP jumped over t]
In each of the scenarios, the experimenters manipulated the story so that there were two possible antecedents for the pronoun: the character denoted by the target R-expression in the sentence and another salient character of the same gender (e.g. Mickey Mouse in (), Jabba the Hutt in (), or Cookie Monster in ()). Moreover, when the puppet delivered the target sentence, she mentioned the key players in the scenario, with the possible co-referent from the sentence listed last. In the ungrammatical sentences, this order of mention was intended to highlight the possibility of this character being co-construed with the pronoun. Children responded correctly to control ACD sentences without pronominal reference (e.g. Kermit found Emily the same color egg that Gonzo did) at a rate comparable to that of adults: close to per cent of the time. Children accepted sentences like (), but rejected those like () (which has a Principle B violation) and () (which incurs a Principle C violation). () is the key sentence for demonstrating that the landing site for QR is below the subject, while all three sentence types demonstrate something about how children perform QR and interpret elided material.7
.. Evidence for multiple landing sites for QR in child grammar Having established that children can correctly interpret ACD sentences, and moreover that they appear to target the same landing site as adults, we can probe their grammar further and determine whether children have the same range of accessible landing sites as adults do, or whether they are restricted to, for example, the closest landing site (arrived at via the shortest move). In Syrett & Lidz (, ), the goal was to present child and adult participants with ambiguous ACD sentences, where the ambiguity stemmed from the ACD being embedded within multiple VPs. In one set of experiments, ACD was embedded in a non-finite clause, with matrix verbs that selected for an infinitival complement (want, need, ask, invite), including possible restructuring and non-restructuring verbs.8 An example is shown in (). 7
But see discussion in Syrett & Lidz () concerning how a conjunction interpretation and pragmatic reasoning could yield a similar pattern. Thus, the combination of Kiguchi & Thornton () and Syrett & Lidz () provides evidence that children correctly interpret ACD sentences, and therefore that QR is part of child grammar. 8 See discussion in Kennedy () related to claims by Hornstein ().
- ()
The cowgirl needed to jump over every frog that the old cowboy did. Embedded: . . . every frog that the old cowboy 〈jump over〉 Matrix: . . . every frog that the old cowboy 〈need to jump over〉
In a second series of experiments, the matrix verb took a tensed clause as a complement, resulting in ACD being embedded in a finite clause, as in (). There is thus a crucial difference between the two cases in terms of a possible tensed clause boundary present in the latter case. This point will become relevant shortly, and will be taken up again in section . in a series of studies conducted exclusively with adults. ()
Clifford said that Goofy read every book that Scooby did. Embedded: . . . every book that Scooby 〈read〉 Matrix: . . . every book that Scooby 〈said that Goofy read〉
In each case, the target sentence was presented in one of two between-subject conditions: one that made the embedded reading true and the matrix reading false, and one that made the embedded reading false and the matrix reading true. The crucial question in both cases was whether children could access the matrix reading, since this would then indicate that they can perform so-called ‘long QR’ and are not restricted to the first possible landing site and can QR to a position outside of the matrix VP (Larson & May, ; Kennedy, ). When children were presented with sentences like (), where ACD is embedded in a non-finite clause, in either the ‘embedded’ or ‘matrix’ context, they—like adults—were able to access either of the two readings. Furthermore, participants either accepted the target sentence, justifying their response with the reading made true in the context, or rejected it, appealing to the reading that was made false. Thus, children are able to perform both ‘short’ and ‘long’ QR. This conclusion was further bolstered by a follow-up study conducted by Sugawara et al. (), in which a verb and tense mismatch between the matrix and embedded clause disambiguated the readings. When all participants were presented with sentences like the one in (), the adults and the children showed a very different pattern. Adults, for the most part, patterned as one would expect if QR were subject to a clause-boundedness constraint: they accepted the sentence in the context that made the embedded reading true (and the matrix reading false), and rejected it in the context that made the matrix reading true (and the embedded reading false). By contrast, children were prone to rejecting the target sentence, flying in the face of expectations and the principle of charity. This pattern is not only curious, because it seemed as though children are being contrary, but because if children really are overgenerating interpretations of these ACD sentences, and are too lenient with the QR operation, it seems nearly impossible to arrive at a learnability story whereby they might encounter evidence that would allow them to prune away the grammatically barred readings of these ACD sentences, which are rarely encountered in adult speech, let alone childdirected speech.9 Thus, we are faced with a classic Poverty of the Stimulus argument. 9
One need not rely on this precise kind of ACD construction, of course, to learn the limits of application, but then it is extremely challenging to identify an analogue with another linguistic construction that would allow the learner to arrive at the adult state. (See related discussion across Pullum & Schulz (2002) and Legate & Yang ().)
However, a closer look at the adult responses revealed that a small subset of adults were patterning in the same way, consistently accessing the matrix reading of sentences in which ACD was embedded in a finite clause. Thus, for these adults, it would appear that QR was not subject to a clause-boundedness constraint. Consequently, there would be no learnability story to tell: children and these adults have in common that QR is not subject to clause-boundedness. If this is the case, then two questions arise: (a) what licenses the matrix reading of these ACD sentences (i.e. what allows QR to target a matrix-VP external site), and (b) what was preventing the vast majority of adults from accessing the matrix reading in such sentences? These questions are pursued in section ..
. I ACD
..................................................................................................................................
.. ACD as evidence against clause-boundedness of QR If QR is a covert form of Ā movement, on a par with wh-movement, it seems unexpected that QR should be subject to a clause-boundedness constraint, whereas wh-movement is not. A simple comparison between the two cases (as in ()‒()) highlights this difference. In (), the wh-phrase can undergo successive cyclic movement out of the object argument position of a verb in a finite embedded clause to land in a position above the subject, whereas in (), the universally quantified phrase in object position seems to be unable to take wide scope over the indefinite in subject position. ()
What did a footman [VP say [CP that Mr Carson inspected t]]?
()
A footman [VP said [CP that Mr Carson inspected every place]]. a > every, *every > a
However, enough counterexamples have been presented over the years, among them those in () and (), to lead to the conclusion that this is not a hard and fast constraint. (See also Farkas & Giannakidou, .) ()
A quick test confirmed [CP that each drug was psychoactive].
()
A doctor will make sure [CP that we give every new patient a tranquilizer]. Reinhart ()
VanLehn ()
Perhaps more to the point here, the matrix reading of sentences in which QR must occur out of a finite clause appears not only to be accessible, but sometimes forced, as in () (a comparative construction in which the degree phrase undergoes QR) and ()‒(), ACD sentences. ()
John thinks that Mary is taller than Bill does. 〈think that Mary is d-tall〉
Wilder ()
-
()
I expect (that) everyone will visit Mary that you do. Tiedeman (), Fox () 〈expect those people to visit Mary〉
()
John said that you were on every committee that Bill did. 〈say that you were on those committees〉
Wilder ()
Faced with these suggestive examples paired with the data from children and adults in Syrett & Lidz (), Syrett (a) undertook a series of studies with adult participants in order to probe the availability of the matrix reading of sentences in which ACD is embedded in a finite clause, as in (). In this case, the matrix reading would be Woody said he jumped over every frog that Jessie said that he jumped over. ()
Woodyi said hei jumped over every frog that Jessie did.
To facilitate the availability of this reading, the experimental contexts were manipulated to support the discourse conditions that would make the matrix reading felicitous. First, the event denoted by the embedded verb (e.g. jumping) was never shown. Second, the act of reporting was made especially salient, since both characters mentioned in the sentence reported their actions to a third party (Buzz Lightyear), and these speech events were accompanied by a speech bubble above the animated images on a computer screen. Next, instead of saying, ‘I know what happened!’ at the end of each story, the puppet exclaimed, ‘I know what X said!’ In addition to manipulation of the context, the target linguistic sentence was also manipulated to facilitate processing. First, the embedded CP did not have a that complementizer. (However, a follow-up experiment demonstrated that even with the complementizer present, the matrix reading was robustly available.) Second, the embedded subject was a pronoun in lieu of a full name. Next, contrastive focus was placed on the two main characters (see Wilder, , for a related observation). Finally, the ACD constructions were extended beyond just the universal quantifier to those with the same, as in (). ()
Jakei said (that) hei found the same treasure as Captain Hook did.
The response pattern obtained across experiments, accompanied by full-fledged justifications, leaves no doubt that the matrix reading is indeed available to adults, and therefore that QR can cross a tensed boundary. The percentage of matrix readings ranged from per cent to per cent at the lowest, and per cent to per cent at the highest. Thus, speakers allow not only ‘long QR’ out of non-finite clauses, but also finite clauses as well, and do so robustly when the discourse conditions support the matrix reading. This ability to access the matrix reading of ACD embedded in a finite clause is actually predicted under Cecchetto’s () proposal, in which QR proceeds in a successive cyclic fashion, motivated by Scope Economy and constrained by the Phase Impenetrability Condition (Chomsky, /), whereby each sub-link of the movement is semantically motivated and crosses just one instance of v or C (each a phase). In fact, under Cecchetto’s account, once the QNP lands outside the matrix VP, it could raise even higher, again as long as the QR is semantically motivated and respects movement constraints. The
prediction is that if there is an indefinite in subject position that could interact scopally with the QNP, and generate a distinct semantic interpretation, then this would license movement. This prediction is borne out. Syrett (b) followed a methodology similar to the previous TVJTs, and similar to Syrett (a) in particular. Participants were presented with sentences such as () in a context that supported an ‘extra wide scope’ reading of the QNP over the subject indefinite and a matrix reading of the VPE. ()
Someonei said hei could jump over every frog that Jessie did.
In the story corresponding to this target sentence, Jessie claims to Woody she can jump over a set of frogs. Following her departure from the scene, a group of young boys, who have overheard her boasts, enters. Each one approaches one of the frogs and makes a claim that he can jump over that frog. When Jessie returns, Woody reports back to Jessie what the boys have said. Irate, she storms off. The puppet then recaps, ‘I know why Jessie was so mad! Someone said he could jump over every frog that Jessie did.’ Adult participants accepted the sentence over per cent of the time, with close to per cent of the written justifications provided unambiguously pointing towards the extra wide scope reading.
.. ACD and possible evidence for theoretical frameworks There is robust evidence that both young children and adults are able to provide offline judgements indicating their interpretation of ACD sentences, and the role of the grammar, the discourse, and the processor in licensing various interpretations. We might, then, ask whether on-line data—that is, data that are collected from real-time sentence processing studies—can provide us with further insight into the nature of the representations of ACD sentences. This was the goal of Hackl et al. (). Hackl et al. () presented participants with sentences as in ()‒() in a self-paced reading study. The factors that were manipulated included the head of the object phrase containing the relative clause (the/every), the presence of a lexical verb or a site of ellipsis (e.g. funded vs. was), and (in Experiment ) whether the intended antecedent for ellipsis was a small or large VP (e.g. did = treat . . . or was reluctant to treat . . . ) The authors predicted that, since the universal quantifier would have to undergo QR for reasons of type mismatch anyway, the semantically-motivated movement would facilitate processing in the case of ACD relative to the definite determiner, but that there would only be an advantage in the case of the embedded VP (which would implicate short QR) and not with the matrix VP (which would require long QR, and involves a more complex VP to process). ()
The understaffed general hospital was negotiating with {the/every} doctor that the nonprofit medical organization {funded/was} in order to arrange for free vaccination clinics.
()
The doctor was reluctant to treat {the/every} patient that the recently hired nurse {admitted/did/was} after looking over the test results.
-
The results for Experiment demonstrated an initial slowdown for the sentences with a lexical verb in the ‘spillover region’ immediately following the main verb or auxiliary in the matrix sentence, followed by a crossover with a slowdown for the ellipsis cases. In both experiments, the ‘determiner’ the sentences incurred longer processing time than the quantificational every sentences, and the results yielded a determiner-ellipsis interaction in Experiment and a determiner-ellipsis size interaction in Experiment (with the-was sentences taking longer). Hackl et al. () thus concluded that the presence of the universal quantifier facilitates resolution of ACD downstream, and that this finding supports a QR approach to ACD resolution. However, the interpretation of these findings was subsequently challenged by Szabolcsi (), who has pointed out that under a variable-free approach, the slowdown with the can be accounted for, because a default ‘individual-forming’ interpretation of the definite determiner will not work for the interpretation of the sentence. The reinterpretation will trigger a reprocessing of the preceding material, and result in a slowdown of reading time. Moreover, if the default processing interpretation for the antecedent of ellipsis is a small VP antecedent, then this preference may account for the RTs observed in Hackl et al.’s Experiment . Gibson et al. () also took issue with the interpretation of Hackl et al.’s conclusion that the results support a QR theory of ACD resolution, arguing that the target sentences tipped the scales in favour of a slow-down with the, since in such cases a speaker would normally provide an additional expression highlighting the sameness of the action performed on the matrix and relative clause objects, such as same or also (their ‘sameness’ hypothesis), and that the most natural production in the target sentences would be one where there were two different lexical verbs. In Gibson et al.’s experiments, they manipulated the type of determiner (every vs. the), the presence or absence of the words same and also, and the presence or absence of a full VP and whether or not it was different (e.g. negotiate, funded), as illustrated in an abbreviated form in (). They also excluded the spillover region, since they were obtaining acceptability ratings. ()
a. The understaffed general hospital was negotiating with {every/the (same)} doctor that the non-profit medical organization {(also) was/funded/was negotiating with}. b. The choreographer evaluated {every/the (same)} ballerina that the lead dancer {did/evaluated}.
Across tasks, they replicated the determiner-ellipsis interaction, but further demonstrated a significant effect of verb type (with the different verb increasing acceptability at a comparable rate for both every and the), and an effect of the presence of same and also, with both terms increasing acceptability in the sentences with the, and the presence of also lowering acceptability with sentences with every. (Same cannot be used alongside every.) Thus, the authors concluded that the original processing data do not unambiguously signal a QR analysis of ACD sentences, and that the effects could arise via other means. Clearly, this lively exchange regarding the processing and acceptability judgements of ACD sentences highlights the reasons why ACD is a hotbed of research, and how the results arising from experimental work in this area may directly bear upon theoretical proposals about the nature of our linguistic representations.
. C ACD
.................................................................................................................................. The experimental evidence collected over a range of experiments demonstrates the following. First, children as young as years of age can arrive at the correct interpretation of ACD sentences, thereby indicating that the QR mechanism is part of the child grammar. Second, when children perform QR to resolve ACD sentences, they target a landing site that is outside of the antecedent VP but also below the subject, and are not restricted to the nearest landing site. Given an ACD sentence made ambiguous by ACD being embedded in multiple VPs, children and adults are able to access both candidate interpretations. This finding holds not only for ACD embedded in non-finite clauses, but also for ACD embedded in finite clauses—but in the latter case is dependent upon discourse and processing factors being controlled for successfully. Finally, on-line processing and offline acceptability tasks with ACD sentences allow us to obtain evidence that bears potentially not only on how we represent and resolve such configurations, but also on the theoretical framework we adopt. Future research on ACD should further probe the limits of ACD configurations and interpretations, investigate the interaction of wh-words and ellipsis in ACD sentences, examine the role of binding principles in ACD configurations, and determine at a more fine-grained level what our real-time processing of ACD sentences says about the nature of our grammar and the role of extra-grammatical factors.
A This writing of this chapter was made possible by startup funding from Rutgers, The State University of New Jersey—New Brunswick.
......................................................................................................................
- ......................................................................................................................
. I
.................................................................................................................................. C a scenario in which a speaker of English came to believe that Donald lied in some debate. In this scenario, she has the straightforward option to utter a fairly simple sentence such as (a). However, in certain cases she also may opt for using an it-cleft such as (b) to convey the very same truth-conditional content. Then, a natural question for a linguist to ask is just what the difference in meaning between these two forms is. ()
a. Donald lied at the debate. b. It was Donald who lied at the debate.
The literature has distinguished three inferences which may typically be associated with such sentences. These are given in () as rough paraphrases. ()
a. Canonical inference: Donald lied at the debate. b. Existential inference: Somebody lied at the debate. c. Exhaustivity inference: Nobody other than Donald lied at the debate.
From these three inferences, the canonical one is the one which is the least disputed. Both (a) and (b) entail (a). Moreover the canonical inference is—in some sense—the main proffered content of both sentences, or following the terminology of Simons et al. () and Tonhauser et al. (), their at-issue content. After all, the most natural way to understand both agreeing and disagreeing with these utterances is to relate agreement and disagreement to the inference (a). Hence, if some speaker reacts to (a) or to (b) by (a) or (a), she is understood as conveying at least the messages specified in (b) and (b) respectively. ()
a. I agree. b. The speaker believes that Donald lied at the debate.
()
a. No way! b. The speaker does not believe that Donald lied at the debate.
As far as the existential inference is concerned, it is often claimed that (a) merely entails (b) while (b) presupposes (Horn, ) or conventionally implicates it (Halvorsen, ). One particularly clear test showing this point comes from Dryer (). If a structure presupposes existence it should not be compatible with nobody, as shown in (a). Also, it should be incompatible with a contextual question that contradicts the alleged presupposition, as shown in (b). ()
a. A: Who lied at the debate? B: ??It was NOBODY who lied at the debate. B': NOBODY lied at the debate. b. A: Did anyone lie at the debate? B: ?It was DONALD who lied at the debate. B': DONALD lied at the debate.
However, such evidence is mainly limited to so called focus-background clefts, also known as topic-clause clefts (Hedberg, ) and stressed-focus clefts (Prince, ), in which the focus is on the cleft pivot and the cleft-relative encodes known or topical information. Yet there is another type of cleft often discussed in the literature and known as topic-comment clefts, comment-clause clefts (Hedberg, ; Delin & Oberlander, ) or informative presupposition clefts (Prince, ), exemplified in (). It is not so clear whether such clefts come with an existential presupposition as well, cf. Hedberg () for some recent discussion. Consider the web-example in (), for which it is really hard to make the argument that the existential inference of the marked it-cleft is somehow (even pragmatically) presupposed; after all, from the four predications about Trump in the example, the last one is most likely to be new information in the given context, yet it is the last one which appears in a cleft relative. We will assume here, for the sake of explicitness, that at least focus-background clefts have an existential presupposition which distinguishes them from canonical sentences. We will return to this problem in section .. ()
After the GOP national convention, he chose to continue attacking Republican rivals. He chose to attack the Gold Star family of a fallen soldier; and he chose to hire as campaign CEO Steve Bannon, the former head of Breitbart News whose controversial support of the ‘alt-right’ will keep Trump on the defensive on the issue of racism. And it was Trump who recently chose to divide his own forces—as Custer divided his forces—by ‘softening’, ‘hardening’, retreating, denying and delaying on his main line of attack: the issue of immigration and deportation. http://www.wbur.org/politicker////commentary-trumps-last-stand
When it comes to the exhaustivity inference, the situation is much less clear. It is certainly clear that the canonical sentence (a) may give rise to an exhaustivity inference in certain contexts, for instance as an answer to a wh-question, as in (), but this inference is fairly
-
weak in English and can easily be cancelled. Hence, there is a wide agreement that for canonical sentences exhaustivity is a conversational implicature, even if in light of recent developments a lot of disagreement exists about how it should be derived, cf. Chierchia () and related research. It is also fairly clear that clefts are in some sense more exhaustive than canonical sentences. For example, asking a who-else question as followup to an it-cleft is significantly less acceptable than after a canonical sentence, as shown in the contrast in ()‒(), cf. Destruel et al. (). Certainly, A’s follow up in () somehow seems to suggest that she is challenging B’s answer in some sense, whereas this does not seem evident in (). At the same time, it is clear (and experimental research has provided ample evidence for the fact) that the exhaustivity in it-clefts also significantly differs from the exhaustivity encoded by exclusive particles such as English only. This is witnessed in (), in which A’s follow up clearly challenges the statement B made. ()
A: Who lied at the debate? B: DONALD lied at the debate. A: Fine, and who else lied at the debate?
()
A: Who lied at the debate? B: It was DONALD who lied at the debate. A: ?Fine, and who else lied at the debate?
()
A: Who lied at the debate? B: Only DONALD lied at the debate. A: ??Fine, and who else lied at the debate?
Yet it is entirely unclear how the exhaustivity inference of it-clefts comes about, what exactly it amounts to, and what its status in terms of entailments/pressuppositions/ implicatures or at-issue/not-at-issue meaning is supposed to be. This chapter intends to give a critical overview of this discussion. In section ., I will give an overview of the most important positions on this issue that have been adopted in the literature. In section ., I discuss the most important experimental studies on cleft exhaustivity. Finally, in section ., I attempt to discuss the findings from a broader theoretical and empirical perspective. The chapter ends with a brief conclusion. Before concluding this introductory section, I wish to mention that various languages employ various structures which have been suggested to be similar to it-clefts. For instance, pre-verbal focus in Hungarian (also known as structural or moved focus) is quite often equated to it-clefts in meaning, and most specifically in terms of exhaustivity; see prominently Kiss () and many researchers ever since, for instance Horváth (); see Zimmermann & Onea () for an overview. Indeed, similarities exist. However, differences also exist. The latter even applies to languages with very similar forms, like c’est-clefts in French (Destruel, ) and es-clefts in German. To give one example, in Hungarian it is not possible to combine the pre-verbal focus with the anti-exhaustive additive particle is (‘too’), whereas in German there are many attested examples (see Boell, , for a recent study). In the literature, theories about clefts in one language are often assumed to apply to clefts in other languages too. Such claims are often somewhat problematic, but at the moment we simply lack enough studies and detailed theoretical analyses of structures in
individual languages to be able to easily overcome this problem. Destruel et al. () is one study which seriously considers cross-linguistic variation in the theory of it-clefts. Given the length and scope of the current chapter I will not be able to do justice to cross-linguistic variation. While this chapter is mainly about it-clefts in English, it will contain some experimental evidence from other languages without a detailed discussion about potential problems when applying those findings to English.
. T
.................................................................................................................................. In the classical semantic/pragmatic literature, there are three main types of inferences that may associate with some expression E. E may simply entail the inference P under discussion as part of its at-issue content, E may presuppose or conventionally implicate P, which makes it part of its not-at-issue content, or E may conversationally implicate P, which makes it not part of E’s semantic content. While the distinction between these categories is not always entirely straightforward and gets more and more blurred in the most recent literature given new data and experimental studies (Schwarz and Breheny, Chapters and , respectively, this volume), it is very surprising to find that for the exhaustivity inference of it-clefts, indeed, all three positions have been advocated in the literature.
.. At-issue content In the early literature, mainly in Bolinger () and Atlas & Levinson (), as well as in Szabolcsi () for Hungarian pre-verbal focus (often treated on a par with it-clefts as mentioned in section .), it has been claimed that exhaustivity is simply part of the entailed (at-issue) content. Potentially the clearest piece of evidence for such an account is that exhaustivity, at least in some cases, seems to interact with negation. Hence, many speakers would accept an utterance like () (from Szabolcsi, ) both in Hungarian and in English, whereas significantly fewer people tend to accept (). If the cleft or the preverbal focus in Hungarian did not encode exhaustivity as part of their at-issue content, the negation operator could not interact with this inference (as has been known at least since Karttunen, ), and one would expect the sentence in () to be as self-contradictory as (). ()
Nem Péter aludt a padlón, hanem Péter és Pál. Not Peter slept the floor-on but Peter and Paul. ‘It was not Peter who slept on the floor but Peter and Paul.’
()
Peter did not sleep on the floor. Peter and Paul did.
Still, such examples are rare in naturally occurring discourse and to the extent that they exist at all, they seem to be very limited in usage. They resemble classical examples of metalinguistic negation (Horn, : ‒), such as in (), in which negation does not seem to attach to the at-issue truth conditional content of the respective sentences but rather to implicatures or connotations.
- ()
a. Ben Ward is not a black Police Commissioner but a Police Commissioner who is black. b. I’m not his daughter–he’s my father. c. SOME men aren’t chauvinists–ALL men are chauvinists.
Moreover, Horn () has provided compelling evidence against this kind of analysis, among others, by considering embedding sentences with exclusives vs. it-clefts as in (). If it-clefts indeed encoded exhaustivity as plain at-issue content, the continuation with but would, contrary to the fact, not be redundant in (a). As opposed to this, embedding a sentence with an exclusive in the same type of examples is entirely fine precisely because exclusives encode exhaustivity as part of their at-issue content, as shown in (b). ()
a. ??I knew Donald talked about Ted, but I just heard it was Ted he talked about. b. I knew Donald talked about Ted, but I just heard he only talked about Ted.
Such examples have led most scholars to abandon the view that it-clefts (or Hungarian preverbal focus) encode exhaustivity as their at-issue content.
.. Conversational implicature The second main type of approach goes back to Horn (), who argued that the exhaustivity of it-clefts is a generalized conversational implicature. One of the powerful arguments in favour of such an account is that study after study has revealed naturally occurring examples in which exhaustivity does not apply. Consider, for instance the wellknown data in () from a poem by James Oppenheim cited by Horn (). Another famous set of examples has been provided as early as Prince (). ()
As we go marching, marching, unnumbered women dead go crying through our singing their ancient cry for bread. Small art and love and beauty their drudging spirits knew. Yes, it is bread we fight for—but we fight for roses, too!
Horn derives the exhaustivity implicature by suggesting the principle in (). ()
Principle of derivation: If an expression α entails P(a) and presupposes ∃x.P(x) than α has a generalized conversational implicature that :∃y:y ∈ C ∧ PðyÞ ∧ y ⊈ a, where C is some contextually relevant set of individuals.
Scholars have often noted that () seems somewhat ad hoc, especially because it is unclear how it relates to the Gricean maxims (Grice, ). However, I will return to this principle in section . showing that in light of more recent findings it can be translated into a more appealing form. Another reason why many scholars have rejected this type of analysis is that the exhaustivity inference of clefts is not as easy to cancel as one would expect implicatures to be. Indeed, the oddity of examples such as () is hard to overlook. As opposed to this,
the exhaustivity implicature associated with canonical sentences with a narrow focus seems much easier to cancel, as witnessed in (). ()
a. #It was Donald who lied; indeed it was Donald and Ted. b. #It was Donald who lied; indeed Ted lied too.
()
a. DONALD lied; (indeed) Donald and TED lied. b. DONALD lied; (indeed) TED lied too.
While the particular derivation proposed by Horn is widely rejected in the literature, the idea that constructions such as it-clefts conversationally implicate exhaustivity is not unpopular in recent literature. Some examples include Onea & Beaver () for Hungarian pre-verbal focus, Saur (), DeVeaugh-Geiss et al. () for German, Washburn et al. () for English, and Destruel () for French, and others.
.. Presupposed exhaustivity The third main type of approach is that exhaustivity in clefts is some sort of presupposition or not-at-issue content. To my knowledge, this kind of approach was first suggested by Halvorsen (), who specifically referred to the exhaustivity inference as a conventional implicature. More recent similar theories such as Büring & Križ () are explicit that the exhaustivity inference should be a presupposition, whereas Velleman et al. () refer to it more generally as not-at-issue semantic content. The main question of such theories concerns the exact nature of the exhaustivity presupposition. Originally, Halvorsen () proposes a fairly straightforward uniqueness presupposition as in (), which essentially reappears in Hedberg (, ) and Percus (), with a slightly different motivation and formal representation. ()
It is A who P. presupposes: ∃x:8y:PðyÞ ! x ¼ y
However, such an approach runs into a series of problems. For instance, Atlas & Levinson () suggest that, under the very assumptions Halvorsen makes, () is predicted to be infelicitous because a presupposition should survive embedding under negation and contradict the second statement. Moreover, Büring & Križ () argue that even a more standard uniqueness presupposition would be violated by the very content of a sentence like (). After all, if the sentence presupposes that exactly one person lied, the assertion that two persons lied would come out as a contradiction to the presupposition. Yet both () and () are perfectly acceptable. ()
It wasn’t Donald who lied, it was Ted and Jeb.
()
It was Donald and Hillary who lied at the debate.
To avoid such problems, more recent analyses that assume that exhaustivity is presupposed have come up with more complicated proposals. In particular, both Büring & Križ ()
-
and Velleman et al. () suggest that the existential presupposition essentially amounts to a conditional statement. The predicted presupposition is in (): ()
It is A who P. presupposes: PðAÞ ! :∃x:PðxÞ ∧ x ⊈ A
While Velleman et al. () and Büring & Križ () share this conditional presupposition of exhaustivity in a certain sense, the approaches are fundamentally different when it comes to the derivation of this presupposition. I will discuss these approaches in a bit more detail. Büring & Križ () assume that clefts have an underlying copula structure involving two terms: the cleft pivot and the definite description gained from the cleft relative. Hence, one could paraphrase (a) as (b). Büring & Križ () and Križ () observe that for plural definite terms there is a homogeneity requirement for truth evaluation. Hence, in a situation in which the group of liars overlaps with the group of politicians it is hard to evaluate whether the sentence in (b) is true or false. Büring & Križ () suggest that definites in copula sentences presuppose that such overlaps in the extension of the two terms do not apply. For clefts, this amounts to the presupposition that we should not consider scenarios in which the politicians lied but other persons lied as well. ()
a. It was the politicians who lied. b. The ones who lied are the politicians.
As opposed to this, Velleman et al. () do not assume an underlying equative or definite description for clefts. Instead, they assume that it-clefts associate with a Question Under Discussion (QUD) in the sense of Roberts (a) and Beaver & Clark (), which matches the focus semantic value of the cleft in the sense of Rooth (): that is to say, itclefts associate with a wh-question corresponding to their focus structure, as in (). With respect to this question, clefts semantically encode as part of their at-issue meaning that the canonical inference is a true answer, whereas they encode as a presupposition that the canonical inference is the strongest true answer to that question. () a. QUD: Who lied? b. ANS: It was DONALD who lied. at-issue: There is at least one true alternative in the denotation of the question which is at least as strong as Donald lied. not-at-issue: There is no true alternative in the denotation of the question which is stronger than Donald lied. The fact that Velleman et al. () make the exhaustivity inference depend on the alternatives present in the QUD gives them a significant empirical advantage as compared to Büring & Križ (). In particular, in cases in which the focus is only on a part of the cleft pivot, as in (), the exhaustivity inference predicted by Büring & Križ (), given in (a), does not appear to be empirically correct. Since, however, the QUD compatible with
() only involves alternatives about John’s daughters, Velleman et al. () correctly predicts the exhaustivity inference in (b). ()
It was John’s ELDEST daughter who attended. a. The only person who attended was John’s eldest daughter. b. The only daughter of John’s that attended was his eldest daughter.
At the same time, a potential weakness of Velleman et al. () is closely connected to its strength. For now, we have no sufficient understanding of the range of possible QUDs and what kind of configurations of alternatives are allowed or disallowed according to general principles of discourse. Consider the example in (). If the QUD includes at least the alternatives in (a) and (c), () is predicted to be contradictory. However, if the QUD for whatever reason does not include the alternative (c) or both (b) and (c), the discourse is predicted to be felicitous. One would then expect that as an answer to a question like (), the answer could be maximal without contradicting (b) because (c) is (at least intuitively) not part of the question alternatives. Yet the preferred interpretation of the answer (b) seems to still contradict (c). To some extent, this seems to apply to (c) as well. (See Bade, , for a possible independent explanation, but one which does not apply to (d).) ()
It was Donald who lied, and Hillary lied too. a. Donald lied. b. Hillary lied. c. Donald and Hillary lied.
()
a. b. c. d.
We all know that Hillary lied. But who else lied? It was Donald who lied. Donald lied. Donald.
Apart from such issues, there is a further problem all presuppositional accounts face that was already quite extensively discussed in Horn (). Presuppositions and conventional implicatures, like all not-at-issue content, are expected to project from entailmentcancelling environments. But Horn argued that there is no evidence that the exhaustiveness inference of it-clefts projects, as neither utterance in () commits the speaker to the inference that Mary ate nothing other than a pizza. Intuitions are not very clear, especially on the conditional presupposition. Yet preliminary evidence from Beaver & Tonhauser () suggests that cleft exhaustiveness may indeed not project. ()
a. Was it a pizza that Mary ate? b. It wasn’t a pizza that Mary ate.
.. Indirect exhaustivity There is a final set of approaches which has occasionally been grouped with presuppositional accounts because they assume a strong similarity between definite descriptions and
-
it-clefts. In this group of accounts, there is indeed a presupposition associated with clefts; however, it is not exhaustivity itself which is presupposed. In approaches such as Szabolcsi’s () and Pollard & Yasavul’s (), clefts, similarly to definite descriptions, presuppose some discourse referent and then they contribute an identification of that discourse referent as their at-issue content. So, far these approaches seem to be on a par with Percus (). However, the essential difference is that these approaches do not assume that clefts presuppose uniqueness. Szabolcsi () argues that clefts presuppose a maximal discourse referent that satisfies the predicate of the cleft relative. So, (b), repeated as (), would simply presuppose that there is a maximal (non-empty) set of liars. Büring () correctly diagnoses that this presupposition trivially follows from the existence presupposition of clefts, for if there is someone who lied, than there must also be a maximal set of liars, and if there is a maximal non-empty set of liars, then someone must have been lying. Crucially, however, Szabolcsi () argues that it is precisely that maximal set of liars which is identified by the cleft pivot. Then, the exhaustivity of the clefts immediately follows, as shown in (). ()
It was DONALD who lied at the debate. a. Presupposition: x : LðxÞ ∧ 8y:LðyÞ ! y ⊆ x b. At-issue content: x = Donald
This approach has one considerable advantage: it does not predict projection of exhaustivity, since the only presupposition is in fact the existence presupposition (after all, the maximality presupposition is truth-conditionally equivalent to the existence presupposition). The main disadvantage is, however, that it makes exhaustivity essentially part of the at-issue content, since it predicts interaction with entailment-cancelling operators. In particular, examples such as () and () are clearly predicted to be acceptable. Finally, Pollard & Yasavul () further develop this indirect kind of approach by essentially leaving it up to discourse rules to decide whether the presupposed discourse referent is maximal or not. In particular, they assume that clefts presuppose some discourse referent (essentially meaning that they have an existential presupposition) and this discourse referent may end up being maximal or not. If the discourse referent is maximal, exhaustivity is predicted; if it is not maximal, exhaustivity is not predicted. Pollard & Yasavul () suggest that as a follow-up on overt questions, it-clefts tend to be exhaustive because overt wh-questions introduce a maximal discourse referent, whereas following up on indefinite antecedents, it-clefts tend to be non-exhaustive. () a. Who lied? It was Donald who lied. So, nobody else did. b. Someone lied. It was Donald. Maybe others did too. Apart from some technical stipulations worthy of further investigation, Pollard & Yasavul () share the fate of Szabolcsi () and the original at-issue exhaustivity approaches such as Bolinger (), Atlas & Levinson (), and Szabolcsi (), in that again exhaustivity ends up being at-issue, or at least not conventionally not-at-issue. However, there is an interesting new feature in the approach of Pollard & Yasavul (): exhaustivity may well be at-issue whenever it exists, but it may just as well fail to exist. It is not clear to
me whether the predictions are borne out empirically. It certainly does not seem to me that an example such as () is better than () in isolation. Similarly, it is not obvious how on either of these approaches Horn’s counterexamples like () can be accounted for. After all, there should be a reading for (a) in which exhaustivity ends up as at-issue content and then it is not clear why the example does not seem acceptable. ()
A: Who lied? B: It wasn’t Hillary who lied. It was Donald and Hillary.
. E
.................................................................................................................................. In recent years, a significant number of experimental studies on clefts have been presented. The experimental studies tend to have in common that they compare it-clefts with plain narrow focus and exclusives (and more recently, especially after the publication of Büring & Križ, , even with definite descriptions). In general, the findings are essentially the same in all languages, though the theoretical implications depend somewhat on further properties of the experimental studies. The main finding is that in some sense it-clefts are less exhaustive than exclusives such as only but significantly more exhaustive than plain narrow focus. One of the first studies about the exhaustivity of pre-verbal focus in Hungarian was reported in Onea & Beaver (). Later versions of the experiment have been replicated for French, German and English: see Destruel () and Destruel et al. (); see also Cummins et al. () and Amaral & Cummins () for related experiments. The dependent variable of the experiment was the choice of the reply of participants to utterances containing clefts (a), exclusives as in (c) or plain narrow foci as in (b) when the exhaustivity inference was violated. The three available choices were as shown in (). ()
Alternative Target sentences: a. cleft: It was Donald who lied at the debate. b. narrow focus: DONALD lied at the debate. c. exclusive: Only Donald lied at the debate.
()
Reply choices: a. Yes, and Hillary lied at the debate too. b. Yes, but Hillary lied at the debate too. c. No, Hillary lied at the debate too.
The results clearly show for all investigated languages that for plain narrow foci, such as (b), as target sentences, most participants tend to choose yes, and continuations, whereas for exclusives, such as (c), most participants tend to choose no-continuations. For it-clefts, as in (a), in the various investigated languages, and for Hungarian pre-verbal focus, the results generally show a lower rate of yes, and answers as compared to narrow foci, and a lower rate of no answers as compared to exclusives. Originally, Onea & Beaver
-
() took this to suggest that the exhaustivity inference triggered by pre-verbal focus in Hungarian, and by analogy it-clefts, was a conversational implicature, which they took to obtain from the tendency to interpret answers to wh-questions as complete and, hence, exhaustive. Following observations made in Mayol & Castroviejo () and Xue & Onea (), in a more recent study, Destruel et al. () eventually concluded that the experiment does not really show that the exhaustivity of clefts is pragmatic (or an implicature in particular), but rather that the exhaustivity inference is not-at-issue regardless of its semantic or pragmatic status. While these findings are compatible with analyses such as Velleman et al. () and Büring & Križ (), they do not elucidate the greatest weakness of these analyses, namely the lack of evidence for the predicted projection of cleft exhaustivity (in any way). Another study that confirms that there is a clear difference between the processing of exhaustivity violations with only and it-clefts is Drenhaus et al. (). In particular, Drenhaus et al. () confronted participants in an ERP study with sentences that violate the exhaustivity inference associated with it-clefts, as in (), or with exclusives, as in (). They observed that, when faced with overt exhaustivity violations, it-clefts in German give rise to N effects while exclusives give rise to P effects. From this, the authors conclude that the exhaustivity inference in it-clefts must be pragmatic (as opposed to the semantic effect in exclusives). Crucially, however, the alternative explanation that exclusives provide at-issue, while it-clefts provide non-at-issue exhaustivity, offered in Velleman et al. () or Destruel et al. (), would also account for the observed results. ()
Es ist Maria, die das Klavier spielen kann und außerdem noch Luise und Jana . . . It is Mary who the piano play can and also too Luise and Jana ‘It is Mary who plays the piano, and so do—in addition—Luise and Jana.’
()
Nur Maria kann Klavier spielen und außerdem noch Luise und Jana . . . only Mary can piano play and also too Luise and Jana ‘Only Mary plays the piano, and so do—in addition—Luise and Jana.’
Destruel et al. () claimed that the reason why the exhaustivity inference of it-clefts leads to lower levels of contradiction as compared to exclusives is the not-at-issue status of the exhaustivity inference in it-clefts. This claim came under scrutiny by DeVeaugh-Geiss et al. (). DeVeaugh-Geiss et al. () suggest that if at-issueness explains the difference between exclusives and it-clefts in the design of Onea & Beaver (), one would expect to find that difference in other cases in which two inferences differ only in their at-issue status. Hence, one would expect that contradicting the not-at-issue inference for exclusives would yield similar results to contradicting the not-at-issue inference of it-clefts, and that both should differ from contradicting the respective at-issue inferences. Hence, one would expect that the two sentences in () should be acceptable to a comparable level, and certainly more acceptable than the ones in (). ()
Contradicting at-issue inferences: a. It was Donald who lied and he did not lie. b. Only Donald lied and Ted lied.
()
Contradicting not-at-issue inferences: a. It was Donald who lied and Ted lied. b. Only Donald lied and he did not lie.
Contrary to this prediction, in a rating study on it-clefts in German, DeVeaugh-Geiss et al. () found that out of all these sentences only (a) is somewhat acceptable. Hence they conclude that it is unlikely that at-issueness alone can explain the difference observed between it-clefts and exclusives in the experiments previously mentioned. They suggest that, in addition to being potentially not-at-issue, the exhaustivity inference of it-clefts is likely to be a pragmatic inference, in particular a quantity implicature. Once again, however, it is not entirely clear what follows from the results. In particular, a potential experimental confound comes from the fact that when contradicting the not-atissue content of exclusives, the overtly mentioned alternative in the QUD has been contradicted. By contrast, when contradicting the exhaustivity inference of clefts, the overtly mentioned alternative was not contradicted. Such uncontrolled factors in the experiment need further investigation in future experiments. Another recent study reaching a very similar conclusion is Washburn et al. (). They observe that the violation of cleft exhaustiveness does not lead to a very strong decrease in acceptability. Instead, it-clefts are much less acceptable when they lack an additional contrastiveness property they have been argued to have by several scholars, most recently by Zimmermann (), Destruel & Velleman (), and Destruel et al. (). In particular, Destruel & Velleman () argue that it-clefts are more felicitous when they conflict with the expressed expectations of interlocutors in discourse, as for example in (), which is a corrective use of it-clefts. ()
A: I wonder why Alex cooked so much beans. B: Actually, it was John who cooked the beans.
In a recent study DeVeaugh-Geiss et al. () compared the way in which definite descriptions, exclusives, narrow focus and it-clefts behave in an incremental picture validation or falsification task experiment. In this experiment participants were required to judge as soon as possible whether a sentence giving rise to an exhaustivity inference is true or false, while boxes were uncovered, incrementally increasing the available information. The critical observation is that, in a scenario in which the participant was confronted with one of the targets from (a) and had the information (b) but had no information whatsoever regarding the question (c), most participants judged the sentence true if the target was a narrow focus sentence and did not make a judgement when the sentence contained an exclusive. Similarly, in the falsification task, if the person knew (a) but had no information regarding the question (b), participants judged sentences with exclusives as false whereas they did not make a judgement for targets with narrow focus. () a. Target: cleft: It was Donald who lied. exclusive: Only Donald lied. narrow focus: DONALD lied. definite: The person who lied was DONALD.
-
b. Information: Donald lied. c. Question: Did anyone else lie? ()
a. Information: Hillary lied. b. Question: Did Donald lie?
Both observations above are expected according to all theoretical approaches. Crucially, for both it-clefts and definite descriptions, participants appeared to behave in two clearly distinguishable patterns, which led to a post hoc division into an exhaustive and a nonexhaustive group. The exhaustive group treated clefts like exclusives; the non-exhaustive group treated clefts like narrow focus. In one sense, this seems to be perfectly in line with definite-based theories of cleft exhaustivity such as Büring & Križ (), except that Büring & Križ () did not actually predict the existence of a group of participants who find neither definites nor it-clefts exhaustive. DeVeaugh-Geiss et al. () then suggest that their findings are rather compatible with a theory such as Pollard & Yasavul (), in which the exhaustivity in it-clefts is not a semantic inference, but still should be derived in a similar way as the uniqueness inference of definite descriptions. I will return to this issue in the next section. DeVeaugh-Geiss et al. () further suggest that even under such a mixed approach, a number of implicatures may apply which tend to bias towards the exhaustivity inference in it-clefts. Finally, I wish to mention a series of experiments that have recently been conducted on shi . . . de clefts in Mandarin and presented in Liu & Yang (). Participants needed to rate target sentences in which the exhaustivity was violated for clefts, plain focus, and exclusives. Interestingly, even though the contexts in which the stimuli were shown, as well as the range of overt alternatives, have been significantly varied in the experiments, the results show that clefts remain in between canonical sentences and exclusives on the rating scale in each of the cases. This seems to suggest that, at least in Mandarin, the derivation of cleft exhaustivity either does not significantly depend on the linguistic context (contrary to the suggestions of Pollard & Yasavul, ; and DeVeaugh-Geiss et al., ) or that it depends on factors not yet considered in the existing theories. I would like to conclude this section by noting that experimental studies and theoretical studies seem to diverge quite significantly. While most of the experimental studies have found that the exhaustivity inference of it-clefts is not particularly strong and is unlikely to be semantically coded, most of the theoretical literature—the most notable exception being Horn (), Horn (), and Pollard & Yasavul ()—tend to derive exhaustivity in itclefts semantically. While it seems fair to say that the experimental studies only provide partial evidence for their claims, the mere mass of studies and their converging verdict seems to be something that needs to be taken quite seriously in the theoretical literature.
. D
.................................................................................................................................. In this section, I discuss a potential way to bridge between the existing experimental evidence and the theoretical literature. As shown in the last section, the experimental literature tends to favour a pragmatic derivation of exhaustivity for it-clefts. At the same
time, it should be noted in all fairness that a clear and principled derivation of the exhaustivity of it-clefts as an implicature has never been given. I will try to therefore reevaluate the principle of derivation in Horn () given in () in light of recent developments and suggest at least a potential direction for future research. As a starting point, it should be made clear that any analysis of exhaustivity of it-clefts should somehow make sure that the exhaustivity inference depends on the focus in the itcleft. This point has been elucidated by Velleman et al. (). But what has been discussed less in the recent literature (even though mentioned in Velleman et al. ()) is that the very same is true for the existential presupposition of it-clefts. This is shown in (). In (), the classical existential presupposition is not satisfied; after all, the very QUD indicates that it is not clear that anyone attacked Megyn. It is a possibility entertained by the speaker that actually Megyn was the agent rather than the patient of the attack. Yet, if the existential presupposition depends on the focus of the cleft rather than the pivotrelative-clause division, the data is entirely predictable. ()
a. Who attacked whom? b. It was DONALD who attacked MEGYN, and not the other way around.
Moreover, in the introductory section, I have suggested that when the focus is not or at least not only on the cleft pivot, it is often unclear whether an existential presupposition comes about. In fact, it is also quite unclear whether topic-comment clefts exhibit a similar level of exhaustivity to focus-background clefts. Consider the random examples in (), for which defending both exhaustivity and existential presupposition appears to be hard. In fact, one might even conjecture that the reason why exhaustivity of such clefts appears less clear or non-existent is because it is not clear at all what is presupposed by such examples. ()
a. James Sharman. He was with Lord Nelson at Trafalgar. It was he who helped to carry the admiral from the upper deck to the cockpit. (William Kingston, A Yacht Voyage Round England, p. ) b. Some individuals do stand out, chief among them, in terms of breadth of engagement, was undoubtedly Lord Kames. It was Kames who contributed to the important volume of Essays published by the Philosophical Society in [ . . . ] (John Finlay, The Community of the College of Justice: Edinburgh and the Court of Session, ‒, p. )
So, if Horn’s principle of derivation of exhaustivity of it-clefts in () (that was based on the existential presupposition of clefts) is to somehow cover such data as in () and (), it has to be modified. One of the easiest ways to do so is, following Velleman et al. (), to suggest that clefts presuppose an issue. This issue, under existential closure, will be truth-conditionally equivalent to the existential presupposition of clefts. Hence, (a) essentially conveys the presupposition that someone attacked someone. In view of the treatment of presuppositions as anaphora in dynamic frameworks ever since Van der Sandt (), and especially in Pollard & Yasavul (), we can say that a cleft anaphorically refers to an issue. Hence, we may want to understand Horn’s idea that clefts have an existential presupposition as the claim that clefts anaphorically refer to an
-
issue. With this in mind, we can even translate Horn’s claim that clefts implicate exhaustivity into the claim that clefts imply that no other true answer to that issue is more informative than the canonical inference. Then, Horn’s principle of derivation can be understood as ().1 ()
If an expression E refers to an issue Q while answering it, the addressee should infer that Q is answered in a maximal true way by E.
Such a principle seems to allow quite a large range of possible theories. Firstly, the principle could be a pragmatic principle derived in a Gricean way, making the exhaustivity inference a classical generalized conversational implicature. Deriving the principle would mainly involve the observation that by using a plain focus construction, in many languages, the QUD would be sufficiently identifiable.2 By adding an additional marker that the question is being answered, the speaker would invite the hearer to draw an additional inference that would not be licit if some other question was addressed. The maximality with respect to the QUD would appear to be a good candidate for such an inference. Second, if one follows Pollard & Yasavul (), the principle could follow from the very meaning of clefts, which would identify the maximal answer of the issue which is being addressed to the cleft pivot. But at the same time, in the theory of Pollard & Yasavul () it-clefts are not limited to addressing a QUD. Their anaphoric antecedent could be some other discourse entity. In this case, it-clefts would not be exhaustive, but the principle would still vacuously hold. Yet another option is that, as suggested by Beaver & Onea (), it-clefts always refer to an issue in discourse, but maximality of the answer only applies in cases in which the question itself requests a maximal answer. Hence, for a mention-some question, for instance, one would not expect an it-cleft answer to be exhaustive. Such a theory would in essence restate Velleman et al. () with the main difference that question alternatives need not form a lattice structure. This does not entirely exhaust the range of possibilities. Crucially, however, all of these possibilities seem to be closely related and, more importantly, they all allow both for exhaustive and non-exhaustive interpretations of clefts. In all of these versions, the observation made by experimental research that it-clefts are somehow less exhaustive than exclusives and more exhaustive than plain narrow focus canonical sentences seems to be quite expected. It would seem that from such a perspective the remaining puzzle is the relation between definite descriptions and it-clefts. In the most recent literature the connection between it-clefts and definite descriptions has mainly been invoked to account for the exhaustivity of clefts by making use of the maximality or uniqueness presupposition of definite
1
I do not know whether Horn would accept this paraphrase of his principle of derivation but as far as I can see these amendments are unavoidable. 2 This is the core idea of Roberts (), who analyses focus as an indicator of the QUD. The idea has been further developed by Beaver & Clark () and many others; see also Tonhauser (Chapter in this volume).
descriptions, as standardly assumed following Russell () or Strawson (). In less recent literature, mainly in Delin (, ), however, the anaphoric nature of the existential presupposition of it-clefts is emphasized (for focus, see Geurts & Van der Sandt, ). In a whole line of research since Heim (b) and Van der Sandt (), definite descriptions have been considered anaphoric. In such dynamic theories, the uniqueness inference seems to be some default inference drawn by the hearer in an attempt to make definite descriptions identifiable in discourse whenever a proper antecedent is missing: see Abbott () for a theoretical overview and Schwarz (Chapter in this volume) for an overview of experimental research. From this perspective, the anaphoricity of it-clefts would not be surprising at all if they have an underlying definite description structure (or even independently of that; see DeVeaugh-Geiss et al. ). In fact, the parallel between the two structures would be quite complete if plain definite descriptions could also be anaphoric to issues. To a certain extent a large number of data discussed in Križ () that show cleft-like focus effects with definite descriptions would be accounted for if that were the case. Thus, it seems that in view of the experimental evidence an anaphoric theory of clefts is necessary. Moreover, there seems to be no other option than assuming that (at least in some cases) clefts can be anaphoric to some very specific QUD, as suggested by Velleman et al. (); Pollard & Yasavul (); and more recently by DeVeaugh-Geiss et al. (). How exactly the exhaustivity should be derived is not sufficiently clarified in the literature, but it seems that if one intends to account for the variation observed in the experimental literature one of the theories along the lines suggested earlier could be expected to be on the right track.
. C
.................................................................................................................................. In this chapter, I have provided an overview of the main theoretical approaches and the most important experimental studies on the exhaustivity of it-clefts in English and other languages. A significant discrepancy between the theoretical and the empirical literature has become evident to the extent that while most theoretical literature views exhaustivity of it-clefts as a semantic feature, most empirical literature finds the inference to be too weak for a semantic one. This discrepancy can partly be explained by a number of issues with the experimental studies, yet in its entirety the experimental evidence does seem to have a fairly clear impact on the direction in which future theoretical analyses will need to go. The observation that it-clefts are anaphoric in nature suggested by Pollard & Yasavul () seems to require further control of contextual variables when designing experiments. In particular, manipulating potential anaphoric antecedents appears to be a good way of getting at contrasts in exhaustivity. Crucially, any theory of the exhaustivity of it-clefts will need to take into account the effect of information structure discussed in Velleman et al. () and it should somehow account for the parallel between it-clefts and definite descriptions in terms of exhaustivity theoretically predicted by Büring & Križ () and observed in DeVeaugh-Geiss et al. ().
-
A This research profited from many fruitful discussions with David Beaver and Malte Zimmermann on various occasions, which I gratefully acknowledge. Special thanks also to Chris Cummins for very helpful editorial comments on this chapter. This research has been funded by the German Science Foundation (DFG) as part of the priority programme SPP ‘XPrag.de: New Pragmatic Theories Based on Experimental Evidence’ (grant ‘Exhaustiveness in It-Clefts’).
......................................................................................................................
......................................................................................................................
.
A news article in The Guardian from July attributes the statement in () about new Prime Minister Theresa May to a ‘liberal Tory’ (Behr, ). ()
‘Theresa will at least stab you in the front.’
The article goes on to clarify that the statement was made as a compliment. Indeed, any speaker of English could only interpret () by taking its focus structure into account: the phrase in the front is contrastively focused with an implicit alternative, in the back, yielding an overall interpretation that could be paraphrased as ‘Given circumstances where stabbing is inevitable, Theresa will take the alternative of stabbing you in the front, rather than stabbing you in the back.’ Interpreting the statement as a compliment therefore crucially relies on a comprehender’s ability to recognize the focus structure of the sentence, and infer the (in this case very small) contrast set with respect to which the focused element must be interpreted. Contemporary research on focus within linguistic theory falls primarily into two strands: the study of the semantics of focus—its contribution to truth conditional and non-truth conditional meaning (Chafe, : Rooth, , ; Krifka, ; Schwarzschild, ), and the study of the intonational realization of focus in speech (Ladd, ; Gussenhoven, ; Selkirk, ; Pierrehumbert & Hirschberg, ; Selkirk, ). Experimental investigations into focus have been going on alongside theoretical work throughout. Along the way, they have occasionally become entangled in important debates within psycholinguistics, such as the debate over staged versus constraint-based parsing models (Ni et al., : Sedivy, ; Paterson et al., ). Experimental questions have also often come from theoretical frameworks: for instance, experiments on focus projection and on the nature of focus alternatives were inspired by theoretical claims. This chapter surveys the main lines of experimental research into focus. It is organized thematically rather than chronologically or by methodology, though sometimes certain questions attract certain experimental paradigms. It is therefore important to keep differences in methodology in mind, particularly when studies asking similar questions use different tasks, and report different outcomes. Also important are differences in how focus is ‘operationalized’ in experimental studies. While manipulating the preceding context,
using focusing syntactic constructions such as clefts, using focus particles like only, and manipulating the type and placement of pitch accents are all meant to manipulate focus, these different ways of manipulating focus come with their own linguistic properties, and quite often do not yield identical results. Experimental studies also differ importantly in how they characterize the abstract notion of focus that these manipulations are meant to probe. Some studies define focus in terms of discourse status (new versus given information, focus versus background). Others strictly tie focus to its surface realization, whether that is the presence of a pitch accent, the use of a focusing syntactic construction, or correspondence to explicit wh-elements in the discourse. Taking such an approach, a problem then becomes how to discover the focus structure of a sentence by way of overt cues to focus and principles about how focus can project from overtly-marked elements to unmarked ones. For many studies, ‘focus’ is used to mean contrastive focus—information that is contrastive with another (explicit or implicit) discourse element, as opposed to informational focus—focus used to mark new information. In such cases, contrast is often defined in terms of intended meaning, and inferring the intended contrast set is a matter affecting a sentence’s truth-conditions. As we will see, even how the notion ‘contrastive’ is implemented can vary from study to study. Finally, ‘focus’ is sometimes used in experimental studies to mean ‘discourse focus’, which is related to attentional focus rather than the notion of focus defined in terms of linguistic representations. While we might want to set research into discourse focus aside because it is defined differently from linguistically-characterized focus, it is not always clear-cut which cases should be set aside: notions like discourse focus have been invoked in explanations of some of the same linguistic phenomena that linguistic focus is used to explain. For example, one influential theory of discourse structure (Grosz & Sidner, ) proposes that two different levels of focus structure are responsible for different interpretive phenomena: Centering Theory works at the sentential/inter-sentential level and is responsible for resolving anaphoric reference, while discourse-level intentional structure is responsible for the interpretation of definite descriptions. Such an explanation would seem to have much in common with e.g. the proposal in Cummins & Rohde (), which does appeal to contrastive focus in connection with pronoun resolution. The organization of this chapter cross-cuts the earlier issues, and instead focus on lines of research that are held together by the questions they ask about memory representations, time-course, interpretation, and cues to focus structure. The final section concludes with discussion of some issues that are likely to be important in future experimental research on focus.
. A, ,
.................................................................................................................................. Experimental studies going back to the s have used recall paradigms to show that focused material is remembered better than unfocused material, whether focus is manipulated using pitch accent (Cutler, ; Cutler & Foss, ; Hornby, ), using the discourse context (Cutler & Fodor, ), or using syntactic focusing constructions such
.
as clefts (Bredart & Modolo, ; Morris & Folk, ; Birch et al., ). Enhanced memory representations in such studies were construed as evidence that more attention was dedicated to processing focused material. Increased attention at the point of memory encoding was also consistent with focused material being easier to integrate into an existing discourse model (Morris & Folk, ), and the increased likelihood of referring to focused elements in story continuations (Birch et al., ) due to increased availability of these items in memory. A line of research attempts to use effects of focusing on attention to explain enhanced sensitivity in change-detection tasks. Bredart & Modolo () found that the ‘Moses illusion’ described by Erickson & Mattson ()—in which many participants provide an answer for the question ‘How many animals of each kind did Moses take on the Ark?’ and fail to notice that ‘Moses’ should have been ‘Noah’—was sensitive to clefting: prticipants in a sentence verification task were more susceptible to the illusion when material other than the problematic name was the focus of a cleft (e.g. ‘It was two animals of each kind that Moses took on the Ark.’). Sturt et al. () (also Sanford & Sturt, ) suggest such change-blindness phenomena have to do with the influence of focus on depth of processing: focused material is more likely to be processed fully, rather than being left underspecified. In a text-change detection task, participants had to detect changes (e.g. ‘cider’ being replaced by ‘beer’; changes underlined in ()–()) between two repetitions of a sentence. Changes were more likely to be noticed when the changed word was focused by clefting or pseudoclefting () than when a different word was focused (). ()
What Jamie really liked was the CIDER/BEER.
()
It was JAMIE who really liked the cider/beer.
Similar results obtained when focus was manipulated using the prior context instead of syntactically (Sturt et al., ) (), and using both prosody and prior context (Sanford et al., ) (). ()
a. Everybody was wondering which man got into trouble./Everybody was wondering what was going on that night. b. In fact, the man with the hat/cap was arrested.
()
a. They wanted to know which money had been stolen. /They wanted to find out what had happened. b. The money from the WALLET/PURSE had gone missing. c. Thefts in the area were becoming all too common.
Eye-tracking during the same change-detection task showed more fixations and longer viewing times on a changed than on an unchanged word, only when the changed word was in focus (Ward & Sturt, ). Explanations in terms of depth of processing are in principle compatible with either focus leading to more detailed processing of lexical material, or focused material requiring more processing effort. According to the first type of explanation, unfocused material has a less detailed (perhaps underspecified) memory representation to begin with, leading to differences in retrievability between focused and unfocused material. By contrast,
according to an effort-based explanation, the same information may be encoded better for focused compared to unfocused material. In addition, according to Birch et al. (; see also Birch & Garnsey, ), the effects of focusing on memory only appear after a delay: in probe recognition experiments, when an element in the subject position of a sentence was probed immediately, syntactic clefting had no effect on recognition latency. However, when the probe was delayed, syntactically focused words were recognized faster than those that had not been focused. Similarly, Foraker & McElree () argued on the basis of SpeedAccuracy Tradeoff experiments that clefting did not affect the speed of accessing an antecedent representation, but did increase the likelihood of retrieving the antecedent representation. These results were taken to suggest that focusing using clefts strengthens the focused material’s memory representation, but does not result in focused elements being actively maintained in focal attention (see Onea, Chapter in this volume, for more on clefts). Turning to effects on the processing of the focused element itself, there have been mixed findings regarding the impact of focus on reading time. Birch & Rayner () used clefting to manipulate focus/prominence as in ()‒(). ()
The tenants at the complex were sick and tired of all the noise coming from #.
()
a. It was the landlady who confronted the woman who lived there. b. The landlady confronted the woman who lived there.
They reported fewer fixations and shorter reading times for focused words, in both early measures of reading (first-fixation time, gaze duration) and late measures (total reading time, total number of fixations), suggesting that both encoding and integration processes were facilitated for syntactically focused words. In another reading study using clefts and pseudoclefts to manipulate focus, however, Lowder & Gordon () found increased reading times on a word (memo) when it was syntactically focused (a) compared to conditions without clefting or pseudoclefting (b) or conditions where a different word was focused (c), as well as more time regressing to words immediately preceding the focused word. ()
a. What the secretary typed was the official memo about the new office policy. b. Yesterday the secretary typed the official memo about the new office policy. c. It was the secretary that typed the official memo about the new office policy.
The authors suggested that these findings reflected deeper encoding of focused material, and deeper integration of focused material with its sentential context. The difference in findings between Birch & Rayner () and Lowder & Gordon () might be related to the difficulty of integrating new information into the discourse representation: while the focused element was discourse-new in both studies, only Birch & Rayner () provided a preceding context sentence (). Processing the focused element might have been facilitated in that study because the focus (the landlady) was implicitly associated with the individuals mentioned in the context sentence. In fact, a study by Chen et al. () independently manipulated newness and focus, and suggested that being new in the discourse makes information more difficult to integrate,
.
while focused information is processed more quickly. In a reading study in Chinese, focus was marked by the focus-particle shi (). Newness was manipulated independently using the discourse context (). ()
()
a. Heren was persuading his friends to go on an outing. (He) ignored that the weather forecast had predicted bad weather. b. Heren was persuading Zhongying and others to go on an outing. (He) ignored that the weather forecast had predicted a bad weather. a. At that time shi Zhongying reasonably opposed him. ‘At that time it was Zhongying who opposed him reasonably.’ b. At that time Zhongying reasonably opposed him. ‘At that time Zhongying opposed him reasonably.’
In the critical region (Zhongying), focused information took less time to read than nonfocused information, while new information took longer to read than given information. In the immediately post-critical region, fixation durations were longer for new information than given information, only in the focus conditions. Thus, it may be important to take into account the extent to which focused elements are implicitly or explicitly represented in the discourse context when asking about the effects of focusing on processing time or complexity. When integrating new material requires revising the discourse model, not simply adding to it, focused material is processed more slowly than unfocused material. In Benatar & Clifton (), a context question (b) was followed by an answer in which a critical word was either new or contrastively (correctively) focused (c). ()
a. John and Mary are working today. b. Speaker A: Did you tell someone to go home early?/Did you tell Mary to go home early? c. Speaker B: I told JOHN, but I don’t know if it was a good idea.
Target words were read more slowly when the critical word corrected prior information than when it provided new information. By contrast, target words introducing new information were read more slowly than those providing given information, in sentence pairs like (). ()
a. Speaker A: I’m confused, does Kyle care about Natalie?/Natalie is confused, does Kyle care about someone? b. Speaker B: Kyle cares about Natalie but he doesn’t show it.
The presence of explicit contrast sets in the discourse seems to facilitate integration of focused elements. Braun & Tagliapietra () and Fraundorf et al. () showed that focus alternatives explicitly introduced in the prior discourse are better remembered when the focused element receives a contrastive (L+H*) pitch accent. In a direct comparison of explicitly introduced and inferred contrast, Chen & Yang () had participants read short discourses that introduced one or two characters, with the character being emphasized or
non-emphasized in subsequent texts. Eye movements showed that early processing of the emphasized character was facilitated (reflected in first fixation duration and gaze duration), which may have been due to increased attention allocation, whereas late integration of the emphasized character was inhibited when the discourse did not explicitly introduce a contrastive character (reflected in total reading times, second pass reading times, total number of fixations, and regressions). The integration of emphasized names was therefore facilitated only when the context provided a contrastive character. In addition to strengthening memory representations for focused elements themselves, focus has been shown to enhance memory for alternatives to the focused element— members of its contrast set. Fraundorf et al. () showed that, in a sentence verification task following discourses which introduced an explicit contrast set followed by a focused element (), a contrastive L+H* accent selectively facilitated correct rejections of explicitly-mentioned contrast set members (b), and did not facilitate correct rejections of same-category associates that were not mentioned as part of the contrast set (c). ()
a. Both the British and the French biologists had been searching Malaysia and Indonesia for the endangered monkeys. b. Finally, the BRITISH spotted one of the monkeys in MALAYSIA and planted a radio tag on it.
()
a. The British scientists spotted the endangered monkey and tagged it. b. The French scientists spotted the endangered monkey and tagged it. c. The Portuguese scientists spotted the endangered monkey and tagged it.
Similar findings were reported by Spalek et al. (), who manipulated focus using the German exclusive particle nur (‘only’) and the inclusive particle sogar (‘even’). Participants heard discourses that explicitly introduced a contrast set, then continued by mentioning an element from that set with nur, sogar, or no particle. When recalling elements from the initial context sentence after a delay, participants recalled alternatives to the focused element better in both focus particle conditions, relative to the no particle condition. This was despite lexical differences between the two particles: while inclusive particles mark the alternatives to be part of the predication, exclusives express that the alternatives to the focused constituent do not hold. These findings suggest that focus and contrast are processed differently from unfocused material, but in a way that may not be sensitive to lexical differences.
. F
.................................................................................................................................. Focus, and particularly the presence of explicit contrast sets in the discourse, played an important role in the debates over structural versus constraint-based models of sentence processing (Frazier, ; MacDonald et al., ; Trueswell & Tanenhaus, ). Much of this literature centred around temporary syntactic ambiguities like (), where loaned is temporarily compatible with being the main verb of the sentence (‘The businessmen loaned
.
people money at low interest’) or starting a reduced relative clause (‘The businessmen who were loaned money at low interest . . . ’). ()
The businessmen loaned money at low interest were told to record their expenses.
()
Only businessmen loaned money at low interest were told to record their expenses.
According to structural theories, any sentence should receive an initial parse based solely on structurally-defined principles. However, Ni et al. () showed that replacing the definite determiner the with the focus particle only (as in ()) reduced the garden path effect in the disambiguating region. What only does here is introduce an implicit contrast set into the discourse: the businessmen mentioned in the sentence are contrasted with some unmentioned set of individuals. A continuation where loaned introduces a restrictive relative clause would supply exactly the kind of contrast set needed (businessmen who were loaned money at low interest, businessmen who were not loaned money at low interest). By creating an expectation for contrastive information and thereby making the relative clause parse more likely, only makes it easier to revise an incorrect main verb parse to the correct relative clause parse later in the sentence.1 The increased reading times normally associated with making this revision is therefore reduced in sentences with only. Sedivy () replicated Ni et al.’s only effect, and built on this by manipulating whether an explicit contrast set was introduced in the prior discourse. The inclusion of an explicit contrast in the discourse context () eliminated the facilitatory effect of focus on reading of reduced relatives. ()
All of the secretaries and accountants were made to take a tough computing course.
()
Only the secretaries prepared for the exam . . . a. . . . and earned significant pay raises. (main verb continuation) b. . . . passed and earned pay raises. (relative clause continuation)
Paterson et al. (; see also Filik et al., ) argue that the reduced garden path observed for reduced relative clause sentences with only does not hold for all reduced relative clauses. In particular, they argue that relative clause sentences starting with a noun phrase, verb, noun phrase sequence () were resistant to only-induced contrast effects because of the overwhelming preference for a Subject-Verb-Object parse (cf. the SVO Heuristic from Bever, ). ()
The/Only teenagers (who were) allowed a party invited a juggler straightaway.
In an eye-tracking reading study, participants read reduced and unreduced sentences with and without the focus operator only. There were longer first-pass reading times in the Note that if (16) is initially construed as ‘Only businessmen and no other individuals’, a relative clause parse would be no more expected than in (15). The reduced garden path in (16) suggests that only creates an expectation for a more concrete set of implicit alternatives—in other words, {businessmen who were loaned money at low interest, businessmen who were not loaned money at low interest} is more expected as a contrast set than {businessmen, non-businessmen}. 1
critical region of reduced sentences than in the same region of unreduced sentences, regardless of the inclusion of only. The authors conclude from this result that only is sometimes unable to modulate a garden-path effect when one parse is very strongly preferred. However, it should be noted that Ni et al. () and Sedivy () did not claim that only should affect parsing uniformly; rather, they were arguing against processing models that excluded the possibility of discourse-level contrast influencing the initial parse of a sentence. Prosodically-marked focus has also been shown to influence syntactic attachment (Speer et al., ; Schafer, ; Schafer et al., ; see also Tonhauser, Chapter in this volume, for in-depth discussion of prosodically realized focus). Schafer () showed that prosodically-signalled focus can influence syntactic attachment preferences. A H* accent on a noun increased the likelihood that a subsequent relative clause was interpreted as modifying that noun; when a sentence contained two nouns that could serve as the head of a relative clause, an accent on one of the nouns disambiguated the sentence in favour of the accented noun. Thus, the relative clause was more frequently interpreted as modifying the noun plane when it was accented () than when the noun propeller was accented (), and vice versa. ()
The sun sparkled on the propeller near the PLANE that the mechanic was so carefully examining.
()
The sun sparkled on the PROPELLER near the plane that the mechanic was so carefully examining.
A similar attachment effect is observed with embedded wh-words that are ambiguous between an interrogative wh-element introducing an embedded question and a relativizer (Schafer et al., ): a H* pitch accent on the wh-word () led to more embedded question interpretations than when a different constituent was accented (). ()
I asked the pretty little girl WHO is cold.
()
I asked the pretty little girl who is COLD.
Because interrogative wh-elements must be focused, accent on an embedded wh-word supports an analysis where the wh-word is functioning as an interrogative constituent and introducing an embedded question. On the other hand, wh-relativizers are not normally focused, therefore accenting would be incompatible with an analysis where the wh-word introduces a relative clause.
. F
..................................................................................................................................
.. Focus as cue to discourse structure One way in which focusing has been shown to influence interpretation is by serving as a cue to some aspect of the discourse structure. In an environment where an elided verb phrase
.
had two potential antecedents, a contrastive pitch accent (L+H*) on a subject preceding one of the potential antecedents (‘said Maria went to the rally’ in (a), ‘went to the rally’ in (b)) made that verb phrase more likely to be interpreted as the antecedent of the elided verb phrase (a)‒(b) (Frazier et al., ). ()
a. JULIE said Maria went to the rally . . . b. Julie said MARIA went to the rally . . .
()
. . . and GREG did too. a. (Greg said Maria went to the rally) b. (Greg went to the rally)
As Kehler () has argued (see also Kehler, ; Kertz, ), antecedent and ellipsis clause pairs like those in ()‒() exemplify the Resemblance coherence relation, supported by the use of the connective and, and the interpretation of the elided VP as syntactically parallel to the antecedent in terms of argument realization. What the contrastive pitch accents in () might be doing, then, is serving as a cue to the underlying discourse structure—whether that structure is ultimately construed in terms of coherence relations (Kehler, ) or information structure (Kertz, ). The Frazier et al. () findings might also be explained by appealing to Question Under Discussion (QUD; Roberts, ): the difference in interpretation between (a) and (b) could be due to a difference between the QUD ‘What did each person say?’ and ‘What did Julie say?’. Clifton & Frazier () directly tested the effects of focus-evoked QUD on discourse interpretation. In an eye-tracking during reading study, reading was disrupted when an answer to an explicit question did not answer the question in a way that conformed to the subcategorization biases of the verb used in the question. Further, when a previously introduced question ended up not being addressed by the end of a discourse, reading was disrupted, suggesting that readers maintained an expectation for QUD resolution while reading a discourse. Relatedly, Cummins & Rohde () demonstrated that focus placement could indirectly influence interpretation in a number of pragmatic phenomena by influencing the QUD inferred by listeners. They used a listening task where contrastive pitch accent was placed on a target element: a scalar adjective (), an element embedded under a negated presupposition trigger (), or a potential antecedent of a pronoun (). ()
The view from the hotel window is pretty/PRETTY.
()
Bill doesn’t regret arguing with his boss/BOSS.
()
Charles/CHARLES congratulated Simon. He had criticized Stephanie.
In each case, focus highlighted the QUD, perhaps making listeners infer a contrastive question rather than a non-contrastive informational one. For scalar adjectives, this increased the availability of scalar alternatives, increasing the availability of scalar implicatures (‘The view from the hotel window is not gorgeous’). Similarly, contrastively focusing an element embedded under a presupposition trigger in () may have highlighted the QUD ‘Who is it that Bill doesn’t regret arguing with?’ (as opposed to e.g. ‘Does Bill regret arguing with his boss?’), making the presupposition that Bill argued with his boss less
likely to project. Relative to a baseline preference for interpreting ‘Simon’ as the antecedent of ‘he’ in (), contrastive focus on ‘Charles’ increased interpretations of ‘Charles’ as the antecedent. In all cases, focus placement could be argued to have influenced interpretation indirectly, by serving as a cue to QUD (see Degen & Tanenhaus, Chapter in this volume, for discussion of the role of QUD in pragmatic processing, including the interpretation of scalar implicatures).
.. Anaphoric dependencies Much of the discussion of focus in relation to pronoun resolution has centred on the timecourse of implicit causality effects (see Rohde, Chapter in this volume, for discussion of implicit causality effects on pronoun interpretation). The implicit causality of a verb refers to the bias to attribute causality to one of its arguments. When asked to provide a cause for a sentence like ‘Ann envied Bill’, for example, people tend to produce causes associated with Bill (e.g. ‘ . . . because he had a beach house’). NP-biasing verbs like envy contrast with NP-biasing verbs like anger, which tend to elicit causes associated with the subject NP (‘Ann angered Bill because she knocked over his drink’). The influence of verb bias on pronoun resolution was demonstrated by studies in the s (Garvey and Caramazza, ; Garvey et al., ; Brown and Fish, ; Au, ); for example, Caramazza et al. () showed that participants’ interpretation of the pronoun in a sentence like () (from Stewart et al. ()) was consistent with the verb’s implicit causality, and that responses were faster when the subordinate clause was congruent with verb bias. ()
Roy questioned Anthony because he wanted to learn the truth.
The debate over the mechanism underlying implicit causality effects has often boiled down to a difference in predictions about time-course. One prominent theory about the source of these effects appeals to the focusing properties of implicit causality: by focusing attention on the cause of an action, implicit causality facilitates comprehension of a subsequent pronoun referring to that individual. Because proponents of the Focusing account link implicit causality effects directly to verb bias, they predict that these effects should emerge in early stages of processing, very soon after the pronoun is encountered. By contrast, Integration accounts associate implicit causality effects with later-stage processing, when the two clauses are integrated based on their causal relationship; on such accounts, implicit causality effects on pronoun resolution are expected to emerge late. McDonald & MacWhinney () argued for the Focusing account using a cross-modal probe task involving auditory presentation of a sentence like () and visual presentation of a probe word. For NP-biasing verbs, NP was processed faster than NP at points before the pronoun and milliseconds (ms) after the pronoun (perhaps reflecting a firstmention advantage), but immediately after the pronoun, reaction times were similar for both NP and NP. Greene & McKoon () also reported a greater reduction in reaction times for NP probe words than for NP probe words, when probe words were presented after an NP-biasing verb rather than before the verb. Because these effects occurred either immediately after the pronoun or before the pronoun but after the biasing verb, these results were used to argue for the Focusing account.
.
Garnham et al. () employed an all-visual probe task, and tested before and after the pronoun and at the end of the sentence. Self-paced reading was used to assess effects of congruency—whether the implicit cause associated with the verb matched the explicit cause given in the sentence—and anaphor type (name or pronoun). ()
Jean congratulated Rita (vigorously) because she/Rita had won the championship. (congruent)
()
Jean congratulated Rita (vigorously) because she/Jean was very impressed. (incongruent)
The absence of a congruency-by-anaphor type interaction, with a congruency effect appearing for pronouns but not for names, was construed as evidence against the Focusing account, and for the Integration account. Because the debate is about whether attentional focus plays a role in the mechanism underlying implicit causality effects, studies like these do not typically involve manipulating linguistic focus. One exception is Itzhak & Baum (): against the background of implicit causality effects, they asked how contrastive focus (conveyed by a contrastive pitch accent) interacted with verb bias to influence reference resolution in a visual world experiment (Tanenhaus et al., ). In critical conditions, implicit causality bias was pitted against contrastive accent: in (), the verb envy biases the pronoun to be interpreted as referring to Bill, but the antecedent John is contrastively accented. ()
John and Bill both care about money a lot.
()
JOHN/John envied Bill when they were young because he came from a rich family.
While the bias-inconsistent referent (John) received fewer fixations overall, looks to the bias-inconsistent referent increased when the corresponding name was contrastively accented. This effect emerges at the pronoun, and prior to disambiguating information; contrastive focus therefore seems to be able to resolve the potentially misleading information from verb bias.
.. Focus and reference resolution in visual displays The Itzhak & Baum () study builds on a line of research investigating prosodicallymarked contrast on reference resolution in visual displays. Dahan et al. () found in a visual world task that participants rapidly interpret accented noun phrases as referring to discourse-new referents and deaccented noun phrases as referring to given referents. In a first instruction, participants were asked to move an object in a display (); a second instruction used either an accented (L+H* or H*) or a deaccented noun to refer to the same object (candles), or to a discourse-new competitor (candy) (a)‒(b). ()
Put the candle above the triangle.
()
a. Now put the CANDLE/CANDY above the square. b. Now put the candle/candy ABOVE THE SQUARE.
Participants looked more often to the competitor (candy) when the noun in the second instruction was accented and less often when it was deaccented, indicating that listeners made immediate use of pitch accent information and its relationship to discourse status in order to resolve reference. In another visual world eye-tracking study, Weber et al. () found that German listeners looked earlier at the picture of a referent belonging to a contrast pair (red scissors, as opposed to red vase) when instructions to click on it carried a contrastive accent on the colour adjective (a) than when the adjective was not accented (b). ()
Click on the purple scissors.
()
a. Now click on the RED scissors/vase. b. Now click on the red SCISSORS/VASE.
In addition, there was a general preference to interpret adjectives contrastively, whether the adjective was accented or not; this was reflected in a general preference for the contrast set member (red scissors) relative to its competitor (red vase). In a second experiment, however, introducing the first member of the contrast pair (purple scissors) with a contrastive accent resulted in the subsequent adjective being interpreted contrastively only when contrastively accented. This suggests that contrastive focus is sensitive to the prior prosodic context, in addition to the discourse status of potential referents. Kurumada et al. () provided further evidence for the incremental computation of contrastive inferences on the basis of visual context and contrastive prosody. Participants in a visual world experiment heard ‘It looks like an X’ pronounced with either a noncontrastive accent on the final noun () or a contrastive accent on the verb (). ()
It looks like a ZEbra. H* L-L%
()
It LOOKS like a zebra. L+H* L-H%
When the visual display contained a single contrast set (e.g. a zebra and a zebra-like animal), listeners made anticipatory eye movements to the contrast set upon hearing contrastively-accented LOOKS. Presumably, this is because the contrastive focus on looks made the contrast between ‘It LOOKS like X’ and ‘It IS X’ highly salient. As the effects of contrastive focus emerged before auditory input about the target noun (zebra) was available, this result shows that contrastive prosody, together with information from the visual context, gives rise to immediate inferences about reference. Prosodic information is also used to signal which element under the scope of a focus particle should be focused. In a visual world study, Gennari et al. () manipulated whether the element that associated with only was unstressed or stressed (). ()
The mother only brought some milk/SOME MILK to the boy.
Participants looked more often to the referent corresponding to the focused element and a contrast set member when the focused element was stressed, suggesting that prosodic
.
information was integrated rapidly and used to identify focus under the scope of the focus particle. In a similar study in Dutch, Mulders & Szendrői () used accent on one of two arguments of the verb to signal which argument should associate with the Dutch focus particle alleen (‘only’). In a picture matching task, participants judged whether sentences like ()‒() matched a picture. ()
Ik heb alleen SELDERIJ aan de brandweerman gegeven. ‘I only gave CELERY to the fireman.’
()
Ik heb alleen selderij aan de BRANDWEERMAN gegeven. ‘I only gave celery to the FIREMAN’.
.. Inferring implicit alternatives As in the studies described earlier (section .) on the impacts of focus on memory (e.g. Fraundorf et al., ), some recent research related to interpretive effects of focus has turned its attention to the status of the alternatives evoked by focus, rather than the focused element itself. In many prior studies, contrast sets associated with focused elements were explicitly introduced in the surrounding discourse: reading studies investigating effects of only on subsequent syntactic attachment introduced explicit contrast set members in a preceding sentence (e.g. Sedivy, ); other studies showing that focus-incongruent continuations disrupted reading did so by explicitly mentioning a contrast set member (e.g. Paterson et al., ; ‘Jane passed only the salt . . . and not the pepper’). Kim et al. () showed that, as in Sedivy ()’s self-paced reading study, comprehenders made rapid inferences about contrast based on the discourse context and the presence of only: listeners in a visual world experiment made anticipatory looks to a focused element in the scope of only, when explicit alternatives had been mentioned in the preceding discourse. In addition, listeners readily interpreted unmentioned associates based on either lexical () or situation-based () categories as focus alternatives. ()
a. Neil has some pears and some oranges. b. Alex only has some apples.
()
a. Neil and Alex are at the baseball game. b. Alex wants to buy some Coke and some nachos. c. Neil only wants to buy some hot dogs.
Interestingly, the implicit associates showed general facilitation irrespective of the presence of only, unlike explicitly mentioned alternatives, which were selectively facilitated in sentences with only. Using cross-modal priming, Husband & Ferreira () investigated the time-course of activation of contrastive (a) versus non-contrastive (b) alternatives to a contrastively accented prime word (NURSE in ()). ()
The murderer killed the nurse/NURSE last Tuesday night.
()
a. contrastive: doctor b. non-contrastive: clinic c. unrelated: plug
In early processing, contrastive focus led to activation of both contrastive and noncontrastive associates relative to unrelated items. By contrast, in later processing, contrastive associates remained activated while non-contrastive associates were deactivated. Similarly, Götzner & Spalek () used a probe recognition experiment to compare the activation of contrastive alternatives to a focused element with non-contrastive conceptual associates under the scope of a focus particle (German nur ‘only’ or auch ‘also’). ()
Carsten wollte gern Obst essen und griff in einen Korb. ‘Carsten wanted to eat some fruit and reached into a basket’
()
Er nahm sich {nur,auch,___} (Äpfel)F heraus. ‘He {only,also,___} took (apples)F out of it’
Contrastive probes (e.g. ‘berries’) took longer for participants to reject, while noncontrastive probes (nouns related to the focused element by situational knowledge, e.g. ‘maggots’) were unaffected. The authors concluded that the computation of focus alternatives takes into account whether an alternative can replace the focused element. One question raised by these studies is why Braun & Tagliapietra () and Husband & Ferreira () observed (initial) facilitation for non-contrastive associates, while Götzner & Spalek () observed inhibition. Götzner & Spalek () suggest the difference is related to methodology: their study uses difficulty of rejecting probes as a measure of activation, whereas Husband & Ferreira () use facilitation of a cross-modal prime. Differences in sensitivity to low-level activation between these paradigms could be responsible for the different findings for non-contrastive alternatives. However, another possibly relevant factor is how the difference between contrastive and non-contrastive associates has been operationalized in different studies. Recall that, in Fraundorf et al. ()’s sentence verification experiment, a contrastive accent on a focused element increased correct rejections of explicitly mentioned contrast set members, but did not affect correct rejections of same-category unmentioned associates. In that experiment, non-contrastive alternatives were unmentioned associates that could replace the focused element (Portuguese vs. British). Kim et al. () did not distinguish between contrastive and non-contrastive alternatives, but used conceptual associates (both lexical associates and associates evoked by situational information) which were able to replace the focused element. Husband & Ferreira () used Latent Semantic Analysis to select contrastive associates that had higher similarity scores to prime words than noncontrastive associates did; in addition, contrastive associates were able to replace prime words and preserve both grammaticality and plausibility, whereas substituting non-contrastive associates for prime words would result in either ungrammaticality or implausibility. According to Götzner & Spalek (), their non-contrastive associates were related to the focused element by world knowledge, but the key criterion distinguishing contrastive from non-contrastive associates was replaceability—whether an associate could have replaced the focused element. Thus ‘contrastive’ in these studies is characterized by explicit mention, conceptual similarity,
.
replaceability, or some combination of these properties. In evaluating future studies, it will be important to keep track of the way in which contrastive and non-contrastive alternatives are characterized.
. C
..................................................................................................................................
.. Conflicting cues to focus Most prior studies manipulated a single cue to focus—clefting, use of a focus particle, the prior discourse context—while holding other cues constant. When different cues signal different focus structures, how are these conflicts resolved? Paterson et al. () showed that reading was disrupted when an explicit contrast item introduced after the focus particle only was incompatible with the focus structure introduced by only. In dative or double object sentences where either the direct object or the indirect object was preceded by ‘only’, readers showed increased regression path and total reading times when reading a continuation mentioning an incongruent alternative to the focused element (e.g. (a) continued with (b)). ()
a. At dinner last night, Jane passed only the salt to her mother . . . b. At dinner last night, Jane passed the salt to only her mother . . .
()
a. . . . but not the pepper as well because she couldn’t reach. b. . . . but not her father as well because she couldn’t reach.
In another reading study using eye-tracking, Sauermann et al. () similarly pitted contextual cues to focus against lexical cues. Embedded wh-questions were used to focus either the direct or indirect object in a double object sentence, which was followed by a continuation in which the placement of only signalled a focus structure that was either congruent or incongruent with the preceding context. When only preceded both objects as in (), continuations with an explicitly excluded alternative (not the cherries/grownups) compatible with the focus structure associated with only (a) resulted in slower reading times for both the excluded alternative and the following region, compared to continuations where the excluded alternative was compatible with the focus structure supported by the context (b). This was taken to suggest that the conflict between lexical and contextual focus cues was resolved in favour of the contextually-cued focus structure. ()
John wondered what Sally would pass the children.
()
Sally passed only the children the apples . . .
()
a. . . . but not the cherries, because they did not want them. b. . . . but not the grownups, because they did not want them.
However, when only appeared after the indirect object and associated with the direct object as in (), cue conflict disrupted processing earlier in the sentence, and reading times
during and after the excluded alternative did not favour either lexically cued (a) or contextually cued (b) focus structures. ()
John wondered who Sally would pass the apples.
()
Sally passed the children only the apples . . .
()
a. . . . but not the grownups, because they did not want them. b. . . . but not the cherries, because they did not want them.
It remains unclear why the conflicting cues to focus structure seem to resolve in different ways depending on the placement of the focus particle. However, it may be relevant that, in sentences like (), only can associate with the indirect object the children as well as with the direct object the apples, a difference that would have been signalled prosodically in auditory stimuli. Event-related brain potential (ERP) studies have also investigated cases of conflicting cues to focus structure (Hruska et al., ; Stolterfoht et al., ). For example, Cowles et al. () measured ERP responses to violations of focus structure, where focus was manipulated contextually, using wh-questions, and using syntactic clefts. In addition to finding a large positivity associated with clefting, in general consistent with sentence-final integration effects, they found that focusing a element incongruent with the focus structure determined by the wh-question elicited an N-like effect.
.. Cues to focus projection Another important question has been how an accented element projects focus to a larger constituent (Gussenhoven, ; Selkirk, ; Gussenhoven, ; see also Schwarz, Chapter in this volume). Birch & Clifton () manipulated both accent placement and context in auditorily presented question-answer pairs; listeners made judgements about the appropriateness of the prosody or made judgements based on their comprehension of the exchange. In the appropriateness judgements, accenting a discourse-new noun phrase (b) was judged less appropriate than accenting both the verb and noun phrase comprising the verb phrase (a), suggesting that the accent on the noun phrase was not successful in projecting focus to the entire verb phrase. ()
Isn’t Kerry pretty smart?
()
a. Yes, she TEACHES MATH. b. Yes, she teaches MATH.
By contrast, comprehension of sentences with VP focus was as good when only an argument NP was accented as when both the NP and V were accented. This disparity suggests a task-related difference: NP focus did project to the VP in terms of conveying the intended meaning to listeners, but listeners also showed a preference for both NP and V to be accented in conveying that meaning. Using the same methodology, Birch & Clifton () showed that accented adjuncts do not project to larger constituents they are part of. Linguistic analyses of intonational
.
structure have indicated that some syntactic constituents, when accented, can project focus to an entire phrase, but other constituents, including adjuncts, cannot project focus. When both the verb and the adjunct were discourse-new, both appropriateness ratings and comprehensibility showed a preference for both verb and adjunct being accented (a) compared to just the verb (b). ()
a. He DROVE SPEEDILY. b. He DROVE speedily.
In addition to the presence of a pitch accent on the focused constituent, Welby () shows that focus projection is sensitive to the deaccenting of following unfocused material.
. F
..................................................................................................................................
.. Different focus operators The focus particle only has been a prominent feature in reading time studies. However, the alternative-triggering properties of only are also meant to be shared by other focus-sensitive lexical items (e.g. also, even, too). More recently, studies have begun to investigate differences between focus particles in terms of the time-course of their influence of interpretation, and the nature of the relationship they specify between the focused element and its alternatives. For example, Filik et al. () compared sentences with either only or even with continuations that were either congruent or incongruent with expectations triggered by the focus particle, as in (). ()
Only/Even students taught by the best/worst teacher passed the examination in the summer.
While reading times in the only sentences reflected a difference between likely and unlikely continuations in the critical region (passed the examination), comparable differences only emerged in the post-critical region (in the summer) for sentences with even, suggesting that different focusing mechanisms—even different focus particles— can differ in the time-course and perhaps complexity of evoking contrast. In the case of even, the alternatives triggered may be more complex because they must be ordered on a scale of likelihood (or another contextually salient scale): ‘Even students taught by the worst teacher passed the examination’ gives rise to the inference that the students taught by the worst teacher were the least likely to pass the examination. The corresponding only sentences also encourage a scalar interpretation (the students taught by the best teacher were the most likely to pass the exam), but a non-scalar exclusive reading is also always possible. The complexity of the alternative set required to interpret a focus particle may in turn influence the richness of the context needed to support rapid online interpretation.
.. Contrastiveness and the sources of contrast sets As discussed in section .., recent studies have begun to investigate the mechanisms underlying the computation of focus alternatives. An important question in future research will be how contrastiveness is characterized—what makes contrastive alternatives different from non-contrastive ones? While alternatives explicitly introduced into the discourse are clearly facilitated in both processing and recall measures, the status of implicit alternatives that are either conceptual associates of the focused element or cannot substitute for the focus is less clear. Part of understanding where contrast sets come from will involve understanding how non-linguistic context informs comprehenders’ inferences about linguistic alternatives. In reference resolution studies, the non-linguistic context is approximated by visual context. However, it will be important to understand how speakers and hearers estimate contextually salient alternatives based not only on visual information, but also prior knowledge, and knowledge about shared and unshared knowledge among interlocutors.
.. Cross-linguistic variation Investigations into focus in English are necessarily constrained by properties of English, such as its rigid word order. Languages with different ordering, morphosyntactic, or tonal properties allow for more nuanced understanding of how the grammatical properties of a language influence how different cues to focus are used in a language. For example, Keller & Alexopoulou () investigated the relative contributions of word order, accent placement, and clitic doubling to the realization of focus in Greek, a free word order language. Ouyang & Kaiser () explore how information status and corrective focus are realized in the prosody of Mandarin, a tone language where acoustic cues are also used to distinguish lexical items. Even across languages that share relative flexibility in terms of word order, focus phenomena may differ in terms of their sensitivity to discourse factors. In a comparative study, Skopeteas & Fanselow () found that contrastive interpretation of object-fronting is context-dependent and sensitive to referent predictability in some languages (German, Spanish, Greek), but not others (Hungarian). Similarly, Turnbull et al. () suggest that contextual predictability affects the prosodic realization of focus in a way that may vary across languages. More broadly, different cues to focus vary in terms of their inherent reliability—for example, prosodic cues are inherently more gradient than word order cues. Looking at typologically different languages, future research might also ask whether languages are efficient in how they combine the cues available to them for conveying focus.
......................................................................................................................
......................................................................................................................
. I: W NPI?
.................................................................................................................................. N Polarity Items (NPIs) are a class of lexical items that need to be licensed by certain conditions. For instance, the English NPIs any and ever typically appear in the semantic and syntactic scope (i.e. the c-command domain) of an NPI licensor, such as negation (Klima, ; Ladusaw, , ). As shown in (), any and ever are grammatically licensed when they appear within the scope of negation (a,b), but they are ungrammatical when there is no negation present (c,d), or when a negation is present, but doesn’t c-command the NPI (e,f). ()
a. b. *c. *d. *e. *f.
John didn’t talk to anybody. John hasn’t ever talked to Bill. John has ever talked to Bill. John has talked to anybody. Anybody didn’t talk to John. The debate that nobody cared about will ever end.
Because of their apparent sensitivity to the presence of negation, any and ever are labelled ‘negative’ polarity items (NPIs).1 Although negation is a cross-linguistically attested licensor for NPIs, it must be noted that their distribution (within a single language or crosslinguistically) is quite broad and includes a vast range of negative and non-negative licensors, including negative quantifiers, conditionals, modal verbs, generic sentences, imperatives, questions, the scope of certain universal quantifiers and disjunctions (see Giannakidou, , for a review). At the same time, many other lexical items, in addition to any and ever, can be NPIs, and the set of NPIs can vary from language to language. 1
Some NPIs, such as any, also obtain so-called free choice readings in modal environments and with imperatives, such as You may talk to any student, and Pick any card! We won’t discuss the free choice use in the current chapter.
We give some English examples of both NPIs and licensors in () (examples taken from Linebarger, ), with each licensor and NPI underlined: ()
a. John didn’tknow any French. b. Few people have any interest in this. c. Only John has a hope in hell of passing. d. Everyone who knows a damn thing about English knows this word. e. If you steal any food, they’ll arrest you. f. He was taller than we ever thought he would be. g. He refused to budge an inch.
Given the broad distribution of NPIs in English and cross-linguistically, a large body of the formal linguistic literature has focused on the licensing question. What is the essential grammatical property that can systematically account for the distribution of NPIs? Since NPIs do not seem to form a homogenous class, either within a single language or crosslinguistically (Zwarts, , ; Haspelmath, ; Giannakidou, ), it is likely that there isn’t one single licensing condition that can explain everything, and an ultimate characterization of the licensing conditions cannot be derived without a close look at the variation in NPIs themselves (within and across languages). Indeed, various notions of licensing have been proposed in the literature (Baker, ; Fauconnier, ; Ladusaw, , ; Linebarger, , ; Kadmon & Landman, ; Hoeksema, ; Krifka, a; Zwarts, , ; Giannakidou, , ; Lahiri, ; Von Fintel, ; Chierchia, ). The primary focus of this chapter is to present an overview of the experimental studies that have investigated the comprehension and acquisition of NPIs. But before we do that, we need to first understand the complexity of the linguistic properties of NPIs. In the sections later, we will first give a brief overview of the major linguistic notions that have been shown to be important for NPI licensing (section .), and then we will present the recent findings and puzzles revealed by experimental studies in sentence processing and first language acquisition (sections . and .).
. L
.................................................................................................................................. In this section we briefly review three lines of research that have been important for our understanding of NPI licensing as a natural phenomenon.
.. Downward entailment According to the influential Downward Entailing (DE) hypothesis (Fauconnier, ; Ladusaw, , ; see also Hoeksema, ; Von Fintel, ; among others), NPIs are licensed in the scope of the DE operators. ()
DE function: A function f is DE iff for every arbitrary element X and Y, it holds that X Y ! f(Y) f(X)
DE functions reverse entailment relations. Negation is a typical DE operator. Consider () as an example. The set of red cars is a subset of all the cars. In an affirmative statement (b), the entailment relation holds from the subset (the red cars) to the superset (the cars), but not vice versa (a). The entailment relation reverses, however, in (c) and (d), which contain a sentential negation. ()
a. b. c. d.
John bought a car. ⊭ John bought a red car. John bought a red car. ⊨ John bought a car. John didn’t buy a car. ⊨ John didn’t buy a red car. John didn’t buy a red car. ⊭ John didn’t buy a car.
The DE hypothesis is extremely influential because it provides a unified semantic characterization of NPI licensing for the first time. It elegantly captures the fact that a set of seemingly unrelated items could all license NPIs because of their logical properties. Take the universal quantifier every as an example. It licenses NPIs in its restrictor, but not its scope. This is predicted because only the restrictor of every is DE, whereas its scope is not. The correlation between the NPI-licensing behaviour of every and its DEness is demonstrated in the following two examples. ()
The restrictor of every is DE and licenses NPIs a. Every student wrote a paper for this class. ⊨ Every student who enjoyed the class wrote a paper for this class. b. Every student who did any homework for this class got an A on the final exam.
()
The scope of every is not DE and doesn’t license NPIs a. Every student wrote a paper for this class. ⊭ Every student wrote a paper on language change for this class. *b. Every student who did their homework for this class ever got an A on the exams.
Despite its intuitive appeal, one empirical problem for the DE hypothesis is that NPIs can be licensed in non-DE contexts as well. The following is a (non-exhaustive) list of the NPI-licensing contexts that are non-monotonic (examples are taken from Giannakidou, ; Von Fintel, ; and Rothschild, ). ()
a. Antecedents of conditionals If John had walked on any weeds he would have tracked dirt into the house. b. ‘most’ Most men with any revolutionary commitments were executed. c. ‘only’ Only John did any work. d. Exactly n Exactly three people with any money showed up. e. Adversative predicates Sandy regrets that Robin bought any car.
f. Superlatives (and comparatives) Emma is the tallest girl to ever win the dance contest. g. yes-no non-rhetorical questions Did Lucy see anyone? One approach to solving this problem is to refine the original notion of DE. Von Fintel () proposed to replace DE with a weakened condition, Strawson DE. A sentence S Strawson-downward-entails another sentence S’ if, when the presuppositions of both S and S’ are satisfied, S entails S’. For example, the sentence Only John ate vegetables doesn’t entail that Only John ate kale in the strict sense, because the presupposition of the second sentence John ate kale isn’t entailed by the first sentence. However, with the notion of Strawson DE, we would restrict the inference test to only those situations in which the presuppositions of both sentences are fulfilled. In a situation in which we know that John ate kale, then if Only John ate vegetables is true, Only John ate kale is also true. The Strawson DE criterion would correctly classify only as a NPI licensor, whereas the traditional DE criterion fails. Limited by space, we won’t go into a full discussion of the theoretical and empirical advantages of the Strawson DE condition discussed in Von Fintel () (also see Gajewski, ; Guerzoni & Sharvit, ). But we note that it has also been argued that Strawson DE still doesn’t suffice to account for all the empirical problems that the traditional DE account encounters. We refer readers to discussion of the problems in Giannakidou (), Rothschild (), and Homer ().
.. Non-veridicality Observing that DE operators capture a subset but not all of the NPI licensors, Giannakidou (, ) proposed (non-)veridicality sensitivity as the licensing condition for NPIs (see also Zwarts, ). Non-veridical contexts are those that do not entail or presuppose the truth of a proposition p. Sentential negation, being the most robust NPI licensor cross-linguistically, is not only non-veridical, but anti-veridical, since the proposition under negation is made necessarily false. The antecedent of a conditional, which also licenses NPIs, is non-veridical, because the truth of the proposition in the antecedent of a conditional is not warranted. The non-veridicality condition achieves broad empirical coverage. Giannakidou (, ) argued that many of the licensing environments that are not DE are non-veridical, including questions, conditionals, modal contexts, future tense, imperatives, etc. The non-veridicality condition connects NPI licensing to a previous observation that NPI licensors appear to have graded strength depending on how negative they are (Zwarts, , , ). In Zwarts’s system, the strength of negation is a function of how many of De Morgan’s laws are satisfied by an expression. Some expressions, such as anti-additive and anti-morphic negations, are strong negations; mere DE quantifiers, such as few, are forms of the weaker negations. But the non-veridical licensors include also non-DE expressions (as discussed earlier), and they satisfy the fewest of De Morgan’s laws. From this perspective, there is a hierarchy among different classes of NPI licensors, and nonveridical licensors form a superset of the others.
.. Domain widening and exhaustifying the alternatives Kadmon & Landman () made the argument that the semantic/pragmatic contribution of any is to make a stronger statement than a regular indefinite noun phrase. They proposed that the semantics of any is the regular semantics of an indefinite noun phrase plus an additional domain widening function. Domain widening widens the interpretation of the regular noun phrase along a certain contextual dimension. Any is licensed only if the widened domain it introduces creates a stronger statement. This analysis essentially introduces a scalar component to the semantics of any. By uttering I don’t have any potatoes, the speaker could mean that he doesn’t have even a rotten potato (not to mention a normal potato), or even one single potato (not to mention a larger quantity). The exact scalar dimension is determined by context. The (implicit) scalar component in English any is explicitly introduced in some other languages. For instance, in Hindi, an overt morpheme even is part of the NPI form (Lahiri, ). Strengthening with domain widening can be naturally achieved with DE environments, since the assertion in which a DE operator quantifies over a larger/widened domain would entail the assertion with a smaller domain. This account therefore explains rather than stipulates why DE-operators are prominent NPI licensors (see also Krifka, a). Building upon the insights of Kadmon & Landman () and Krifka (a), Chierchia (, ) further implemented a compositional analysis. Without going into all the technical details, in a nutshell, the basic idea is this. The NPI any lexically activates alternatives. A covert ‘only’ is introduced such that during the semantic composition of a proposition, alternatives that are not entailed by the assertion need to be eliminated (i.e. exhaustifying the alternatives). It turns out that the exhaustification process leads to contradictions in Upward Entailing (UE) contexts, but not in DE contexts, and therefore NPIs (specifically the scalar type)2 are licensed in DE but not UE contexts.
.. Summary The quick and brief overview in this section, although not doing full justice to the large literature, already reveals that NPI licensing involves satisfying constraints at multiple levels, including syntactic, semantic, and pragmatic constraints. Each of the proposals outlined earlier emphasized a particular dimension of NPI licensing, but it is plausible that we need more than one single condition to account for the full range of data. The multi-dimensionality of NPI licensing becomes even more apparent when we consider how NPIs are processed and comprehended in real time. The experimental studies we introduce in the next section will suggest that to successfully integrate an NPI into its sentential context, the comprehension system needs to recruit multiple different mechanisms, including both grammatical and extra-grammatical ones. 2 There are also many non-scalar NPIs (see Lin, ; Giannakidou, ; Giannakidou & Yoon, ). Chatzikonstantinou () provided experimental evidence to show that scalar and non-scalar NPIs could have different prosodic properties.
. P N P I
..................................................................................................................................
.. Basic sensitivity to the licensing condition Given the complexity of the licensing condition for NPIs, a basic processing question is how quickly people can utilize the relevant grammatical constraints to integrate an NPI word into a sentence. To address this question, a commonly used experimental paradigm is to compare the processing profile of licensed and unlicensed NPIs, and examine how quickly people can detect the linguistic anomaly from the unlicensed NPI. To understand the fine-grained time course of the NPI-licensing process in comprehension, a number of studies have adapted the Event-Related Potentials (ERP) technique, due to its excellent temporal resolution. The two most common ERP components that are relevant for sentence-level language comprehension are the so-called N and P effects. The N is a negative-going waveform that peaks at approximately milliseconds (ms) from the onset of the stimulus presentation, with a primarily centroposterior scalp distribution. Generally speaking, the amplitude of the N evoked by an incoming word indexes the degree to which that word’s semantic features match the semantic features that have been pre-activated by its context at the time it is encountered (Lau, Wagers, Stroud, & Phillips, ; Kutas & Federmeier, ; Kuperberg, ). For example, a highly predictable word sugar in the sentence I like my coffee with cream and sugar elicited a much smaller N response than a semantically incongruent word sock in the same sentential context I like my coffee with cream and sock (Kutas & Hillyard, ). Even though the term ‘pre-activation’ could be associated with active prediction of specific lexical items, we use the term here in a more neutral sense: it refers to the activation of relevant semantic features, regardless of whether active prediction or expectation of the upcoming word is at work, ahead of encountering the full linguistic input. In situations where the context is highly constrained, it is possible the comprehension system may actually anticipate a specific lexical item, especially if it is a content word. But predicting a specific lexical item is not a necessary condition to evoke reduced N amplitude. In the context of NPI licensing, before an NPI is encountered, it is not very plausible to assume that the comprehension system could have been anticipating the specific lexical form of an NPI (such as any or ever). But as long as, when an NPI is encountered, the semantic features/properties processed in the context match the relevant lexical semantic features of the NPI, we should expect to see a reduced N on the NPI. In a series of studies on German (Saddy et al., ; Drenhaus et al., , ; Drenhaus et al., ), a reduced N with a central maximum was found on the German NPI jemals (‘ever’) when it was licensed by negation, compared to the ungrammatical counterpart when the word jemals was not licensed. An example from Saddy et al. () is given in (): ()
a. Kein Mann, der einen Bart hatte, war jemals froh. no man who a beard had was ever happy No man who had a beard was ever happy.
*b. Ein Mann, der einen Bart hatte, war jemals froh. a man who a beard had was ever happy a man who had a beard was ever happy.
N effects were also found for Dutch (Yurchenko et al., ), Turkish (Yanılmaz & Drury, ) and English NPIs (Shao & Neville, ), and in a magnetoencephalography (MEG) study by Tesan et al. (). These findings provide strong evidence that comprehenders can very rapidly integrate the relevant linguistic information to license NPIs in a sentence—by ms after an NPI is presented, language users can already detect whether the NPI appears in the proper context. It should also be noted that the rapid incorporation of the NPI licensing constraints in online comprehension is equally available in languages in which an NPI can linearly precede its licensor. Pablos et al. () found that when the Dutch NPI ook maar iets precedes a negation in a sentence, comprehenders actively anticipate a downstream negation after encountering the NPI, resulting in larger central anterior negativity on the negation if the distance between the NPI and the negation is long. Furthermore, such an effect is most salient when the negation is actually properly licensing the NPI by being in a c-commanding position, but the effect diminishes when the negation is in a non-c-commanding embedded position (see also Yanılmaz & Drury, ). The particular ERP component associated with the NPI licensing process, namely the N response, also sheds important light on the specific mechanisms involved in NPI processing. Under the functional interpretation that N indexes the degree of semantic feature match between a word and its preceding context, the reduced N on a licensed NPI (and conversely the larger N amplitude on an unlicensed NPI) suggests that some semantic features/properties of the sentential context, pre-activated prior to the point at which an NPI is encountered, match certain lexical semantic features of the NPI. Since these studies all manipulated whether the sentential context contains negation or not, these results at least suggest that some property of negation is implicated in the lexical processing of NPIs. However, we do not have enough information to precisely identify the exact NPI licensing property that triggered the N effect. For example, is it the [+Neg] feature that is responsible for the early stage of the NPI licensing process, or is it the non-veridicality of the context, or is it the downward entailment property? All of these possibilities are consistent with the current findings. To tease them apart, one would need to examine a wide range of NPI licensors with non-overlapping properties, and see whether they all facilitate the lexical processing of an NPI during the N time window. In this regard, it is interesting to note that not all the studies on NPI licensing have found increased N amplitudes on the unlicensed NPIs. Drenhaus et al. () compared licensed and unlicensed NPI jemals (‘ever’) in German. Two kinds of licensors were included—negation and wh-questions. NPIs licensed under negation elicited smaller N than the unlicensed NPIs, but NPIs licensed under a wh-question did not lead to any N reduction. The authors attributed the difference between negation and whquestions to their strength of licensing (see Zwarts, , ). But it is worth noting that wh-questions are not negative, and it is possible that the N reduction on licensed NPIs is primarily triggered by negation. In another study, Steinhauer et al. () examined the licensing of three different NPIs in English ever, any, and at all. Although there was an N difference between the licensed and the unlicensed NPI at all, they did not find an
N difference between licensed and unlicensed ever and any. One important difference between this study and the others mentioned earlier is that Steinhauer et al. () tested a larger set of licensors in their stimuli, including various negative licensors such as not, without, rarely/hardly, and also licensors that are not negative per se, but non-veridical, such as every, before, whether, and yes-no questions. It is possible that the N reduction on the NPI ever or any can only be triggered by pre-activated negative features and therefore, that the N effect in Steinhauer et al. () might have been washed out by the use of both negative and non-negative licensors. This post-hoc explanation is obviously very tentative, but it is worth further exploration. Results from both Drenhaus et al. () and Steinhauer et al. () suggest that the lexical processing of NPIs seems to be primarily sensitive to negation, but not nonveridicality in general. This issue is addressed in more detail in a later study in Xiang et al. (), which we will discuss later in this section. An additional finding from Steinhauer et al. () is that not all NPIs are licensed in the same way either. In particular, although no N reduction was found for the licensed ever and any in this study (compared to their respective unlicensed controls), the morphologically complex NPI at all did induce an N effect. Steinhauer et al. () suggested that the morphological complexity of the NPIs themselves may also matter. The majority of the studies reviewed earlier also reported a posteriorly distributed Plike late positivity effect, which was larger for unlicensed NPIs than for licensed ones.3 The P component is a positive-going waveform that peaks at about ms after the onset of a stimulus. This effect was originally associated with syntactic processing, since it is reliably elicited by syntactic errors (Hagoort et al., ; Osterhout & Holcomb, ) or grammatical but syntactically complex constructions (Osterhout et al., ; Kaan et al., ; Phillips et al., ; Gouvêa et al., ); on the other hand, there is a growing body of work on the ‘semantic P’ effect (Kim & Osterhout, ; Kuperberg, ; BornkesselSchlesewsky & Schlesewsky, ; Van de Meerendonk et al., ; Brouwer et al., ; Paczynski & Kuperberg, ; Chow & Phillips, ), showing that words that are semantically implausible within their context can also elicit a large P. Although the precise functional interpretation of the P is yet to be determined, a broad generalization that has emerged is that it reflects costs associated with a processing stage in which information from different sources is integrated into one coherent representation (Friederici & Weissenborn, ; Kuperberg, ; Bornkessel & Schlesewsky, ; Van Petten & Luka, ). Increased P amplitudes signal the detection of an integration error or integration difficulty, including costs associated with the process of reanalysis. In the particular context of NPI licensing, multiple streams of information—syntactic, semantic, and pragmatic— are recruited to construct a grammatical representation that can license NPIs. In an ungrammatical sentence that does not license NPIs, the comprehension system fails or has great difficulty integrating an NPI into the current grammatical representation, and therefore produces a large P. To our knowledge, the only two studies that did not find a P are Saddy et al. () and Yurchenko et al. (). The original data from Saddy et al. () was reanalysed in Drenhaus et al. () using a symbolic resonance analysis and a hidden P was discovered. For Yurchenko et al. (), the authors acknowledged that the lack of a P may be due to insufficient power in the data, as well as, potentially, to task-specific effects. 3
If the N and the P components index the workings of different processing mechanisms at different stages of the online comprehension process, combining the two together provides us with an opportunity to examine the finer-grained and more precise mechanisms that support the comprehension of NPIs. In particular, since it is likely that NPI licensing is not a uniform phenomenon, different licensors (or licensor-NPI pairs) may show different responses on one or both of the ERP components. Xiang et al. () examined the English NPI ever under different licensors, and argued that there are different classes of licensors, grouped by the neurophysiological patterns they elicit. Their results suggest multiple different licensing mechanisms. Four kinds of NPI licensors were tested in this study, no, few, only, and emotive factives, such as surprised, glad, etc. An example of the stimuli is given in (): ()
It is hard to train a dog. a. No dogs Andrew owns have ever responded to commands. b. Few dogs Andrew owns have ever responded to commands. c. Only dogs Andrew owns have ever responded to commands. d. Andrew is surprised that the dogs he owns have ever responded to commands. e. *The dogs Andrew owns have ever responded to commands.
ERP responses were measured from the onset of the critical word ever. The critical finding is that, when compared to the unlicensed ever in (e), all the other conditions (a)‒(d) showed a reduced N on ever, but only (a)‒(c) showed a smaller P. That is to say, in the N time window, all four licensors no, few, only, and the emotives patterned together and triggered a reduced N on ever; but in the later P time window, the emotive factive licensor deviated from other three kinds of licensors. In particular, in the P time window, the emotive factive condition elicited similar P amplitude to the ungrammatical condition (e), suggesting processing cost associated with NPIs licensed by emotive factive licensors during this time window. These findings are difficult to explain under any account that assumes a single unified licensing mechanism for different licensors, be it DE, Strawson DE, or non-veridicalitybased. Any single-route licensing account would predict no difference between the four licensors in (). Xiang et al. () interpreted their results as providing evidence for a qualitative distinction between a grammatical licensing and a pragmatic licensing mechanism, as represented by licensors no, few, and only on the one hand, and the emotive factives on the other. The relevant dimension of difference is how these licensors introduce negation into a proposition. No and few are negative determiners, and they introduce negation into the asserted meaning of a proposition. Only also introduces a negative meaning into the assertion by an exclusive component in its semantics. The example in (c) means something like ‘only Andrew’s dog but not other dogs have ever responded to commands’. On the other hand, the emotive factive predicate surprised in (d) introduces a negative implicature—that Andrew is surprised by something implicates that he did not expect it. Under this perspective, all four licensors activate some kind of negative feature. It is this pre-activated negative feature that triggered the reduced N amplitude on ever, since the lexical processing of ever is sensitive to negation. However, it is crucial to distinguish the lexical stage of processing from the full licensing of ever, since the latter involves a more combinatorial process of integrating the NPI at the sentential level. Xiang
et al. () argued that even though the comprehension system can extract the negative information from the four kinds of licensors equally quickly (as shown by the similar N effects), it nevertheless draws a line between a primary grammatical licensing mechanism and a secondary pragmatic licensing mechanism. The increased P amplitude triggered by the NPI word under an emotive factive licensor reflects the failure to license the NPI with a grammatical licensing mechanism, so that the comprehension system ultimately resorts to a pragmatic licensing mechanism, relying on the negative implicatures.
.. Illusory licensing effect To understand the precise mechanisms that license NPIs, it is also useful to look at the errors people make in comprehension, because the errors that comprehenders make when they compute the proper licensing conditions potentially provide important information about the underlying processes. It has been observed that sometimes comprehenders fall for ‘illusory licensing’ of NPIs. This often happens if a licensor is inserted into the preceding context—but crucially, it is not in the right structural (c-commanding) position (Drenhaus et al., ; Vasishth, Drenhaus, et al., ; Vasishth, Brüssow, et al., ). An example is given in () (examples taken from Drenhaus et al. ). ()
a. No man who had a beard was ever happy. *b. A man who had a beard was ever happy. *c. A man who had no beard was ever happy.
(a) is grammatical, and (b) is not, since there is no licensor in (b). But the crucial sentence is (c). In the (c) example, there is a negation in the sentence, but the NPI ever is not in the scope of the negation (i.e. the negation does not c-command the NPI). The NPI ever therefore remains unlicensed. In comprehension experiments, it was consistently observed that (c) was more acceptable than (b), and the unlicensed NPI word in (c) elicited shorter reading times or smaller ERP amplitudes than the NPI word in (b) as well (Drenhaus et al., ; Vasishth, Drenhaus, et al., ; Vasishth, Brüssow, et al., ; Xiang et al., ; Parker & Phillips, ). All of these findings suggest that the negation in (c) triggered some sort of illusory licensing or interference effect. On the surface, illusory NPI licensing looks very similar to the well-documented agreement attraction/illusion effect (Bock & Eberhard, ; Pearlmutter et al., ; Wagers et al., ) and interference in reference resolution (Badecker & Straub ; Gordon et al. ). For instance, a sentence like *the key to the cabinets are . . . , contains an agreement dependency between the singular subject the key and the plural predicate are, which is ungrammatical due to a number mismatch. However, the intervening noun cabinets creates an illusion/interference effect on various processing measures: such sentences are not uncommon in spontaneous production, they can be elicited in controlled laboratory experiments, they are judged to be relatively acceptable, and on-line reading times on the otherwise problematic verb are generally reduced compared to number mismatched verbs without interference (Bock & Miller, ; Bock & Eberhard, ; Pearlmutter et al., ; Wagers et al., ; Dillon et al., ).
The standard account of the agreement attraction effect has modelled it as an instance of memory interference. During incremental parsing of a long distance dependency, the tail of the dependency initiates the memory retrieval of the head. This retrieval is prone to interference due to overlapping features between the set of retrieval cues and other material stored in working memory (McElree et al., ; Lewis & Vasishth, ; Lewis et al., ; Wagers et al., ). Memory interference could be driven by partially matched morphosyntactic features, as has been repeatedly shown by agreement attraction errors like the example earlier, and examples like ‘The new executive who oversaw the middle managers were dishonest’ (example from Dillon et al., ). In such cases, at the plural verb, for example, were, the memory retrieval mechanism sets out to retrieve a plural subject, guided by the retrieval cues [+plural] and [+subject]. A non-subject plural noun has some probability of being misretrieved due to its partial feature match with the retrieval cues. At first glance, such a memory (mis)retrieval-based account could also be applied to NPI illusory licensing. For instance, Vasishth et al. () argued that the parser uses semantic cues such as [+negative] and syntactic cues such as [+c-command] to retrieve a proper licensor for ever from previously processed material in memory. For (b), no such match is found, and the sentence is determined to be unacceptable. For (c), however, the quantifier no in the embedded subject position partially matches the search criteria: although it doesn’t match the syntactic cue [+c-command], it does satisfy the semantic cue [+negative]. During retrieval of a licensor, this partial feature match may boost the activation level of the memory representation of the embedded quantifier no, causing it to be more likely to be retrieved once its activation level goes beyond a certain threshold. Although such an account is plausible, it seems to miss a crucial distinction between NPI licensing and other types of long distance dependencies such as subject-verb agreement. Specifically, while the latter dependency types involve mainly syntactic relations between lexical items (e.g. a subject and an agreeing verb, or a head noun and a verb in a relative clause), NPI licensing involves not only syntactic licensing (e.g. the c-command requirement on a proper licensor), but also logical-semantic and pragmatic licensing conditions. Xiang et al. () argued that the NPI interference effect stems from over-application of an inference-based licensing mechanism that is already in place in the grammar (Baker, ; Linebarger, , , Giannakidou, ; Xiang et al., ). Some evidence for this comes from a study in Xiang et al. (), which showed that NPI interference is sensitive to speakers’ individual pragmatic reasoning abilities, as measured by the Communication Subscale from the Autism-spectrum Quotient (AQ; Baron-Cohen et al. ). The AQ test has been widely used to assess individuals’ Autistic traits, which were known to affect semantic and pragmatic processing. In this study, NPI interference was compared to agreement attraction on the same group of participants. It was found that participants with worse pragmatic skills were better at rejecting the illusory NPI licensing items, and people with better pragmatic skills accepted them more often. But there was no interaction between AQ scores and the acceptability judgments on the set of number agreement items. All the studies reviewed earlier assumed that the illusory NPI licensing effect is a static phenomenon, but a set of experiments from Parker & Phillips () clearly demonstrated that illusory licensing can be switched on or off systematically. In particular, the amount of processing time the comprehender has between encountering the licensor and the NPI appears to be an important factor in modulating the illusory licensing effect. Parker & Phillips showed that the standard illusory licensing effect in (a) disappears if more material is
inserted between the licensor and the NPI. This could be done by moving the NPI to a post-verbal position (b), or embedding the NPI in a subordinate clause (c), or inserting a parenthetical clause (d). Although comprehenders accept (a) to some degree, they fully recognize that (b)‒(d) are as ungrammatical as sentences without any licensors. ()
a. The authors that no critics recommended have ever received acknowledgement. b. The authors that no critics recommended have received any acknowledgement. c. The journalists that the editors recommended for the assignment thought that the readers would ever understand the complicated situation. d. The authors that no critics recommended for the assignment have, as the editor mentioned, ever received a pay raise.
Very importantly, similar manipulations do not switch off the illusory agreement effect, as in The key to the cabinets, according to the janitor, probably were destroyed by the fire. These interesting findings provide further evidence that NPI licensing, unlike morphosyntactic agreement, involves semantic and pragmatic processing. More importantly, data like this also raises important questions about how semantic/pragmatic representations are dynamically consolidated over time in on-line comprehension. Parker & Phillips (in press) suggested that the memory encoding of the semantic/pragmatic representations may evolve over time such that the specific format of the encoding is different from the time the licensor is encountered to the time an NPI is encountered. It is this change in memory encoding that would affect the illusory licensing effect, although the details of the memory consolidation of semantic/pragmatic representations are yet to be worked out.
.. Summary The experimental studies on NPI processing reviewed in this section form the initial basis for understanding how NPIs are comprehended in real time. The emerging picture is that incremental comprehension is very sensitive to the relevant information that contributes to NPI licensing. Current findings already reveal that small changes, such as switching from one licensor to another, or changing the position of the NPI word in the sentence, can produce different processing results. To model results like these, we need to look at both NPI-specific linguistic properties, including the syntactic, semantic, and pragmatic constraints on NPI licensing, and more domain-general comprehension mechanisms, such as how linguistic representations are encoded and retrieved over time.
. A N P I
.................................................................................................................................. The complexity of NPI licensing and processing naturally raises the question of learnability. Children need to learn, from their input, the distributional constraints on NPIs. More importantly, the acquisition task is not simply for children to memorize a one-to-one relationship between a certain licensor-NPI pair. Instead, children need to learn the
relevant syntactic, semantic, and pragmatic features that allow an NPI to be licensed. As we will see later, some licensors (and possibly some NPIs) are extremely rare in the early input of a child. Given the diversity of both the set of NPIs and the set of licensors, it is unlikely that children will be exposed to a sufficient number of exemplars of every possible kind of NPI licensing in early life. The challenge for a child is to make a leap from the limited observed input to a generalized set of semantic, pragmatic, and syntactic properties that are essential to licensing NPIs.
.. Children’s knowledge about NPI licensing Studies that analysed children’s spontaneous speech from corpora have generally observed that young children make very few errors with NPI licensing. Tieu (, ) analysed the spontaneous production from forty children who speak either American English or British English, between and years old. Focusing on the production of one singular NPI any, these studies found that children produced very few unlicensed any—the error rate is only about per cent. Furthermore, even though, for the majority of the cases, children’s use of any appeared with sentential negation, they also correctly produced instances of some other licensors, such as negative quantifiers, conditionals, comparatives, etc. Lin et al. () and Lin () analysed about forty monolingual Mandarin-speaking children, between and years old. The Chinese NPI shenme has overlapping but not identical distribution with the English NPI any. The most salient difference between the two is that Chinese shenme is lexically ambiguous between an NPI and a wh-question morpheme. For instance, the wh-word what in a wh-question like ‘what did you eat?’ would be translated into Mandarin as shenme. The dual function of shenme in Mandarin makes the learning task more complicated. Lin () found that children’s early production of shenme predominantly represents its function as a question morpheme. But after years of age, children start producing shenme in various NPI-licensing contexts. Like English-speaking children, Mandarin-speaking children made very few errors—they do not use the NPI shenme in contexts that do not license NPIs. In addition to observations from children’s spontaneous production, experiments that examined children’s elicited comprehension and production also demonstrated that children have, at least some, non-trivial understanding of the licensing conditions of NPIs. A number of studies used the Truth-Value Judgment Task (TVJT) to probe children’s comprehension of NPIs (Thornton, ; Xiang et al., ; Lin, , ; Tieu & Lidz, ). It was consistently found that young children have some understanding that the NPI any should be interpreted under the scope of sentential negation. Consider the following comparison between an NPI any and a regular indefinite: ()
a. Donald and Daisy both can’t find any stars. b. Donald and Daisy both can’t find a star.
The indefinite a star can take either narrow or wide scope relative to the sentential negation. But the NPI any can only take narrow scope under the negation. Results from TVJTs (Xiang et al., ; Tieu & Lidz, ) have found that, in contexts under which the scope interpretations of any and a regular indefinite need to be distinguished, children were
very sensitive to their differences. For instance, Tieu & Lidz () set up a scenario in which the two toy characters Donald and Daisy needed to find all the hidden stars. There were different types of hidden stars, including wooden stars, metal stars, fuzzy stars, etc. They succeeded in finding all the other stars except for the fuzzy one. Given this scenario, children were asked to judge whether the sentences in () was true or false. For the most part, they correctly rejected (a), but accepted (b) (see similar findings in Xiang et al. ). Findings like these suggest that children have some sophisticated understanding of what an NPI means. At its semantic core, any in (a) is very similar to a regular indefinite, except that it introduces domain widening (Kadmon & Landman, ), which prevents any from taking wide scope over negation. Lin (; see also Lin, ) carried out an elicited imitation task to investigate children’s knowledge of a range of different licensors in Mandarin Chinese. In this study, children were given a set of pre-recorded sentences, and they were instructed to repeat, as precisely as they could, every sentence they heard from the recording. A crucial assumption of the authors is that a repetition task is not simply a task of passive verbatim recall. Instead, while repeating the stimuli, children were also reconstructing the stimuli based on their internal grammatical rules and representations. If children failed to repeat certain constructions correctly, it may indicate that the particular stimulus presented to them was not totally compatible with their current grammar. Constructions involving four different kinds of NPI licensors were tested, including sentential negation, conditionals, epistemic uncertainty modals, and polar questions. Wh-questions were also included since the Chinese NPI shenme is also a wh-interrogative morpheme. In addition, there were also ungrammatical controls in which shenme is not licensed at all. The results suggested that the first use children acquire is the wh-use of shenme. Children as young as months old were able to repeat back the wh-construction accurately. The NPI use of shenme under sentential negation and conditional and polar questions was acquired slightly later, reaching per cent accuracy by months old. Interestingly, epistemic uncertainty modals were the most difficult for children—the imitation accuracy for this construction is not reliably different from the ungrammatical controls even for the oldest children tested in this study ( months). To our knowledge, this study is one of the very few that examined and revealed different developmental paths for acquiring different NPI licensing contexts.
.. What is in the input? Our empirical understanding of children’s knowledge state regarding NPI licensing is still quite limited. There are only a handful of studies that have looked at the acquisition of NPIs. Most of these studies focused on a single NPI, or a single licensing environment like sentential negation. But even with the limited data, it is clear that children have robust knowledge of at least some aspects of NPI licensing. What is a plausible learning mechanism that could guide children to this state? A close look at the input data presented to children reveals some interesting but puzzling patterns. It is not obvious that there is sufficient evidence in the direct input that would help children make the full set of generalizations about the syntactic, semantic, and pragmatic properties of NPI licensing. Tieu (, ) searched through parental speech samples for two children (based on the CHILDES corpora) and found that the NPI any appeared in about . per cent of all
the utterances in one sample, and . per cent in the other sample. Tieu argued that the small number of tokens may not in itself cause learning problems, since the distribution of any in the parental speech did provide some positive evidence for children to make hypotheses about the syntactic or even semantic properties of any and its various licensors. For instance, any predominately appeared in the scope of negation in the parental speech, and some instances of other DE licensors were also observed. Lin (, ) found that, in a sample of utterances of child-directed speech that contain the Mandarin NPI shenme, (.%) of them are wh-questions, and the remaining handful of examples are almost evenly distributed among different licensing environments for the NPI shenme (see appendix A in Lin, ). Such imbalanced distribution in the input does not completely predict children’s performance in the elicited imitation task discussed earlier. On the one hand, wh-questions are indeed the first thing children understand about shenme, which is predicted given how frequent wh-questions are in the input; but at the same time children nevertheless showed some knowledge about the NPI any in the elicited imitation task, despite the extremely sparse distribution of NPI shenme in the input. Moreover, frequency in the input data alone could not explain why epistemic uncertainty modals are particularly difficult for children to acquire as NPI licensors, since epistemic modals do not seem to be less frequent than other licensors. The distributional patterns of different licensing environments, although they may be imperfect, at least provide some positive evidence for children to start bootstrapping from the restricted distribution of NPIs. But as Tieu () critically pointed out, the observable distributional patterns of NPIs do not automatically ensure that the evidence for acquiring the semantic representations of NPIs is immediately available from the input. For instance, the scalar use of English any introduces alternatives and widens the domain of the indefinite (as discussed in Kadmon & Landman, ; Chierchia, , ), achieving a similar effect to focus. Tieu noted that the domain widening effect of any, however, is not immediately transparent by just observing that any co-occurs with negation or other DE operators. Additional contextual support is necessary to signal the domain widening effect. Tieu () reported instances of any through a search of twenty transcripts of conversations between parents and children, and found only two cases that presented clear contextual cues of domain widening or focus. Given the limited input, it remains a puzzle what features of the input provided the most informative cues to facilitate children’s development of the semantic and pragmatic representations.
.. Summary The major research question on the acquisition of NPIs is about understanding how children acquire the full complexity of NPI licensing. Empirically speaking, we have very limited data to support any robust conclusions. But it already appears that children have some non-trivial understanding of NPIs at an early age. At the same time, it also appears that not all NPIs and licensors are acquired equally well or equally early. Current discussion of the input has mostly focused on finding out the distributional frequency patterns of NPIs. Given that distributional patterns alone do not seem to provide an adequate account of the developmental path, it is possible that there are other types of subtle cues in the input that correlate with learning, which the current research has yet to fully identify.
. C
.................................................................................................................................. The linguistic complexity of NPI licensing makes it a rich empirical domain for investigating the cognitive architecture of language processing and acquisition. An adequate processing model needs to explain how information from different sources, syntactic, semantic, and pragmatic, are incrementally combined, stored, accessed, and continuously consolidated over time to form an appropriate licensing context for NPIs. At the same time, what adults can do as language users has to be learned with a specific learning mechanism that can transform information from the external input to an internal grammar. Previous studies, as we discussed in this overview, have revealed exciting new empirical findings, but at the same time have also raised more questions. For future research, we suggest two possible directions that are worth exploration. First, more varieties of NPIs and different licensing contexts should be experimentally investigated. Previous studies have largely focused on negation as the licensing context, and examined only the two NPIs ever and any. The empirical horizon needs to be widened to include other kinds of licensors and NPIs. Cross-linguistic investigation would be especially valuable since the distribution of NPIs and likely the precise licensing mechanisms vary from language to language. Second, the investigation of NPI licensing should engage with other research domains. For instance, downward entailment inferences are argued to be important for NPI licensing. But very little has been explored as to how people process logical inferences like downward entailment, and to what extent people’s ability to compute different downward entailment relations carry over to NPI processing. Similarly, for a subset of the NPIs, the scalar NPIs such as any, the proposed pragmatic strengthening function bears similarities to contrastive focus marking. Independent findings from research on focus processing/acquisition could shed light on questions such as what factors modulate the set of alternatives activated by any, or what cues in the discourse are salient to signal the presence of alternatives. It is possible that these related processes are supported by shared cognitive mechanisms.
......................................................................................................................
......................................................................................................................
W do pronouns matter to the study of semantics and pragmatics? A long-standing answer to that question has been that pronouns are the fruit flies of discourse—that their interpretation reflects the deeper semantic meaning of a passage and can thus provide a window into the forces at play in establishing meaning across clauses. This view has motivated much of the experimental work on pronoun interpretation and production. What that work has revealed, however, is that the story is in fact more complex. Pronouns are not transparent windows into underlying discourse structure and meaning; rather the pronominal form itself imposes unique and important pragmatic constraints on what speakers and comprehenders do with pronouns. These two viewpoints on the role of pronouns—as fruit flies to deeper meaning and as linguistic elements with independent effects—are illustrated later with two well-known examples from the pronoun literature. Through these examples and a discussion of a number of experimental studies, this chapter presents a history of the pronoun puzzle. It then introduces a recent attempt to reconcile competing approaches by situating pronouns within a more general model of pragmatic communication. To start, consider examples (a‒b). These examples shows how ambiguous pronouns receive their interpretation via the general reasoning that comprehenders use in order to make a discourse make sense. ()
Effects of general reasoning in pronoun interpretation (from Winograd, a) a. The city council denied the demonstrators a permit because they feared violence. b. The city council denied the demonstrators a permit because they advocated violence.
A comprehender who encounters (a) or (b) must figure out how the speaker intended the clauses to relate. For (a), doing so requires reasoning about how the denial of a permit could be linked to fear and identifying which event participants’ fear could most plausibly account for that denial. The establishment of a coherent discourse thus yields an interpretation of the pronoun they as referring to the city council. A similar process for (b) yields a different interpretation: It is not clear how a preference for violence from the city council would result
in the demonstrators being denied a permit, but if the pronoun is taken to refer to the demonstrators, then it is possible to reason how demonstrators who advocate violence could be worrisome to the council and how that worry could lead to the denial of a permit. In these cases, it seems that the pronoun itself dictates little; it is the reasoning in pursuit of coherence that determines the pronoun’s referent. An approach which takes into account the role of coherence thus assigns pronouns an epiphenomenal status. Their meaning is cast simply as a side effect of the comprehender’s process of establishing a coherent discourse (Hobbs, ; Hobbs et al., ; Kehler, ; Rohde et al., ). But that approach is not the only one, and in the end, it leaves unexplained certain interesting facts about pronoun use. If pronouns are simply unbound variables whose meaning is assigned in the course of general discourse processing (Hobbs, , among others), one might expect processing a name to be simpler than processing a pronoun. After all, a name can pick out a unique referent whereas a pronoun may require additional reasoning to resolve. On the other hand, if the pronominal form itself dictates the contexts in which it is more or less appropriate, using a name in such a context could induce a penalty. Indeed, models like Centering Theory (Grosz et al., ) encode a preference for speakers to use pronouns to refer back to the most prominent referent in the discourse, where prominence is linked to a referent’s syntactic position in the previous utterance. The two passages in (a‒b) represent materials from a reading time study comparing the processing of pronouns and names when the antecedent is the previous syntactic subject (Gordon et al., ). ()
Comparison of pronouns and names: the Repeated Name Penalty (Gordon et al., ) a. Pronoun-Name Condition Bruno was the bully of the neighbourhood. He chased Tommy all the way home from school one day. He watched Tommy hide behind a big tree and start to cry. He yelled at Tommy so loudly that all the neighbours came outside. b. Name-Name Condition Bruno was the bully of the neighbourhood. Bruno chased Tommy all the way home from school one day. Bruno watched Tommy hide behind a big tree and start to cry. Bruno yelled at Tommy so loudly that all the neighbours came outside.
The results of the study showed that pronouns do not necessarily give rise to comprehension difficulty; rather, the passage containing a number of ambiguous pronouns (a) was read faster than the version with only unambiguous names (b). This suggests that there may be contexts in which pronouns are expected over names and in which comprehension difficulty emerges not because of the content of the passage’s message, but the form that that message takes. An approach which takes into account a passage’s surface form (its structure and word choices) thus is concerned with a speaker’s decision of how to realize an intended message. Reconciling the facts illustrated in ()–() requires understanding when and why pronouns are used in discourse, questions that have interested researchers from across linguistics, philosophy, psychology, and computer science. The facts are further complicated by a
number of syntactic, semantic, and pragmatic biases that have been identified in experimental studies on pronoun use. This chapter reviews that work and the questions raised. It then steps back to ask whether it’s really possible to understand pronouns by looking at pronouns themselves. A new approach is laid out that shifts perspective from pronouns specifically to a generative model of language more broadly. Such a model captures the choices speakers make in choosing messages to convey and choosing words with which to convey those messages. In the case of pronouns, a generative model makes an important distinction between the choice of which referent a speaker will mention and the form that that reference may take, thereby providing an account of the patterns observed in ()‒(). In this way, pronouns are still fruit flies, but they serve a new purpose of providing a test case for how generative models can inform research in pragmatics.
. O , ,
.................................................................................................................................. The history of research on pronouns overwhelmingly targets one particular type—the third person singular pronoun—almost to the exclusion of other pronominal forms.1 Despite this focus, the resulting picture has not been simple. Many studies have grappled with the question of what a comprehender does when she encounters a pronoun, and a number of different factors have been identified, some of which appear to contradict each other. More recent work manages to resolve some of these discrepancies by showing how other mechanisms active in a discourse may determine the role a given factor plays in a particular context. These puzzles, and hints at directions for reconcilation, are reviewed in the following subsections.
.. Does syntax matter? To start, an oft-cited constraint on pronoun interpretation involves surface syntactic structure: this is a preference for antecedents that have appeared in position (Frederiksen, ; Gernsbacher & Hargreaves, ; Kameyama, ; Arnold, ; Arnold & Lao, ; Hartshorne, Nappa, & Snedeker, ; see also the interpretation algorithm for Centering Theory, Brennan et al., ). For example, the ambiguous pronoun him in (a) is said to resolve preferentially to John, the subject of the context sentence about the hitting event. However, if that same event is described in a passive construction with Bill promoted to subject position, as in (b), the preference is said to switch to Bill.
1
The focus in this chapter will be further restricted to work conducted on English third person singular pronouns. Work on other pronominal forms (null pronouns, demonstratives) has been pursued for other languages with different pronominal systems. Indeed, pronouns can be used in varied and surprising ways (e.g. they in contexts without a readily available plural antecedent, abstract event-referring this, etc.), but the third person pronouns he/she will be sufficient to get us started. In section . we return briefly to these issues.
()
Subjecthood preference (examples from Kameyama, ) a. John hit Bill. Mary told him to go home. [him=John] b. Bill was hit by John. Mary told him to go home. [him=Bill]
Intuitions about examples like (a‒b) suggest a role for syntax in antecedent selection, as does experimental work (Gernsbacher, ; Crawley et al., ; Speelman & Kirsner, ; Arnold, ) and corpus work (Arnold, ) that shows preferential resolution to the first-mentioned or subject referent over other referents. The subject preference has been assessed with a variety of procedures: for example, reading time and timed/untimed antecedent selection tasks. Another way that syntax has been implicated in pronoun interpretation is via a preference for antecedents in syntactically parallel positions. Antecedent selection tasks and reading time tasks (Sheldon, ; Smyth, ; Chambers & Smyth, ) show a preference for the subject referent in passages like (a) and the object referent in passages like (b). ()
Parallelism preference a. John kicked Bill in the leg. He punched Mary in the arm. [he=subject John] b. John kicked Bill in the leg. Mary punched him in the arm. [him=object Bill]
Already we have two factors (subjecthood, parallelism) that can conflict: in passages like (b), what dictates whether the pronoun him will be resolved to the subject antecedent (John) or the parallel antecedent (Bill)? One proposal is that discourses in which these two preferences manifest differ in the inference processes underlying coherence establishment (Kehler, ). What is notable about passages like (a‒b) and (a‒b) is the way the sentences within them relate to each other. The examples used to illustrate a subject preference describe narrative sequences of events (compatible with the connective then), whereas the examples in support of a parallelism preference both convey parallel events (compatible with connectives like similarly or likewise). In answer to the question Does syntax matter?, one can see that the appearance of syntax-based preferences may depend on the way that the sentences combine to form a coherent discourse. Narration and Parallel relations represent just two out of a larger inventory of possible relations (‘coherence relations’) that can be inferred to hold between clauses (Hobbs, ; Mann & Thompson, ; Sanders et al., ; Kehler, ; Asher & Lascarides, ; Prasad et al., ). The pronoun literature that targets surface order and syntactic constraints (Crawley et al., ; Smyth, ) concedes a role for real-world knowledge and plausibility. However, explicitly testing such a role requires a different tactic of direct assessments of semantic biases and coherence relations (see Stevenson et al., ; Stewart et al., ; Crinean & Garnham, ; Rohde & Kehler, ). To understand the role of general reasoning in pronoun interpretation, we turn next to a particular type of semantic bias. This will connect up with the generative model alluded to at the outset of this chapter.
.. Does thematic role matter? In contrast to posited surface structural constraints, biases associated with the lexical semantics of particular verbs have been shown to cut across syntactic categories. Examples
(a‒b) contain verbs belonging to the class of so-called Implicit Causality (IC) verbs (Garvey & Caramazza, , among others). ()
IC bias to causally implicated referent (examples from Caramazza et al., ) a. Mary annoyed Sue because she had stolen a tennis racket. [she=Mary] b. Mary scolded Sue because she had stolen a tennis racket. [she=Sue]
(a‒b) are minimal pairs, differing only in the first verb. The two referents Mary and Sue appear as subject and object, respectively, in both examples. However, the pronoun she in (a) is interpreted most plausibly to refer to Mary, in (b) to Sue. This difference is attributed to the way annoy and scold assign their thematic roles: Mary as the subject of annoy in (a) is the causally implicated stimulus referent and Sue in object position is the experiencer, whereas in (b), Mary is the agent and it is Sue who is the causally implicated referent. Verbs like these have been shown to induce a pronoun interpretation preference to the causally implicated referent, independent of the syntactic position of that referent. This has been demonstrated using a number of methodologies (for a review, see Koornneef & Sanders, ). For example, in story continuation tasks in which participants are prompted with a passage up to and including the pronoun, the participants show a preference to use the pronoun to refer to the causally implicated referent (e.g. Garvey & Caramazza, ; Stevenson et al., ; Stewart et al., ; Ferstl et al., ; Hartshorne & Snedeker, ). On-line measures show a corresponding pattern. Participants read more quickly when the causally implicated referent is mentioned than an alternate referent (e.g. via a gendered pronoun in a context with one male and one female referent; Koornneef & Van Berkum, ); in visual-world eyetracking paradigms, participants look anticipatorily to a picture of the causally implicated referent even before hearing a pronoun (Pyykkönen & Järvikivi, ). Thematic role effects can also be seen in transfer-of-possession contexts in which the source and goal thematic roles occupy the subject or non-subject position. Examples (a‒b) show how such sentences have been presented in story-continuation tasks (Stevenson et al., ; Arnold, ; Kehler et al., ). In continuations following both (a‒b), participants use the pronoun to refer to the goal referent, Bob, as often or more often than the source referent, John. ()
Transfer-of-possession contexts with goal bias a. John handed a book to Bob. He _______ b. Bob received a book from John. He _______
However, the bias to the causally implicated referent for IC contexts and the bias to the goal for transfer-of-possession contexts are not uniform. When pronoun interpretation patterns are broken down by coherence relation, either by explicitly indicating the relation with a connective (Stevenson et al., ; Koornneef & Van Berkum, ) or by inferring the relation via annotation of participants’ continuations (Arnold, ; Kehler et al., ), sub-patterns emerge. The bias to the causally implicated referent is strongest in continuations that describe, unsurprisingly, a cause (Explanation relations); the bias to the goal referent is strongest in relations that describe what happened next (Narration or Result relations), presumably due to the goal’s association with the end state of a transfer event. The interpretation of the pronoun is thus conditioned on the operative coherence relation. This can be seen in examples (a‒b) and (a‒b).
()
Effect of coherence relation in IC contexts a. John amazed Bill. He could walk on his hands. [He=Johnstimulus] b. John amazed Bill. He clapped his hands in awe. [He=Billexperiencer]
()
Effect of coherence relation in transfer-of-possession contexts a. John handed a book to Bill. He opened it to the first page. [He=Billgoal] b. John handed a book to Bill. He handed him War and Peace. [He=Johnsource]
The first example in each pair shows coreference in accordance with the oft-reported thematic role biases: the causally implicated referent for Explanations following IC verbs in (a), the goal for Narration relations following transfer-of-possession verbs in (a). The second example illustrates alternative coreference biases associated with different coherence relations: the experiencer for a Result relation in (b) (see ‘implicit consequentiality’, Stewart et al., ; Crinean & Garnham, ; Pickering & Majid, ), the source for an Elaboration relation in (b). Early reports of thematic role biases were assessed by collapsing across coherence relations. The picture that emerges when coherence is taken into account is that the purported bias to the causally implicated referent arises from the bias to that referent in Explanations and, crucially, the frequency of Explanation relations in IC contexts; similarly, the purported bias to the goal arises from the frequency of Narration/Result relations in transfer-of-possession contexts (Kehler et al., ). What do these observations about thematic role biases and coherence relations mean for a model of pronoun interpretation? If we think of a comprehender’s task as one of reverseengineering the speaker’s intended message, then verb class can provide a cue to the message the speaker may be trying to convey. For example, if a speaker describes a situation using an IC verb, the comprehender may expect that the speaker will use the next utterance to provide an explanation and, furthermore, that the causally implicated referent will be mentioned as part of that explanation. In this way, a comprehender can attempt to reconstruct the speaker’s intended meaning by considering how the discourse could most plausibly have been generated—which coherence relation is most likely to hold following an IC verb and which referent is most likely to be mentioned given that relation? We return to generative models in section .. See also Degen & Tanenhaus (Chapter in this volume) on the role of prediction in constraint-based pragmatic processing, specifically the relevance of discourse structure, world knowledge, and estimates of the speaker’s epistemic state to inform priors over possible messages. Having identified a role for discourse-level coherence relations in pronoun interpretation, the next two sections consider other ways in which discourse structure and information structure guide pronoun use.
.. Does distance matter? Beyond the short passages discussed thus far, there are claims regarding the interpretation of pronouns in larger discourses—for example, the claim that pronouns favour antecedents that were mentioned . Intuitively, a referent introduced in a previous chapter of this handbook is unlikely to be re-mentioned with a pronoun in the sentence you’re currently reading. Indeed, pronouns have been found to yield more processing difficulty if the antecedent is mentioned two or three sentences back than if the antecedent appears in
the immediately preceding sentence, as assessed via full-sentence reading times, response times for antecedent selection, and reaction times for judging the familiarity of a probe word related to a potential antecedent (Clark & Sengul, ; see also Chang, ; Ehrlich & Rayner, ; Gernsbacher, ; Arnold, ). But how is recency determined? The distance to some referents may be large if one counts words or clauses, but small if the metric reflects the hierarchical combination of sentences in a discourse. Earlier work on discourse parsing suggested that the availability of particular pronoun∼antecedent relationships was constrained by the possible attachment positions of subsequent clauses in an unfolding discourse. The Right Frontier Constraint limits the addition of new clauses to open nodes on the right edge of the discourse structure (Polanyi, ; see also Malt, ; Asher, ; Asher & Lascarides, ). Figure . illustrates the right frontier of a discourse, representing the sentences of the discourse as nodes in a tree with semantic links shown as branches (e.g. the coherence relation expressed by because in () would link a parent and daughter node in a tree like Figure .). When asked to select an antecedent for a pronoun, comprehenders appear sensitive to the role of structure, favouring antecedents in open positions over closed positions, even when the linear distance is greater (Holler & Irmen, ). To illustrate the role of the right frontier in pronoun resolution, consider the pronoun it in the last utterance of the dialogue excerpt in (). Despite a half dozen intervening utterances and the presence of several other available referents, it in (i) manages to pick out an antecedent (the stuck bolt) mentioned only at the beginning of the dialogue. In an analysis of this dialogue and the broader role of intentional structure in discourse, Roberts () identifies an overarching goal shared by the dialogue participants (to disassemble the engine) and a sub-goal introduced at (a) (to remove the stuck bolt). The subsequent utterances (b‒h) address the execution of plans and sub-plans for achieving the sub-goal. In this way, the utterance at (a) about the problem (One bolt is stuck) is an open node in the discourse structure, making it a possible attachment point for the utterance at (i) about the problem’s resolution (I already got it loosened). The pronoun in (i) can thus be resolved to boltv as a structurally accessible referent on the right frontier. ()
Naturally occurring task-oriented dialogue showing effect of discourse structure on pronoun interpretation (Grosz, ). Discourse participants are an Expert (E) and an Apprentice (A) who are working together remotely to disassemble an engine. A camera feed shows the Expert the Apprentice’s work space. a. A: One boltv is stuck. b. A: I’m trying to use both the pliersy and the wrenchz to get itv unstuck, but I haven’t had much luck. c. E: Don’t use the pliersy. d. E: Show me what you’re doing. e. A: I’m pointing at the boltsx. f. E: Show me the 12” combination wrenchw, please. g. A: OK. h. E: Good, now show me the 12” box wrenchu. i. A: I already got itv loosened.
Open
closed
closed
closed
closed
closed
closed closed
open
closed closed
open closed
open
.. A sample discourse structure containing open positions to which subsequent clauses may attach (from Polanyi, : ; reprinted with permission from Elsevier); pronouns favour antecedents in open positions
.. Does topichood matter? Example () demonstrates how recency can depend on the model of the discourse. In a discourse structure, each sentence is said to stand in a coherence relation with prior material (Hobbs, ; Mann & Thompson, ; Sanders et al., ; Kehler, ; Asher & Lascarides, ; Prasad et al., ) or to address a question or goal relevant to the preceding discourse (Roberts, ). Inferring discourse goals involves inferring what is at issue, or what is topical. One way of understanding the use of the pronoun it in the dialogue in () is to consider the information status of its antecedent, boltv, in the dialogue. In information structural terms, boltv is the topic—it is what the discourse is about. In the context of a dialogue, topichood can be determined by the goals and intentions of the interlocutors. Information status can also be signalled more locally via sentence structure. The subject preference illustrated in examples (a‒b) and the Repeated Name Penalty in (a‒b) may reflect the information status of syntactic subjects. In English, subject position is the default position for topics (Chafe, ; Zubin, ; Reinhart, ). Moreover, being the subject of a passive (as in (b) Bill was hit by John) is an even stronger signal that that referent is likely to be the topic (Prince, ; Shibatani, ; Ward & Birner, ). Since speakers are known to use pronouns when they maintain the current topic, it seems reasonable that comprehenders ought to take topichood into account when they encounter a pronoun. The field has long noted this link between topicality and the use of underspecified referential forms like pronouns, with researchers assessing a referent’s status via discourse structure and sentence structure as well as measures of cognitive accessibility (Sanford & Garrod, ; Givón, ; Prince, ; Ariel, ; Gernsbacher, ; Gundel et al., ; Walker et al., ; Grosz et al., ; Cowles & Ferreira, ; Arnold & Lao, ). The importance of a referent’s status in a passage has been confirmed in a range of studies on pronoun interpretation and production. In reading time studies, sentences with pronouns are found to be processed faster when the pronoun refers to a topical antecedent than a nontopical antecedent (as controlled by primacy of mention and scenario dependence; Anderson et al., ; see also Garrod et al., ) or when the antecedent is introduced with a name
instead of a definite noun phrase (perhaps because names are used for principal protagonists; Sanford et al., ). In a study of children’s storytelling, more sophisticated speakers were found to use pronouns to refer to the protagonist (Karmiloff-Smith, ). In adult productions, again the pronoun is used to refer to the main actor (Marslen-Wilson et al., ). Lastly, to see the effects of passivization, consider the examples in (a‒b). These represent stimuli for a set of studies using familiarity with a probe word as a measure of antecedent accessibility (Speelman & Kirsner, ). ()
Effect of passivization on topicality (examples from Speelman & Kirsner, ) a. A nurse threw the shoe over the balcony. b. The shoe was thrown by a nurse over the balcony.
Participants in the study read a sentence like (a) or (b) and then saw the beginning of a subsequent sentence: either a pronoun She or another noun phrase A neighbor. The task required them to indicate at that point whether a probe word was familiar or not. The findings showed that the probe word nurse was recognized more quickly following the pronoun She than A neighbor, but that this pattern only held for active voice stimuli like that in (a), not for the passive voice (b) when the nurse is no longer in a topical position. For the Repeated Name Penalty and the passivization effects, it is only the sentence’s surface structure and form that changes, not the event-level semantics. But in many of the results discussed earlier, surface structure provided no cue to the intended interpretation, which depended instead on semantic biases or general reasoning. The next section summarizes this distinction between models built around meaning and models built around form.
.. Prior models of pronoun use On the one hand, a coherence-driven model of pronouns (Hobbs, ; Kehler, ) highlights the importance of comprehenders’ reasoning about the way that sentences relate. Such a model, however, provides no mechanism to account for behaviour that varies with surface form—for example, alternations that arise with different syntactic structures or different anaphoric forms. This modelling bias may reflect the artificial intelligence work of the time. Computational systems were being built that could reason about the logical form of an utterance and draw conclusions with respect to a knowledge base of axioms about the world (e.g. Hobbs et al., ). Sentence form was treated as a separate postscript link between the underlying meaning and the surface realization. On the other hand, models that take into account the surface forms of the utterances in which a pronoun appears (Grosz et al., ) make no effort to incorporate deep semantic reasoning, yet they capture patterns that cannot be ignored—for example, a link between subject antecedents and pronominal forms. The development of Centering Theory reflects the pressure in computational linguistics to build systems around directly observable features—the theory’s definitions and rules make use of grammatical role and referential form, properties of a sentence that are easily discernable from surface input and don’t require deeper reasoning. Centering Theory can make predictions across a wide variety of discourse contexts precisely because it doesn’t depend on a sophisticated engine for deriving domain-specific meaning. Its effectiveness likely reflects the fact that grammatical role and referential form are symptoms of deeper information structural factors like topichood.
The results of the experimental work reviewed earlier can be linked to each of these two approaches. The coherence-driven approach accounts for evidence of thematic role biases and some of the interpretations of the syntactic and distance-based metrics. The formdriven approach accounts for topic-driven effects like the subjecthood preference and the Repeated Name Penalty. It is only recent work that has proposed a reconciliation of the two approaches (Kehler & Rohde, ). This new work, to be described in what follows, makes an important distinction between factors that influence which referent is the most likely candidate for re-mention and factors that influence how that referent will be mentioned. A need for such a distinction is apparent in several previous studies. For example, Hudson-D’Zmura & Tanenhaus () analyse IC contexts, specifically those in which the causally implicated referent is mentioned in object position. In a pair of studies, one set of participants wrote story continuations (e.g. Max despises Ross because he ______), and another set of participants made speeded sensibility judgements about passages containing either a subject-referring or object-referring pronoun (e.g. Max despises Ross. He always gives Max/Ross a hard time.). The results showed an interesting dissociation between the referents favoured for mention in story continuations (the causally implicated object) and those whose mention with a pronoun yielded the highest sensibility judgements and shortest decision latencies (the subject). At the level of meaning, participants’ story continuations showed that they were attending to the lexical semantics of the verb and the coherence relation signalled by the connective because. At the level of surface form, participants’ speeded ratings showed they were sensitive to the subjecthood of the antecedent. An adequate model of pronoun use must therefore capture both the meaning-driven and form-driven patterns. We need to understand which referents are likely to be mentioned and which referents are likely to be mentioned with a pronoun, since those probabilities need not be the same. Such a model requires recasting the research question from a pronoun-resolution-specific question (‘Given a pronoun, who does it refer to?’) to one that models coreference more generally (‘Given estimates of likely messages and message forms, how does a comprehender infer what the speaker intended?’).
. S
.................................................................................................................................. A range of different methodologies have been used to understand pronoun interpretation, including reading time (e.g. Wolf et al., ), probe-word response (e.g. Greene et al., ), speeded sensibility judgements (e.g. Foraker & McElree, ), brain response (e.g. Van Berkum et al., ), and eyetracking in the visual world paradigm (e.g. Arnold et al., ). This work has tried to understand pronoun interpretation largely by looking at how people interpret pronouns. But therein may lie the mistake. If a pronoun leads to reading time slowdowns, longer response latencies, neuronal activity associated with ambiguity, and delayed looks to a referent in a visual scene, how do we know if that response indicates difficulty with who has been mentioned or with how? We need to include in our analysis contexts in which the form of reference is not a pronoun (see Almor & Nair, ). Consider the following excerpts from two different speakers’ narrations of the same sequence in a video recording of a basketball game (Brennan, ).
()
Choices in referential form (Brennan, ) a. Wolverines movin’ it slowly up the back court, over to number . . . back to forty-one, he shoots from three points . . . no. b. Number thirty passes it off to forty-one. Forty-one goes up for the shot, and he misses.
The nature of the events being described (a pass from one player to another followed by an attempted shot) dictate which referent is most likely to be mentioned after the pass. In both speakers’ versions, player forty-one is clearly in possession of the ball after the pass. However, despite the high probability that the speaker will generate a statement about player forty-one, the first speaker generates a pronominal reference and the second does not. This example distills the distinction that a generative model must make between the probability of who to mention and how, and that those two probabilities need not be the same. Across (a‒b), the observed probability of starting the last utterance with a mention of forty-one is but the probability of pronominalizing that reference is only .. A comprehender encountering the pronoun in the last sentence of (a) would likely have no trouble resolving it to player forty-one since the event semantics guarantees that the person shooting the ball must be the one who received it on the previous pass. So how do we reconcile the fact that a comprehender would easily interpret a pronoun in this context to player forty-one with the fact that a speaker wouldn’t necessarily produce a pronoun for that referent? Why aren’t production and interpretation mirror images of each other? The literature often assumes a direct mapping between production and interpretation. This is apparent in claims that speakers use pronouns to denote referents they believe are associated with a high degree of activation in the cognitive state of the comprehender (e.g. Gundel et al., ; see discussions of salience in Bach, ; Von Heusinger, ). The assumption is typically that a speaker is permitted to use a pronoun to refer to the single most prominent referent precisely because a comprehender will interpret that pronoun to refer to the single most prominent referent.2 Under such an assumption, the remaining task for experimental research is to identify the factors that determine referent prominence. That research has yielded the set of factors reviewed in section .. But examples like (b), in which the speaker uses a non-pronominal form to mention a highly salient referent, are a challenge to the mirror image assumption. A further challenge comes from story continuation data like the following from Stevenson et al. (). ()
2
Story continuation results with and without pronoun prompt (Stevenson et al., ) a. Mary annoyed Sue. ______ [63% Mary] b. Mary scolded Sue. ______ [72% Sue] c. Mary annoyed Sue. She ______ [78% Mary] d. Mary scolded Sue. She ______ [66% Sue]
Other terms for the most prominent referent include salient, accessible, activated, in focus, the centre of attention, etc. All are associated with the same assumption that pronoun production and interpretation are mirror images of each other.
Recall that annoy and scold are classified as IC verbs, with annoy being subject-biased and scold being object-biased in that participants’ continuations favour the causally implicated subject referent Mary for (a) and the causally implicated object referent Sue for (b). What Stevenson et al. found was that these preferences shifted in a condition with a pronoun prompt. The pronoun is fully ambiguous, but its presence served to increase the number of continuations about the subject (meaning more about the causally implicated referent for (c), fewer for (d)). To understand this, we need a model that positions pronouns within the space of possible anaphoric forms a speaker might produce. In section ., generative models are introduced as a framework that can provide the insight we need for a better understanding of pronouns (as argued in section .).
. L
.................................................................................................................................. This section is your statistical interlude to introduce generative models. Generative models provide a way to understand the output of a linguistic system by considering what forces generated that output. We’ll start with an example. Imagine we’re looking ahead to US election night in November . The results are coming in and a speaker—whose political party you know—utters the following potentially ambiguous sentence: ()
The awful one has won!
In order to understand the speaker’s intended message—that is, whether the winner is Democratic candidate Hillary Clinton or Republican candidate Donald Trump—one needs to consider both the probabilities of the possible messages and the probabilities with which this speaker would have generated different utterance forms to convey each message. What’s of interest is how the prior probability of a given message combines with the appropriateness of a particular message form in this context (i.e. when produced by this speaker). Let’s say that the probabilities that the different messages would be generated are determined by real-world knowledge, independent of the speaker, as in (). ()
Probabilities over possible speaker messages3 p(Clinton won) = . p(Trump won) = .
Next, let’s say that the probability the message is generated using the form in () varies with the message and what is known about the discourse context of the utterance, namely the political party of the speaker, as in (). Note that the values associated with these forms are invented and necessarily sidestep the issue of how many possible forms there are that a speaker might produce; here I treat the set of available forms as small and closed, with the Election outcomes reflect the June estimates from http://projects.fivethirtyeight.com/ -election-forecast/. 3
expression ‘The awful one has won’ being one of the only possible epithets. The conditional probability is written p(outcome | condition). ()
Probabilities over possible forms, conditioned on message, with the assumption that the speaker is a Democrat p(‘The awful one has won’ | Clinton won) = . p(‘The awful one has won’ | Trump won) = .
Here, the speaker who is a Democrat is very likely (probability estimate .) to generate a disparaging description of Trump, but since many Democrats are also wary of Clinton, knowing the speaker is a Democrat doesn’t preclude the possibility that a negative description of Clinton will be generated (hence probability estimate .). Note that the probabilities in () do not sum to . This is expected. What must sum to are the probabilities for all the forms that could be generated to convey a particular meaning. For example, utterances that described Trump with kinder words are assumed to be very unlikely for a Democrat (the remaining probability .). The comprehender’s job is to recover a likely meaning given the form of the utterance. That job is encapsulated as the conditional probability of a message given a particular form: p(message | form). Bayes’ Rule, if you’re not familiar with it, can be used to reframe that probability in terms of two probabilities about what the speaker might have done: the prior probability of the intended message and the probability that that message would be conveyed with that particular form. See (). Bayes’ Rule is a general theorem of probability theory that can be used to describe any belief updating (here, about a message) in light of new evidence (here, the utterance form).4 () Bayes’ Rule p(message | form) ∼ p(message) p(form | message) Now let’s apply Bayes’ Rule to the interpretation of the speaker’s utterance on election night, as in (a). Bayes’ Rule states that the probability that the speaker’s intended message is Clinton won, given the words ‘The awful one has won’, is proportional to the prior probability that this speaker would generate such a message (is Clinton won a likely message a priori?) combined with the likelihood that such a message would be conveyed in this context with those words (would this speaker, a known Democrat, produce a negative description of Clinton?). Formula (b) shows a similar calculation for the alternative intended message that Trump won. () a. Democratic speaker intends to convey the message that Clinton won p(Clinton won | ‘The awful one has won’) ∼ p(Clinton won) p(‘The awful one has won’ | Clinton won) ∼ . . = .
The symbol ∼ indicates that the probability on the left side of the formula is proportional to the value computed on the right side. For the exact value, the right side would include a denominator summing over the numerator for all messages. 4
b. Democratic speaker intends to convey the message that Trump won p(Trump won | ‘The awful one has won’) ∼ p(Trump won) p(‘The awful one has won’ | Trump won) ∼ . . = . It’s a toss up: the probability that the sentence in () would be interpreted to mean Clinton won is roughly equal to the probability that it means Trump won. The sentence is confusing when uttered by a Democratic speaker precisely because neither message has a clear advantage given the priors over likely election outcomes and the production probabilities for a Democratic speaker. Now consider what message would be inferred if that same utterance were generated by a Republican. The probabilities of the election outcomes remain the same—assuming that the Republican speaker is following the same polls and election predictions. But the production probabilities for that particular utterance form are different in this case. Here I estimate that the Republican speaker is highly likely (.) to use negative words to describe Clinton, but there may be Republicans who also are wary of Trump (and perhaps more than there are Clinton-wary Democrats, hence an estimate of .). () a. Republican speaker intends to convey that Clinton won p(Clinton won | ‘The awful one has won’) ∼ p(Clinton won) p(‘The awful one has won’ | Clinton won) ∼ . . = . b. Democratic speaker intends to convey that Trump won p(Trump won | ‘The awful one has won’) ∼ p(Trump won) p(‘The awful one has won’ | Trump won) ∼ . . = . In this case, with a Republican speaker, the intended message is much more obvious. The scenario in which the speaker intended to convey Clinton won receives a much higher probability than the one in which he intended to convey Trump won. The example here shows how real-world knowledge can guide the probabilities of different intended messages and how the appropriateness of particular forms varies by context (i.e. depending on the political affiliation of the speaker). In the next section we apply Bayes’ Rule to model pronoun use, for which it can likewise be used to combine biases about intended messages with estimates of the appropriateness of particular forms in a particular discourse context.
. P : A
.................................................................................................................................. Just as a speaker’s ambiguous utterance about the winner of an election can be decoded by modelling the process by which the utterance was generated, an ambiguous pronoun can
likewise be decoded by considering how it was generated. Bayes’ Rule in () describes how to update one’s belief that a particular referent has been mentioned, given that a pronoun has been encountered, (p(referent | pronoun)). It does so by combining the prior belief that that referent would be mentioned, p(referent), with the likelihood that that referent would be mentioned with a pronoun, p(pronoun | referent). ()
Bayes’ Rule for pronouns (as proposed in Kehler et al., ) p(referent | pronoun) ∼ p(referent) p(pronoun | referent)
The probability on the left-hand side of () represents the comprehender’s interpretation problem of decoding an ambiguous pronoun. The probabilities on the right are the estimates of the speaker’s encoding of the intended referent. From (), it is clear that interpretation and production of pronouns are linked. However, the two are not mirror images. Rather, Bayes’ Rule shows how they are related via the prior. Bayes’ Rule is not a model of pronouns; it is simply a statement of mathematical truth. However, it is useful for understanding pronouns because it makes it impossible to conflate the probability of mention and the probability of pronominalization. The proposal in () explicitly links different features to the two probabilities. ()
Proposal for a division of labor in a Bayesian model of pronouns (Kehler & Rohde, ) Different factors influence the probability of mention (the prior) and the probability of pronominalization (the likelihood). The prior probability of mention is influenced primarily by semantic and coherence-driven cues. The likelihood of forms is influenced primarily by structural properties of the preceding discourse.
To see how the Bayesian approach and this division-of-labour proposal play out, this section describes two studies that manipulate coherence-driven and structural factors.
.. Coherence-driven factors influence next mention Let’s return to the class of IC verbs discussed earlier. That earlier work by Stevenson et al. had noted different coreference patterns depending on the presence or absence of a pronoun. The Bayesian approach can help us make sense of that result. The story continuation paradigm allows us to explicitly measure the prior probability of next mention, the likelihood of pronominalization, and the pronoun interpretation probability for the available referents, as shown in (). By using story continuation prompts with and without a pronoun ((a‒c) versus (d‒f)), we can test which factors influence pronoun interpretation versus production. By using verbs with opposing IC biases (subject-bias vs. object-bias) and non-implicit-causality (Non-IC) verbs, we can test for coherence-driven effects and structural effects because the causally implicated referent appears either in subject position or object position (Rohde, ; see also Fukumura & Van Gompel, ; Kehler & Rohde, ). The prompts in () served as materials for a study reported in Rohde & Kehler ().
100
% Subject
80 60 40 20 0
Subj-bias
Obj-bias
Full-stop prompt
Non-IC
Pronoun prompt
.. Implicit causality story continuation results: verb class and prompt type influence next mention (p(referent | pronoun) ∼ p(referent) p(pronoun|referent)), as reported in Rohde & Kehler ()
()
Influence of coherence in interpretation and production of pronouns? a. [Subj-bias] Mary annoyed Sue. She ____ pronoun prompt b. [Obj-bias] Mary scolded Sue. She ____ to estimate c. [Non-IC] Mary babysat Sue. She ____ p(ref|pronoun) d. [Subj-bias] Mary annoyed Sue. ____ e. [Obj-bias] Mary scolded Sue. ____ f. [Non-IC] Mary babysat Sue. ____
g g
full-stop prompt to estimate p(ref), p(pronoun|ref)
The results of the study support the division-of-labour proposal. Figure . shows the percentage of continuations that started with a mention of the subject. The light bars confirm that participants’ choice of who to mention varies by verb class, in keeping with the proposal that the prior is sensitive to coherence-driven factors: IC contexts favour Explanation coherence relations; subject-biased IC verbs favour Explanations about the subject; object-biased IC verbs favour Explanations about the object; Non-IC verbs favour other coherence relations with less consistent subject/object preferences. The dark bars show that participants’ pronoun interpretation also varies by verb class, in keeping with the Bayesian approach. The coherence-driven effects on the prior contribute to the overall pattern of pronoun interpretation. These coherence-driven effects on choice of mention emerge as a main effect of verb class. In addition, the dark bars in Figure . are higher than the light bars, a main effect of prompt type. This would make sense if the likelihood of a pronoun being produced were higher for subjects than non-subjects because, in the Bayesian approach, that likelihood term also contributes to the pronoun interpretation probability. Indeed, Figure . shows that surface structure matters for pronominalization. More pronouns are produced when the antecedent is the subject of the preceding sentence, regardless of verb class. Even though object-biased IC verbs favour the object for remention, those mentions of the object are no more likely to be pronominalized than objects of subject-biased IC verbs (a main effect of antecedent grammatical role but no effect or interaction with verb class). Other contexts show a similar dissociation between prior
100
% Pronouns
80 60 40 20 0
Subj-bias
Obj-bias
Subject Mentions
Non-IC
Non-subject Mentions
.. Implicit causality story continuation results: antecedent grammatical role, but not verb class, influences pronoun production (p(referent | pronoun) ∼ p(referent) p(pronoun | referent)), as reported in Rohde & Kehler ()
probability of next mention and the probability of pronominalization (Kaiser, ; Kehler & Rohde, ; though see Rosa & Arnold, , and Davies & Arnold’s Chapter in this volume) for a competing approach to the role of predictability in pronominalization). Having shown how coherence-driven factors influence pronoun interpretation via the prior, the next study targets the structural effects that influence pronoun interpretation via the likelihood of pronominalization. The question is whether structural effects simply reflect grammatical structure (i.e. subjecthood) or whether such effects themselves are driven by information structure (i.e. topichood; see Bosch & Umbach, , for a similar question regarding the interpretation of personal pronouns and demonstrative pronouns in German).
.. Topichood influences pronominalization The finding that pronouns are produced at a higher rate for subject antecedents than nonsubject antecedents makes sense if the central function of pronouns (in English) is to signal a continuation of the current topic (Ariel, ; Gundel et al., ; Lambrecht, ; Grosz et al., ) and if subject and topic (in English) are highly correlated. Declining rates of pronominalization for antecedents in more oblique structural positions may reflect the declining likelihood that an entity in that position is the topic. A second study reported in Rohde & Kehler () manipulates topichood while holding grammatical role constant in order to test whether the subject bias in pronoun production reflects subjecthood or topichood. The materials for the study used a voice manipulation, as in (). ()
Do pronominalization rates reflect syntactic structure or information structure? a. [active] Amanda amazed Brittany. ___ full-stop prompt shows b. [passive] Brittany was amazed by Amanda. ___ p(ref), p(pronoun|ref) c. [active] Amanda amazed Brittany. She ___ pronoun prompt shows d. [passive] Brittany was amazed by Amanda. She ___ p(ref|pronoun)
g g
As noted earlier, the subject of an active clause in English is considered a default topic, but the passive’s promotion of another constituent to subject position makes that subject even more likely to be the topic. If the choice to produce a pronoun reflects topichood more generally, the prediction is that participants who re-mention the subject of the passive prompt (b) will be even more likely to use a pronoun than those who re-mention the subject of the active prompt (a). As such, a main effect of antecedent grammatical role is expected and would replicate the previous study’s structure-driven effect. The critical new finding would be an antecedent grammatical role voice interaction, whereby subjecthood increases pronominalization more in the passive. The presence of the pronoun prompt is hence expected to have a greater impact on choice of mention in the passive than the active condition: The passive prompt (d) is predicted to yield more subject continuations than the active prompt (c), both of which are predicted to yield more subject continuations than the full-stop prompts (a,b). Note that, because the active/passive manipulation changes the position of the causally implicated referent, the percentage of subject continuations is predicted to vary for coherencedriven reasons between the active and the passive as well. Main effects of voice and prompt type would replicate the previous study’s coherence-driven and structure-driven effects. Of interest is the voice prompt interaction, whereby a pronoun prompt has a greater impact on next mention in the passive condition. The results are shown in Figures . and .. All predictions were borne out. The rates of pronominalization in Figure . replicate the main effect of antecedent grammatical role and show the predicted antecedent grammatical role voice interaction: more pronouns for subject referents than object referents, and even more when the referent is the subject of a passive. The rates of subject re-mention in Figure . likewise replicate the previous study’s effects and show the predicted voice prompt interaction: more subject continuations for the active than the passive and more for the pronoun prompt than the full-stop prompt, and even more for the passive pronoun prompt condition. These two studies point to the importance of pragmatics in pronoun interpretation, via general reasoning underlying the establishment of discourse coherence and information
100
% Pronouns
80 60 40 20 0
Active
Subject Mentions
Passive Non-subject Mentions
.. Voice manipulation story continuation results: antecedent grammatical role and voice combine to influence pronoun production (p(referent | pronoun) ∼ p(referent) p(pronoun | referent)), as reported in Rohde & Kehler ()
100
% Subject
80 60 40 20 0
Active
Full-stop prompt
Passive Pronoun prompt
.. Voice manipulation story continuation results: voice and prompt type combine to influence next mention (p(referent | pronoun) ∼ p(referent) p(pronoun | referent)), as reported in Rohde & Kehler ()
structure underlying the choices in surface realization. The first study highlighted the role of the verb in guiding coherence-driven biases. The second study highlighted the role of syntactic structure in signalling topichood. Other non-structural factors may also influence topichood. For example, the presence of more than one referent in a context may reduce each one’s likelihood of being the topic, and there is experimental evidence showing that speakers indeed produce fewer pronouns when the context includes two referents compared to one (Arnold & Griffin, ). Kim’s chapter (Chapter in this volume) likewise discusses the role of information structure in domains such as coreference, including IC contexts. See also Tonhauser (Chapter , in this volume) on the complexity of establishing focus/topic categories and the repercussions for theories of meaning.
.. Model comparison The findings from these two studies support the Bayesian approach and the proposal of a division of labour between factors that influence choice of mention and choice of form. But how well are these results captured by alternative models? The coherence-driven model described in section . focuses on meaning at the expense of form. In these two studies, coherence-driven effects on pronoun interpretation emerge as biases to the causally implicated referent. These biases in the data can be measured in participants’ preferences in the full-stop condition: p(referent), the prior over referents. Focusing on the prior is the essence of Arnold’s () Expectancy Hypothesis—a model in which comprehenders are posited to interpret a pronoun to refer to the referent they most expect to hear mentioned next, with a variety of semantic and syntactic features being included as cues to this expectation. The problem is that a model based only on the prior overestimates the bias to the object referent for object-biased IC verbs and underestimates the bias to the subject referent for subject-biased verbs. Why? Because it fails to take into account how the presence of a pronominal form increases the number of subject interpretations. In contrast, a model that focuses only on surface structural factors ignores meaningdriven effects. In these two studies, structural effects emerge as a bias to produce more
pronouns for referents which were previously mentioned in subject position, particularly as the subject of a passive construction. This bias can be measured in participants’ behaviour in the full-stop condition: p(pronoun | referent), the likelihood of pronominalization. Focusing on the likelihood of pronominalization as a model of pronoun interpretation is at the heart of the mirror-image assumption—that is, the assumption that speakers produce pronouns to refer to a particular referent only if a comprehender who hears that pronoun will interpret it to refer to the correct referent. The problem is that a model based on the likelihood alone ignores the prior. There are cases in which the prior expectation for a mention of the non-subject referent is strong enough to shift comprehenders’ pronoun interpretation to that referent, even though they may be aware that pronouns are not the preferred form for re-mentioning that referent. The Bayesian approach incorporates the insights of both these types of models. It also shifts the analysis of pronouns from the question of how they are resolved to a question of how they are generated. Older models (e.g. Garnham & Oakhill, ) had focused on comprehenders’ interpretation as the target phenomenon and drawn conclusions about pragmatic inference playing a role only in a later stage of the interpretation process. By switching to a generative model, the focus is on how a comprehender estimates the speaker’s production process, and it suggests that the components of that process rely immediately and inextricably on pragmatic concepts like coherence and information structure.
. H ?
.................................................................................................................................. The Bayesian approach presented here reconciles two competing approaches to pronouns—one that targets coherence (Hobbs, ; Kehler, ) and one that targets surface structure and form (Grosz et al., ). Encouragingly, it also makes specific predictions that are borne out in experimental data. In addition, the Bayesian approach helps clarify the meaning of ‘prominence’, a term whose sense within the coreference literature is often muddled in discussions of speakers’ and comprehenders’ respective biases to produce and interpret pronouns to refer to ‘prominent’ entities. The Bayesian approach makes it clear that a referent can be prominent for coherence-driven reasons, such that a speaker is likely to mention that referent, or it can be prominent for information-structural reasons, such that, if mentioned, it is likely to be referenced with a pronoun. These two senses of prominence—prominence for mention and prominence for pronominalization— need to be understood on their own terms, particularly because they appear to reflect different properties of the discourse. Moving to a generative model also invites a reconsideration of a number of existing results in the literature. It’s possible that questions that were previously framed in terms of resolution can be better addressed from a generative perspective. For example, researchers have asked what factors guide pronoun resolution and whether those factors differ from the factors that guide the processing of names and full noun phrases (Sanford et al., ; Garrod et al., ). Sanford et al. report that referents are more likely to be mentioned again if they have first been introduced with a name than with a description and that
sentences with a pronoun are read faster if the antecedent was introduced with a name than with a description. But without knowing what form of reference is favoured by speakers when re-mentioning the available referents, we can’t tell whether the reading time results reflect facilitation from the form of reference (pronouns may be appropriate for antecedents introduced with names) or the choice of reference (the prior for re-mentioning an antecedent introduced with a name may be high even if the likelihood of using a pronoun is low), or both. This chapter has focused on third person singular pronouns and contexts in which those pronouns refer to a human referent. But this is only the tip of the iceberg for the use of pronouns. For example, the pronoun it and the demonstrative this can be used to refer back to events rather than entities, as in (), or even to a sequence of utterances that comprise a rhetorical argument, as in (). ()
The dog tripped, fell into his food bowl, and splashed the cat. It was a disaster.
()
The cat got annoyed because the food splashed its paws. The food splashed because the dog fell in the bowl. The dog fell because he tripped. This is what someone told me.
One can imagine using the Bayesian approach to model what kinds of messages a speaker might intend—messages about the entities, the events, or the discourse itself—and what kinds of referential forms would be used (see Brown-Schmidt et al., , Kaiser et al., , for contrasts between him/himself and it/that). Under a generative model, a comprehender who encounters the pronoun it in () must consider whether the speaker would have been more likely to reference one of the entities (the dog, the bowl, the cat), one of the events (the dog-falling event), or the sequence of events (the disastrous trip+fall+splash) and what forms would have been used in each case (‘the dog’, ‘it’, ‘Fido’, ‘the fall’, ‘the scene’, etc.). In addition, she may consider the possible ambiguity of the resulting expression. See Davies and Arnold’s Chapter (in this volume) on trade-offs between informativity and ambiguity in the choice of referring expression more broadly. Not only is this an opportunity to reassess pronoun use in English, but the Bayesian approach may also provide a framework for thinking about the referential system in other languages. In some languages, a human referent may be mentioned again with, say, a pronoun or a demonstrative, or a null or overt pronoun. Existing work in this area has pointed to a division of labour between particular referential forms and the antecedents they are compatible with (e.g. Carminati’s () Position of Antecedent Hypothesis), but the kinds of tasks that are used often conflate preferences for who gets mentioned and how. It may also be worth considering how second language learners master these two dimensions of reference in a non-native language, and whether success at estimating who will be mentioned is distinct from understanding what form is most appropriate (see Grüter et al., ). Lastly, the approach advocated here situates pronouns within one of the leading frameworks for thinking about language production and interpretation. Generative models have been applied to a number of pragmatic phenomena (Frank & Goodman, ; Goodman & Lassiter, ; Franke & Jäger, ), and this chapter makes the case that generative models can also clarify our understanding of pronouns. If you’re new to generative models, then pronouns may provide a convenient case study for seeing the explanatory power of such models and the role of experimentation in testing such models.
A The perspectives taken in this chapter reflect my PhD training with Andrew Kehler and our ongoing collaborations. I also am grateful to Jeff Elman for asking me ‘Why pronouns?’ nearly a decade ago. I thank the participants at the Anaphora & Coherence workshop at the North American Summer School on Logic, Language, and Information (NASSLLI) at Rutgers University for their input on this material and their invitation to consider some particularly tricky examples. This chapter was written with support from a Prize in Languages & Literatures to the author from the Leverhulme Trust.
......................................................................................................................
......................................................................................................................
.
. T
.................................................................................................................................. A essential property of language is that it refers to things in the world. Reference specifies this relationship between objects or ideas and linguistic expressions. As speakers plan their utterances, they make rapid decisions about how exactly they will refer to the entities that are part of their intended message. For every object or concept, a range of expressions is available, from highly explicit modified Noun Phrases (NPs), to less explicit deictic or pronominal expressions. Even a familiar, concrete, co-present object can be referred to as my favourite stripy mug, the cup, this one, or it. Though each of these choices would be semantically acceptable, they differ in terms of their pragmatic appropriateness. In this chapter we examine the question of how speakers choose linguistic expressions to refer to things. As we will show, speakers are heavily influenced by context, and exhibit strong preferences for certain referential forms. For example, I would be unlikely to say ‘it’s red’ to talk about a cup that is not either part of the context or known to my addressee. Likewise, a story about my cup would likely not involve repeated explicit descriptions (‘I love my red stripy cup. My red stripy cup was a present from my friend. I use my red stripy cup all the time’). Addressees also have expectations about the referential forms they encounter. They are sensitive to the appropriateness of referential choices, and work to find explanations for the use of unexpected or inappropriately gauged expressions (see section ...). Many researchers have noted that referential expressions fall along a continuum of explicitness (Chafe, , ; Givón, ; Ariel, , ; Gundel et al., ). Yet this continuum results from speakers making choices along numerous dimensions, a few of which are illustrated in Table ..
Table .. Choices between referential expressions Referential dimension
Examples
Modified vs. unmodified Explicit name or description vs. pronoun Definite vs. indefinite Name vs. description
the red cup or the cup I saw vs. the cup Jane/the president vs. she the cup vs. a cup Jane vs. the president
In this review, we focus on two dimensions of referential choice: modification (do I need to use more than just a noun to communicate reference?) and pronominalization (can I use a pronoun instead of a more explicit expression?) Both of these dimensions are critical to successful communication because providing too little information might lead to miscommunication, and producing a too-specific expression may sound clunky and interfere with the listener’s ease of understanding. In the next section, we review two major theoretical approaches accounting for speakers’ referential preferences, both of which highlight the fact that speakers’ choices are highly constrained by what is appropriate in context. We then examine questions about the cognitive processes underlying speakers’ choices, and the impact of these decisions on communication.
. T
..................................................................................................................................
.. Gricean maxims Within the theory of Conversational Implicature (), Grice proposed four maxims of conversation: Quality, Quantity, Relation, and Manner. Most relevant to the notion of informativeness is the maxim of Quantity, itself subdivided into two parts: (i) Make your contribution as informative as is required for the current purposes of the exchange; (ii) Do not make your contribution more informative than is required. That is, speakers are expected not to give less or more information than necessary for the purposes of the interaction. For example, when referring to one type of flower in a garden containing many other types, a rational speaker would be expected to provide sufficient information for her addressee to unambiguously identify the intended flower by using a term more specific than the basic level term ‘flowers’, for example ‘peonies’, or by modifying her expression such that it distinguishes the flower in question from others in the garden, for example ‘the big pink blousy flowers’. In doing so, the Gricean speaker conforms to the hearer’s expectation that she would make her contribution as informative as required. Regarding the second submaxim, a speaker would not be expected to provide that level of detail if the garden contained just one type of flower. Thus, under a Gricean account, speakers are
.
expected to be minimally informative, producing utterances that convey all of the necessary information while being as economical as possible. Crucially for this theory of implicature, if a speaker fulfils these expectations, there is no need for the addressee to engage in extra pragmatic interpretation beyond what is literally stated. So rather than prescribing proper conversational behaviour, the function of the maxims and the Cooperative Principle to which they belong is to account for the derivation of conversational implicatures via comparisons between what is expected and what is heard. For example, if a speaker flouts Quantity- and uses ‘flowers’ to refer to a specific type of flower in a context containing many other types, the addressee might infer that this underinformative speaker does not know the hyponymic term, or that other aspects of the context might disambiguate the apparently underinformative expression (e.g. prior reference). If the speaker flouted Quantity- and used ‘the big pink blousy flowers’ in a single-flower-type context, the addressee might make a contrastive inference, and assume the existence of a non-pink, non-blousy flower. Or, he might make an inference based on speaker goals, hypothesizing that the more detailed description of the flower is critical in the current exchange. Although Grice never intended his theory of conversation to explain the psychological processes by which we understand utterances (Saul, ), his work has influenced the field of reference in two major ways: it has provided a theoretical backbone for the development of subsequent models of reference and informativeness; and it has inspired experimental investigations to test his philosophical ideas (see Noveck and Reboul, ).
.. Informativeness ... What is informativeness? Consistent with Grice’s Cooperative Principle, one theoretical tradition explains how speakers choose referential expressions on the basis of informativeness (e.g. Deutsch & Pechmann, ; Brown-Schmidt & Tanenhaus, , ; Engelhardt et al., ; Davies & Katsos, ; Arts et al., a,b; Pogue et al., ; among others). The basic assumption behind this approach is that speakers aim to be cooperative by providing the right amount of information for communicative success, where the ‘right amount’ is defined by the context. The definition of informativeness depends on the interaction between the context (including visual and discourse features, and speaker intentions) and the referring expression. That is, informativeness is affected as the context and/or the expression changes. Thus, informativeness is defined as a property of expressions within their contexts, such that more informative expressions are those that match a smaller set of candidate referents in the context. For example, the expression ‘the cup’ is informative in the context of a lone cup, but as more cups are introduced into the context, it decreases in informativeness and ambiguity increases. And although ‘the large cup’ is informative in the lone cup context, it is not felicitous (in Gricean terms) since the size modifier is redundant and the expression risks the addressee drawing a false contrastive inference. As more potential (small) cup referents are introduced into the scene, ‘the large cup’ becomes more felicitous. Note that, under this conceptualization, the informativeness of the modified expression would not
increase as more (smaller) cups are introduced, since it fits exactly the same number of candidate referents in both the single- and multiple-cup contexts. Using this reasoning, the literature has typically adopted a three-level taxonomy to categorize levels of felicity: optimally or minimally informative (e.g. ‘the large cup’ when two cups differ in size), underinformative (‘the cup’ in the same situation), or overspecified1 (‘the large cup’ in the lone cup context). Although these levels of referential informativeness are reasonably intuitive, scholars have asked how contextual constraints are calculated and applied to both the speaker’s linguistic choices and the listener’s pragmatic inferences. Recent research has conceptualized informativeness as the extent to which a referring expression reduces uncertainty about the identity of a referent, given the set of potential referents in the concurrent domain (Frank & Goodman, ; Pogue et al., ). Informativeness in this sense has also been cast as surprisal in work formalizing informativeness by modelling human behaviour (Frank & Goodman, ; Goodman & Stuhlmüller, ). Surprisal is an information-theoretic measure of the degree to which a word reduces uncertainty about the speaker’s message (see also Hale, , Levy & Jaeger, ; Mahowald et al., ). All of the conceptualizations discussed are in the spirit of the Gricean expectation of unambiguous referring. Informativeness is discussed both as a property of referring expressions and a property of speakers (Mangold & Pobel ; Davies & Katsos, ; Frank & Goodman, ; Goodman & Stuhlmüller, ; Frank & Goodman, ; Davies et al., ; Davies & Kreysa, ). If a speaker composes her referring expression with words that apply to the intended referent and no/few others, ‘words which pick out relatively smaller sections of the context’ (Frank & Goodman, : ), she is informative in that situation. This behaviour may be considered a trait if speakers habitually use highly informative terms (Grodner & Sedivy, ; Yildirim et al., ). Relatedly, a message is informative about a speaker’s intended meaning in that particular communicative context. If the communication is successful, the addressee’s uncertainty about a speaker’s intended meaning will be reduced. Thus, a referring expression will only be fully informative if it matches the demands of the context (e.g. the presence of a contrast object of the same nominal class) and the speaker’s intention (e.g. for the addressee to uniquely identify a referent and/or to realize the greater or lesser importance that specificity plays in the current communicative context). In sum, informativeness is a function of referring expressions, speakers, and contexts. Even under models of informativeness that contrast with the model sketched earlier, such as those that claim that ambiguous referring expressions are efficient and desirable (Piantadosi et al., ), the role of context is to inform about meaning. There cannot be a measure of informativeness in the absence of any of these components.
... Empirical investigations of informativeness in production A rich seam of research into the incidence and effects of informativeness has been opened in experimental pragmatics and psycholinguistics over the last decade. A number of studies We favour the terms ‘overspecified’ or ‘over-modified’ in contrast to ‘overinformative’ commonly found in the literature (Davies & Katsos, ; Engelhardt et al., ; Pogue et al., ; among others) since informativeness is no greater in overspecified referring expressions than in minimally specified ones. 1
.
take a functionalist approach to understanding informativeness, consistent with Grice’s proposal that speakers make choices for the purpose of communication. This work tends to show that hearers expect felicitously informative expressions in communicative exchanges, and that speakers are largely rational with respect to this. Methodologically, researchers have used referential communication games (Deutsch & Pechmann, ; Mangold & Pobel, ; Maes et al., ; Brown-Schmidt & Tanenhaus, , ; Engelhardt et al., ; Davies & Katsos, ; Nieuwland et al., ; Arts et al., b; Brown-Schmidt & Konopka, ; Engelhardt et al., ; Koolen et al., ; Davies & Kreysa, ), the visual world paradigm (Brown-Schmidt & Tanenhaus, , ; Engelhardt et al., ; Brown-Schmidt & Konopka, ; Davies & Kreysa, ), and Event-Related Potentials (ERPs) (Nieuwland et al., ; Engelhardt et al., ) to manipulate aspects of the discourse context and monitor addressees’ online interpretations of referring expressions. For example, a visual display might contain two objects of the same type, such as a large and a small apple, or just one object of that type, requiring a modified referring expression for disambiguation in the former case, and not in the latter. Typically, levels of informativeness produced by speakers and/or the timing and pattern of eye movements around the referential scene are analysed. These approaches have been successful in providing controlled contexts as a backdrop to reasonably spontaneous language and its processing from both speakers’ and addressees’ perspectives. They have also shed light on the processes linking speakers’ fixations to aspects of a scene and their referential choices. For example, Brown-Schmidt & Tanenhaus () found that earlier fixations to a contrast object were associated with use of a prenominal modifying adjective, whereas later fixations were associated with postnominal modification. Thus, speakers are found to revise their utterance plans on-line. In general, adults2 produce informative expressions, and in that sense they are Gricean. For example, when participants were presented with an array containing both a big and small star, they typically requested a target object (e.g. a big star) using expressions with modifiers (‘the big star’) around per cent of the time (Davies & Katsos, ; Davies & Kreysa, ). Similarly, Brown-Schmidt & Tanenhaus (: Experiment ) found that speakers produced a modifier like ‘with the triangles’ to identify a target square in a context of squares with and without triangles per cent of the time (cf. % for displays with a single square), and Engelhardt et al. () found that modified utterances were more common in a contrast condition than in the singleton condition (% vs. %). When displays are very simple, rates of informativeness can reach ceiling. For example, Nadig & Sedivy () found that adults consistently produced a modifier like ‘large’ to identify a large glass in a context of a large and small glass, and never produced a modifier when there was only one glass. Although speakers are usually informative, they sometimes fall short of this expectation. For example, they may fail to notice that there is another object in the context that could be confused with their intended referent. Ferreira et al. () found that speakers were more likely to provide modification for sets of differently sized identical objects (e.g. a big and small bat) than for sets of objects that were conceptually distinct but had the same linguistic
2
There is also a large developmental literature on reference and informativeness. For reviews, see Graf & Davies (); Serratrice & Allen ().
label (e.g. a bat-for-sports and a bat-as-flying-mammal). This and other work also suggests that, unsurprisingly, modification is more likely when the speaker has attended to the competitor items in the context (Brown-Schmidt & Tanenhaus, ; Wardlow-Lane & Ferreira, ; Davies & Kreysa, ). Thus, knowledge of context or attention to context modulates use of that context in referential choice, a point to which we return in section . on speaker goals and feature salience. Speakers appear to make decisions about modification based on the broad goal of communicative success, not merely as a simple response to the presence of competitor objects in the context. For example, if the conversation has already focused on an object such as a figure that looks like an ice skater, the speaker can use a simpler expression, like ‘the skater’, even if there are other figures in the context that may roughly fit this description (Clark & Wilkes-Gibbs, ). Brown-Schmidt & Tanenhaus (: Experiment ) categorized almost half of the referring expressions as ambiguous in a referential communication task: for example, speakers asked addressees to put an object ‘above the red block’ where there were several red blocks. In such cases, the underinformative referring expression ‘the red block’ was licensed since aspects of the task ruled out many of the alternatives, for instance if only one of the red blocks had a free space above it. So, while speakers generally adhere to expectations of informativeness, expressions that superficially appear to be underinformative do not necessarily present difficulty to addressees, thanks to the integration of additional cues from context. Speakers sometimes provide more information than is strictly needed. For example, Engelhardt et al. () presented participants with a scene including a single apple on a towel, an empty towel, a puppet, and an empty box. Speakers frequently overspecified the target by saying ‘put the apple on the towel on the other towel’ (cf. ‘put the apple on the other towel’), and hearers were equally satisfied with both variants despite the fact that the first one was overspecified. However, Davies & Katsos () argued that this redundancy serves other purposes, for example helping listeners interpret a complex scene, especially when the redundant information is salient (in this case, the duplication of the towel in the scene, and the fact that the target referent was composed of two objects). Thus, if we take human processing biases into account in order to identify the amount of information that is ‘required’, these situations may not represent true violations of the second Quantity submaxim. It seems that there are a number of properties that lead people to provide redundant information when it might help find the referent or identify the intended action. This is backed up by natural language generation literature documenting frequent overspecification in particular conditions (Koolen et al., ; Gatt et al., ; discussed in section ..).
... Empirical investigations of informativeness in comprehension Addressees make fast inferences about what their speaker intends to refer to in an unfolding discourse (Tanenhaus et al., ). For example, on processing modified nouns, they recruit information from prenominal adjectives to restrict the domain of potential reference not only to those objects displaying the property encoded in the adjective but also those objects to which the adjective might plausibly apply. In a classic study of this process, Sedivy et al. () showed that on hearing ‘the tall glass’ speakers were quicker to resolve reference to a glass that was part of a contrast set than to a singleton tall glass, where using the adjective would have been overspecific (see also Wolter et al., ). Other forms (e.g.
.
the ambiguous pronoun ‘she’ when there are two female protagonists in the discourse) may trigger a process of revision as addressees work to override incorrect predictions as the discourse proceeds (Rohde, Chapter in this volume). This is not necessarily problematic: even when speakers deviate from expected forms, addressees are able to accommodate them by using aspects of the communicative context—this is pragmatic processing in a nutshell. Whether speakers diverge from expected forms intentionally or otherwise (e.g. when they misjudge the accessibility of a referent and provide inappropriately explicit referring expressions), addressees ultimately resolve reference. Hearers are generally good at dealing with unexpected or infelicitous referential forms. As mentioned, speakers produce underinformative referring expressions when aspects of the communicative situation restrict the referential domain, ruling out competitors that would otherwise match the expression (Brown-Schmidt & Tanenhaus; ). In that study, addressees asked for clarification in only per cent of ambiguous trials and were not generally confused by referring expressions that would have been ambiguous if the referential domain had not restricted reference (a pattern confirmed by eye movement analyses, which showed that addressees exhibited a strong preference to fixate the target). These findings provide evidence that addressees are well able to resolve ambiguous reference by using extra-linguistic information. But what about in contexts where an underinformative referring expression is truly ambiguous? Although there has been extensive developmental research on children’s responses to underinformative expressions,3 there has been less direct attention on adults’ responses to utterances that do not ultimately disambiguate. Off-line acceptability judgements of underinformative expressions are lower than those of informative expressions (Davies & Katsos, , ). In naturalistic conversation, asking for clarification is a common strategy. This has been described by Clark & Wilkes-Gibbs () as the principle of mutual responsibility, under which interlocutors minimize their referential efforts using underinformative referring expressions in the knowledge that their partner can ask for more detail if required. When encountering ambiguous pronouns, adults are able to disambiguate on the basis of information presented in the preceding discourse, such as mapping a pronominal form onto the most prominent discourse referent (Light & Capps, ; Hendriks et al., ). Thus addressees recruit information from linguistic and non-linguistic sources. If these do not ultimately disambiguate, they may verbally seek clarification. What, then, are the effects of overspecified expressions on the hearer? A speaker might produce a higher level of specification than an addressee is expecting given the referential context, for example ‘the brown rabbit’ to refer to a singleton rabbit. In such situations, the addressee may make a contrastive inference, whereby they assume the presence of a competitor in the discourse (Sedivy et al., ). In a study investigating the incidence of contrastive inference in a card game task, addressees inferred in around per cent of cases that the speaker was holding a hidden card featuring a white rabbit after she had referred to a brown rabbit in the same hand of cards (Kronmüller et al., ). Thus, adults frequently enrich modified utterances. Alternatively, if after doing the necessary inferential work to justify a speaker’s inclusion of an adjective, the addressee concludes that the modifier was
For example, Nadig & Sedivy () found that ‒-year-olds used various strategies for resolving genuinely underinformative reference. See Morisseau et al. () for a review of other work in this area. 3
genuinely gratuitous, due to feedback about the function of the adjective on earlier trials (Engelhardt, ), or having been explicitly told about the speaker’s unreliability (Grodner & Sedivy, ), the contrastive inference may be suspended. Off-line measures complement this finding, revealing that acceptability judgements are lower for overspecified relative to minimally informative expressions, though they are penalized less than underinformative ones (Davies & Katsos, ). A recent debate in the literature has focused on whether overspecified referring expressions help or hinder reference resolution (Rubio-Fernández, ). For example, some studies find that such terms lead to faster identification of the target referent (Mangold & Pobel, ; Arts et al., b). Conversely, other studies have concluded that overspecifications impair comprehension (Engelhardt, Bailey, et al., ; Engelhardt, Demiral, et al., ). One possibility is that the differing outcomes are a methodological artefact. For instance, consider the relative simplicity of Engelhardt et al.’s () materials. In that study, addressees had to identify a target from two-object displays following a modified expression, for example ‘the red circle’. Longer reaction times for the overspecified expression were taken to indicate an impairment to comprehension, though the methodology used does not clarify whether the latencies indexed delayed visual identification and/or implicit pragmatic judgements of the unexpectedly overspecified referring expression (note that overspecification is rare in simple two-figure displays; Rubio-Fernández, submitted). In addition, findings that overspecification facilitates comprehension tend to come from studies of written language (e.g., Arts and colleagues), while evidence that overspecification impairs comprehension stems from studies on spoken language (e.g. Engelhardt and colleagues). Further work on this question is needed. Care should be taken in reconciling approaches using referential communication in the visual world, and those using extended discourses (written or spoken). In the former, evidence suggests that overspecified reference helps the search for a referent, whereas in the latter, inappropriately heavy NPs can impede comprehension (the Repeated Name Penalty; see section .. and Rohde, Chapter in this volume). As when resolving underinformative expressions, addressees recruit multimodal resources to interpret the speaker’s utterance, a strategy not available in reading or unimodal listening tasks.
.. Discourse theories: Framework and empirical evidence A second major theoretical tradition has focused on how referential expressions are constrained by the linguistic discourse context (Chafe, , ; Ariel, , ; Gundel et al., ; Grosz et al., ; Gordon & Hendrick, ), which affects the referents’ information status (see Arnold et al., , for a review). This work builds from the assumption that speakers and their addressees maintain mental representations of the discourse and situation, where information within these representations varies in its cognitive status (Bransford et al., ; Johnson-Laird, ; Van Dijk & Kintsch, ; Kintsch, ; Bower & Morrow, ; Zwaan & Radvansky, ). One of the broadest characterizations of information status is the distinction between given and new information. Things that are physically present or have been mentioned linguistically are considered ‘known’, ‘old’, or ‘given’ in context whereas
.
information that has not been contextually evoked is ‘new’ (Halliday, ; Chafe, , ; Prince, , ). This distinction has been used to explain linguistic choices such as the use of definite (e.g. ‘the cup’) vs. indefinite expressions (‘a cup’), or the contrast between prosodically prominent (‘the CUP’) vs. reduced (‘the cup’) expressions. Yet all given information is not equal. Certain events and entities are considered more salient, topical, accessible, or prominent. While these terms vary, they all refer to some dimension of cognitive status. Scholars in this tradition have proposed that the cognitive status determines the appropriate linguistic form. For example, highly reduced expressions, like pronouns (he, she, they, it), are specialized for highly accessible referents. On the other hand, specific expressions (‘my third grade piano teacher’) are reserved for highly inaccessible referents, such as when new referents are introduced to the discourse. Chafe (, ) defines accessibility as the degree to which the referent is active in the hearer’s consciousness. On this account, a referent can be active (in the current focus of consciousness), semiactive (in peripheral consciousness, i.e. the prior topic) or inactive (neither active nor semiactive for the duration of the exchange). In contrast, starting with the speaker’s choice of referring expression, Ariel () argues that specific linguistic terms signal to the hearer where to ‘look’ for the referent in the current discourse: in current focus, or somewhere more distant. That is, modified NPs signal least accessible information while pronouns signal highly accessible information. Working in the same direction, Gundel et al.’s () givenness hierarchy claims that a pronominal form like ‘it’ signals that the referent’s cognitive status is that it is the current topic of the conversation and thus highly accessible, whereas a NP with an indefinite determiner, such as ‘a cat’, identifies the type of referent with the lowest cognitive status. Many theorists suggest that discourse salience/accessibility can be explained in terms of attention (Bower & Morrow, ). Gundel et al. () suggest that the speaker selects referential forms based on assumptions about their addressee’s attention, and Brennan () says that speakers use word order and other linguistic devices to indicate the focus of their own attention. The discourse-based framework has been used to explain the fact that pronouns and other reduced expressions tend to be used more often in certain linguistic contexts. For example, pronouns are most likely for referents that have been recently mentioned. These patterns emerge in analyses of written texts and elicited narrative production (e.g. Givón, ; Ariel, ; Arnold et al., ). For example, Arnold () analysed the use of pronouns in written stories, and found that pronouns represented per cent of references to something mentioned in the last clause, but only per cent of references to entities mentioned two to five clauses back. This frequency pattern matches the conditions that are easiest for comprehenders. For example, Clark & Sengul () asked people to read stories like A broadloom rug in rose and purple colors covered the floor. The light from a small brass lamp cast shadows on the walls. In one corner of the room was an upholstered chair. The chair appeared to be an antique. They measured the reading time of the last sentence, which was faster when the referent had been in the previous clause, both when the anaphoric expression was a NP (the chair) and
when it was a pronoun, compared to when the referent had appeared two or three sentences previously. They concluded that the previous clause has a privileged place in memory. Syntactic position also has a strong effect on pronoun use (see also Rohde, Chapter in this volume). One of the most robust observations is that pronouns tend to be more common when the referent was last mentioned in subject position, compared with objects or obliques (Brennan et al., ), as shown by corpus and speech elicitation experiments (Stevenson et al., ; Brennan, ; Arnold, , ; Kehler et al., ; Arnold et al., ). Visual world eye-tracking studies show that when listeners hear an ambiguous pronoun, they show a preference to gaze at a picture of the subject of the previous clause, and this preference emerges as early as milliseconds (ms) after the pronoun onset (Arnold et al., ; Järvikivi et al., ; Arnold, ; Arnold & Lao, ; Hartshorne, Nappa, & Snedeker, ). Reading studies find that pronouns are read more quickly when their antecedent was the subject (Fukumura & Van Gompel, ). Moreover, comprehend ers also experience problems in the face of expressions that are more explicit than necessary. For example, Gordon et al. (; see also Gordon & Scearce, ; HudsonD’Zmura & Tanenhaus, ; Almor, ; Fukumura & Van Gompel, ) found that readers slowed down when they read a repeated name referring to a highly prominent entity, such as the subject of the previous clause. That is, it sounds unnatural to read Bruno was the bully of the neighborhood. Bruno chased Tommy . . . , where you would expect a pronoun instead of the second ‘Bruno’, and this unnaturalness slows comprehension processes. The effects of grammatical role are especially strong when the pronoun and its antecedent fall in parallel grammatical roles (Sheldon, ; Chambers & Smyth, ). In addition, pronouns can be considered appropriate for referents that have occurred in syntactically focused positions (Arnold, ; Almor, ; Cowles, Walenski, & Kluender, ; Foraker & McElree, ). The relevance of the discourse context has often been characterized in terms of the topicality of referents—the idea that the discourse is ‘about’ some referents more than others (e.g. Givón, ). The idea that pronouns refer to topics is also frequently instantiated in computational models: for example, Van Rij’s () model suggests that pronouns refer to the subject of the previous sentence due to its high topicality. Centering Theory (Brennan, ; Grosz et al., ; see also Kim and Rohde, Chapters and , respectively, in this volume) recasts the notion of topic as the ‘center of attention’, which is computed on the basis of grammatical function and anaphoric links across utterances (for a related model, see Gordon & Hendrick, ). As in the informativeness tradition, some scholars in the discourse tradition have proposed to account for contextual constraints in terms of predictability. For example, Prince () identifies three ways in which givenness has been defined: () recoverability/ predictability, () salience, and () hearer knowledge, and she goes on to say that the three are related. Givón () explains topicality in terms of properties like persistence, which is the number of clauses over which a referent will continue to be mentioned. Likewise, the Centering formalism proposes that discourse entities are ranked in a list of forward-looking centres, typically based on a grammatical function hierarchy (subject < object < oblique; Brennan et al., ). For example, in ‘The dog buried the bone in the yard’, the three descriptions are ranked (dog, bone, yard). The top-ranked one is the ‘preferred center’, and is the most likely one to be the backward looking centre
.
(similar to the topic) in the next sentence. The definition of centres in terms of likelihood of continuation highlights the relationship between referential predictability and discourse coherence. In addition, corpus analyses show that when a referent has been mentioned recently, or in a prominent syntactic or semantic position, it has a higher likelihood of being mentioned again in the next utterance than less recent or less prominent entities (Arnold, , ). This led Arnold to propose the Expectancy hypothesis, which suggests that linguistic cues to accessibility are relevant in that they indicate likelihood of continued importance to the discourse. Nevertheless, the relation between discourse accessibility and prominence is debated. With respect to reference comprehension, researchers agree that comprehension is facilitated when the context increases the likelihood of the referent being mentioned at all. Much of this research comes from evidence that the semantic structure of utterances affects the predictability of reference to discourse entities. For example, in Sandy admired Kathryn because . . . , people expect Kathryn to be the cause of the event, and when participants are asked to invent a continuation of this story, they are more likely to mention Kathryn than Sandy (Stevenson et al., ; Kehler et al., ; Fukumura & Van Gompel, ; Rohde & Kehler, ). Similarly, in Sandy sent a letter to Kathryn, and then . . . , the transfer event leads to an expectation that the goal of the transfer (Kathryn) will feature in the next event more than the source (Sandy; Stevenson et al., ; Arnold, ; Rosa & Arnold, ). That is, the thematic roles of discourse participants are associated with referential predictability. This expectation guides pronoun comprehension, where readers are faster for expected references and more likely to choose the expected referent (Caramazza et al., ; McDonald & MacWhinney, ; Garnham et al., ; Stewart et al., ; for other types of predictability effects, see Altmann & Kamide, ; Arnold, Tanenhaus, Altmann, & Fagnano, ; Arnold, Hudson Kam, & Tanenhaus, ; Arnold & Lao, ). By contrast, researchers debate whether predictability affects the speaker’s decision to use pronouns vs. names/descriptions. On the one hand, several researchers have found that for the causal type of sentences shown earlier, semantic inferences have no effect on pronoun production choices (Kehler et al., , Fukumura & Van Gompel, ), and instead speakers use pronouns only for syntactically prominent referents. This led Kehler & Rohde (; Kehler et al., ; Rohde & Kehler, ) to make the strong claim that the speaker’s decision to use a pronoun is driven only by syntactic and topicality factors, but not by semantic factors. On the other hand, other studies have shown that pronoun production can be influenced by factors related to predictability. Rosa & Arnold () found that the predictability of goal entities following transfer verbs leads to a small but robust tendency to use pronouns for goals more than for sources (see also Zerkle & Arnold, ). Using a different approach, Tily & Piantadosi () asked participants to play a guessing game, where they gave them newspaper articles with certain words missing. They emphasized that the game was to guess the thing or person that would be mentioned, not the particular word. They found that people were more likely to guess correctly when the original context had contained a pronoun or a proper name, as opposed to a description (e.g. ‘the woman’). In sum, it appears that predictability can affect referential decisions, but it may not be the only relevant contextual property.
In addition to the constraints of information status, the discourse-based tradition recognizes the relevance of ambiguity. As described in section ..., a two-cup situation calls for a modified expression like ‘the red cup’, or ‘the cup on the left’. Likewise, a situation with two boys means that the English pronoun ‘he’ is ambiguous. Indeed, speakers appear to be sensitive to this ambiguity. For example, speakers are more likely to say ‘she’ for Ana following Jacob saw Ana and she . . . than following Liz saw Ana and Ana . . . (Francik, ; Karmiloff-Smith, ; Arnold et al., ; but for a different interpretation, see Arnold & Griffin, ; Fukumura et al., ).
.. Comparing informativeness and discourse approaches The informativeness and discourse-based approaches focus on different linguistic phenomena, but they both make the same general claim: referential forms vary in their appropriateness, on the basis of the constraints on successful communication within a discourse context. Both traditions suggest that less-specific expressions are acceptable when the referent is unambiguous. Both traditions suggest that the informational context constrains the degree of specificity of referential expressions, and both suggest that these constraints are related to the communicative goals of the discourse participants. In addition, scholars within both traditions have proposed that the context is important because it helps comprehenders predict the speaker’s meaning. When the speaker’s meaning is redundant given the context, it is highly predictable, and less explicit linguistic input is needed for successful communication. As explained in section .., computational models of informativeness use surprisal to model contextual effects. Likewise (see section ..), the relevance of the discourse context is at least partly related to predictability, although this issue is debated. Thus, the informativeness and discourse traditions overlap in their goal of accounting for contextual effects on reference form, and share some theoretical proposals. Nevertheless, each tradition represents different dimensions of the reference production process. The informativeness tradition focuses on what it takes to achieve successful communication. That is, have I given enough detail to allow my addressee to pick out the right referent, but without being too verbose? This focuses heavily on the presence of competing referents in the physical or linguistic context. The discourse tradition instead focuses on the functional role of referring expressions. Pronouns do more than just establish reference, they also provide a signal that the utterance should be connected with previous information, a communicative goal that goes beyond simply indicating a referent. This goal is heavily influenced by the structure of the discourse, which has led to a focus on the prior linguistic context. There are also conceptual differences between the two approaches. For example, in the informativeness and computational literature, perceptual salience is associated with things that are new, highly contrasting, or surprising (Gatt et al., ; Clarke et al., ). This is the opposite of discourse salience, which is associated with things that are old or predictable. Care should be taken not to conflate the two.
.
. R- : C
.................................................................................................................................. The informativeness and discourse-based frameworks suggest a regular correspondence between the context and speakers’ choices about how much information to provide when referring. However, real-life references present widespread variation in the application of these contextual constraints. This raises questions about how these models should be instantiated. How precisely is the context calculated, and does it involve calculations about what other people know, or just the speaker’s own knowledge? How is the context applied when selecting referential expressions? How is this process influenced by individual (e.g. speaker goals) or situational (e.g. referential pacts; contextual salience) differences? Do aspects of the referents themselves influence referential choice? We now address these questions by reviewing empirical investigations into some of the constraints involved in referential processing.
.. What’s in the context? Linguistic and non-linguistic sources of information If I want you to pick up an apple that’s on the table in front of us, how do I choose an appropriate referring expression? Which aspects of the context do I use in my mental calculations? Multiple contextual dimensions are at play, including physical, linguistic, and interpersonal. A full theory of reference production requires an understanding of which ones are relevant, and how they interact. The discourse-based tradition focuses heavily on the constraints from the linguistic, textual context. In fact, some computational models make the simplifying assumption that speakers choose pronouns to refer to the grammatical subject of the preceding sentence (Kehler & Rohde, ; Van Rij et al., , or other text-based calculations (e.g. Centering Theory: Brennan, ; Grosz et al., ). On the other hand, scholars agree that the nonlinguistic context matters too. The non-linguistic context clearly affects modification preferences (see section ..), and pronouns can be used deictically, such as pointing at a cat while saying ‘it’s so furry!’ One question is how we should define ‘given’ information. Clark & Marshall () propose that given status is defined in terms of what is mutually known amongst discourse participants, and suggest that people use heuristics based on physical and linguistic co-presence to estimate shared knowledge. If we both see something, or have both heard it, we know it is given. Yet other work suggests that linguistic and conceptual information may have independent effects. One well established generalization is that reference to given information tends to be prosodically reduced, while reference to new information tends to be accented, and acoustically prominent (Halliday, ). Kahn & Arnold () tested how speakers
pronounced their words as they described objects moving on screen, for example ‘the airplane rotates’, where the target object (the airplane) was pictured on screen as one of six objects. Before the object movement, participants were exposed to one of three priming conditions: linguistic priming (a voice spoke the names of three objects), conceptual priming (three objects flashed on screen), or a control condition on which nothing happened. In the linguistic and conceptual priming conditions, the target was the last of the three objects to move, which meant that it was also fully predictable. They found that the target name (airplane) was shorter in the conceptual priming condition than the control condition, but that it was even shorter in the linguistic priming condition, where the target word was given both linguistically and conceptually (since the word also evoked the concept). This suggests that linguistic and conceptual exposure have independent effects (see also Baumann & Hadelich, ). By examining how the non-linguistic context affects speakers’ choices, we can independently address questions about how speech is influenced by properties of the context, as well as questions about what the speaker and hearer know jointly or independently. Later, we discuss each of these types of constraint as we consider how referential choice is influenced by both top-down (e.g. constraints stemming from the interlocutors themselves) and bottom-up (e.g. features of the referents) pressures.
.. Constraints from interlocutors ... Does common ground matter? Researchers agree that the context constrains reference use, but they disagree about the role of common ground. Common ground refers to information that is mutually known—that is, if you can see a cup on the table, and I know you see that cup, and you know that I know that you see the cup, etc., we consider that knowledge to be in common ground (Clark & Marshall, ). For many theorists, common ground defines the relevant context. For example, I cannot say ‘it’s blue’ to you if we have no basis for the assumption that you know what ‘it’ refers to. This leads scholars like Chafe (: ) to suggest that the goal of speakers is ‘categorize a shared referent in a way that allows the listener to identify it’ (see also Gundel et al., ). On this view, common ground is fundamental to understanding the process of both choosing referential forms and understanding them. For example, interlocutors quickly assess their partner’s level of expertise on a subject, and design their utterances accordingly, such as providing more detail when one of the pair is a novice (Isaacs & Clark, ). Speakers also track their addressee’s feedback, and adjust their utterances as needed. For example, Clark & Krych () examined how speakers instructed addressees to build Lego models. When both people could see the workspace, addressees sought feedback, for example pausing before placing a block, and speakers rapidly responded to these gestures. Other work demonstrates sensitivity to cultural or contextually-established common ground (Clark et al., ; Clark & Wilkes-Gibbs, ; Clark & Schaefer, ); see also the section on conceptual pacts (...). In addition, discourse participants actively seek out information about their partner’s knowledge by asking questions (Brown-Schmidt et al., ), and the task goals guide the degree to which people consider their partner’s perspective (Yoon et al., ).
.
On the other hand, other researchers have pointed out that the cognitive means of tracking other people’s knowledge and perspective can be demanding, and that this places limits on the extent to which real-time choices are made. For example, interlocutors’ knowledge states have to be represented based on available information deduced from various cues, such as visual context, textual distance, hearer feedback, etc. (De Cat, : ). Studies have found that when there is a conflict between the speaker’s knowledge and their addressee’s knowledge, they do not always ignore their own information (Keysar et al., ; Barr, ). One popular idea (Horton & Keysar, ; Ferreira et al., ; Wardlow-Lane & Ferreira, ) is that it requires cognitive resources to keep track of what is in common ground, and language users might fail to notice potential ambiguities in their speech. Evidence for this view primarily comes from experiments in which the speaker’s knowledge differs from the listener’s, for example where the speaker can see a hidden object (a large triangle) and uses an unnecessary modifier when referring to an object (a small triangle) in common ground.
... Social cues; gestures The purpose of reference is to communicate: that is, for the speaker to get the addressee to identify the referent. As discussed in section .., the accessibility/prominence of discourse entities has sometimes been explained as increased attention to those entities. If attention is what drives reference form choices, other indicators of attention should matter. A prime candidate is physical gestures or eye gaze, which can signal attention to a physically co-present object. There is substantial evidence that listeners attend to the speaker’s eye gaze, and use it to help resolve ambiguous referring expressions. For example, participants in Hanna & Brennan’s () study were faster to resolve expressions like ‘the blue circle with five dots’ when the speaker gazed in the direction of the referent (see also Staudte et al., ). Goodrich Smith & Hudson Kam () showed that listeners could also follow gestures to positions in space that represented entities in the discourse, and use this to resolve ambiguous pronouns. In addition, - to -year-olds check the gaze of their interlocutor in an attempt to establish which of two possible objects she is referring to with a novel word (Nurmsoo & Bloom, ; see also Diesendruck, ; Grassmann et al., ). Younger still, - to -month-old infants check the gaze of a speaker more when a novel word is produced with two novel objects in view rather than one (Baldwin, ; Vaish et al., ). Nappa & Arnold () tested how the comprehension of ambiguous pronouns is influenced by gazing and pointing, and how these interact with linguistic contextual constraints. Participants in their experiment viewed a video of a woman at a table with two same-gender animal characters, one on either side, and an object in the centre. She told a story, for example This story is about Puppy and Panda Bear. Puppy is having some pizza with Panda Bear. He wants a pepperoni slice. The next screen posed a question, ‘Who wants a pepperoni slice?’, which they answered by pressing a button. In the neutral condition, people tended to pick Puppy about per cent of the time, exhibiting the well-known subject bias for pronouns. On top of this, gazing was a weak cue, either supporting the subject bias when the speaker gazed at Puppy, or leading listeners to respond at chance when the speaker gazed at the non-subject, Panda Bear. Pointing had a stronger influence, leading to a strong bias to choose the pointed-at character, regardless of the linguistic context.
These findings are consistent with the idea that referring is constrained by the attention of the discourse participants. However, Nappa & Arnold () demonstrated that pronoun comprehension is not driven by attention alone. In a second experiment, they manipulated a sudden-onset visual capture cue: a black square that appeared over one character as the speaker said the pronoun. If attention is the only thing that matters, irrelevant attentional capture should also influence pronoun interpretation. However, it did not. Arnold & Lao () found that the listeners’ egocentric attention could partially bias the listener toward one character, but only briefly, during online comprehension. Final interpretation in that task was primarily driven by public, discourse-relevant information— that is, the linguistic context itself. Together, these experiments suggest that discourse accessibility is related to attention, but it cannot be reduced to attention. Instead, pronoun interpretation is driven by evidence about the speaker’s intentions. Pointing provides a strong, intentional indicator of what the speaker is referring to, while eye gaze reflects the speaker’s attention, which is only partially related to the speaker’s intentions.
... Conceptual pacts Reference production and comprehension is influenced by the shared experience of particular speakers and addressees, for example when one person calls an object ‘the triangle shelf ’, and the name is adopted by other discourse participants. If a speaker has used a specific expression taking a particular perspective on the referent, and an addressee has responded correctly to that reference, it becomes subject to a ‘pact’ in that particular discourse situation (Brennan & Clark, , building on earlier work on lexical entrainment by Clark & Wilkes-Gibbs, ). Empirical work in this area has focused on the benefits of maintaining precedents with specific partners and the costs of breaking them (Metzing & Brennan, ; Van der Wege, ). For example, if an existing speaker uses a new expression, for example by switching from using ‘the shiny cylinder’ to ‘the silver pipe’, addressees are slower to look at the intended referent than when that new expression was produced by a new speaker, one with whom a pact had not been formed. Divergent theories have been proposed to account for this partner-specific language use: a socially grounded account, in which addressees meta-represent their interlocutor’s perspective and come to expect consistency in referring expressions (Barr & Keysar, ; Metzing & Brennan, ), and a domain-general episodic priming account, whereby speakers act as memory cues to situationally relevant information through associations with a particular expression (Horton & Gerrig, ; Horton & Slaten, ; see Kronmüller & Barr, , for a metaanalysis of referential pact research). The literature on referential pacts is helpful in explaining apparently overspecified referring expressions (e.g. ‘the shiny cylinder’ to refer to a lone cylinder). That is, expectations of speaker consistency may override expectations of minimal informativeness. Although the precise mechanism behind referential pacts is still being debated, the calculation of context and the production of an informative form may not be a pragmatic, Gricean process: instead there may be more automatic memory-based influences at play.
... Speaker goals Under linguistic theories of intentionality (Searle , ; Grice, ), speakers not only produce utterances but intend their utterances to have some effect, for example to bring
.
their addressee’s attention to a specific referent or aspect of a referent, to teach or persuade their addressee of something, or to end the exchange as quickly as possible. These discourse goals have been shown to affect referential choice. In a study manipulating speaker goals, overspecified referring expressions were common in a condition requiring a speaker to teach a hearer a long-term skill vs. when a hearer only had to execute the action once (Arts, ). Task-criticality also increases overspecification: more detailed referring expressions occurred in a task requiring participants to describe an object in a highimportance condition (long-distance medical surgery) than in a low-importance condition (straightforwardly describing an object; Arts et al., a). While these findings are perhaps intuitive (speakers give more information where precision is prioritized and the risk of misunderstanding is high), higher-level discourse features like speaker goals should be integrated into a comprehensive theory of reference production. That is, when it is critical that information must be communicated precisely, it is safer for the speaker to assume that that information is harder to grasp and thus use more explicit forms (cf. redundancy in Shannon & Weaver’s () information theoretic model). Such assessments may involve grounding considerations, such as community co-membership (Clark & Marshall, ). If an addressee belongs to the same community of say, fishmongers, a speaker may reduce the information contained in a referring expression to ‘the shucker’ from the alternative explicit expression ‘the stubby knife with the black handle’ since she can assume that her addressee can integrate critical specialist information about the best type of knife for opening oysters. In assessing and adjusting knowledge in this way, referring becomes more efficient (Isaacs & Clark, ). Parallels can be seen here with the less accessible/more explicit relationship pervasive in discourse-based models of reference.
.. Constraints from the referent A key part of the context in any referential situation is the referent itself, along with its referential competitors. While higher-level aspects of the discourse situation such as common ground have been extensively investigated, the role of referent features has received less attention. Needless to say, to achieve a comprehensive theory of reference, an understanding of each of the constraints on reference is required. Moreover, from a processing perspective, examining features of the referent itself can shed light on how speakers select the particular modifiers they eventually encode into their expressions. How do speakers identify the particular type of information that will be most informative when referring? For example, if you want to refer to a yellow stripy cup in the context of a plain yellow cup and a blue spotty cup, how would you do it? Informativeness accounts suggest that the optimal expression is ‘stripy cup’, though speakers frequently choose to encode colour too (Koolen et al., ; among others). In this section we discuss how (i) salience of features and (ii) display density might mediate discourse- or hearer-based informativeness expectations. According to discourse-based models of reference, the more conceptually accessible a referent is, the less explicit its informational form. For example, once a referent has been introduced into the discourse, a speaker is licensed to refer to it using a pronoun. However, outside of the linguistic context, there are instances in which a referent is highly accessible due to its perceptual salience and it is precisely that salience (e.g. its brightness, bigness, or redness) which leads speakers to encode the associated feature into their referring expression, even if this would strictly speaking be redundant given the context. For example,
speakers favour colour over other dimensions like size when referring (Pechmann, ; Dale & Reiter, ; Belke, ; Koolen et al., ). As a low-level feature, colour is salient, especially when it differs significantly from variation in the background (Vazquez et al., ). It is also salient due to the fact that it is an absolute rather than a relative property of a referent. Tarenskeen et al. () concluded that colour is more likely to be encoded in referring expressions due to its salience, in turn stemming from how important it is relative to the nominal class it is modifying (e.g. clothing vs. office supplies) and due to the paucity of other attributes in a referent (e.g. a coloured geometric figure). Speakers mention colour when it does not have any discriminatory power (Koolen et al., ), especially when its referent has an atypical colour (Westerbeek et al., ; Rubio-Fernández, ). Aside from colour, increasing the salience or accessibility of a referent’s property by duplicating that property in an array also leads speakers to include that modifier in their referring expressions (Carbary & Tanenhaus, ; Davies & Katsos, ; Koolen et al., ). Thus, increasing the perceptual salience of a referent or a property by rendering it colourful, making the colour more important to the referent itself, including an atypical property of the referent, or duplicating the property across other items in the display increases the likelihood of a speaker referring to it with a more explicit NP. This body of literature provides evidence that visual contrasts are not only easier to detect than linguistic contrasts (see also Ferreira et al., ), but that these contrasts are readily encoded into referring expressions, overriding Gricean predictions and generating overspecified forms. Another referent-based constraint affecting referential choice is display density. Speakers tend to overspecify more when there are two targets rather than a single target to find in a display (Koolen et al., , echoing Arnold & Griffin, , who found that pronouns were more common in a context with a single animate character than contexts with two animate characters). Further, they are more likely to redundantly modify when a visual scene is cluttered, compared to when this is not the case (Paraboni et al., ; Clarke et al., ; Koolen et al., ). A similar effect has been found with referents showing a larger number of attributes (Van Gompel et al., ), and when speakers are put under time pressure (Belke, ). Since it requires cognitive effort to calculate the distinctive features between the target and multiple distractors, speakers might bypass such a calculation and overspecify their referring expression without posing a serious threat to communication. In other words, under live communicative pressures, they may sacrifice Gricean felicity while maintaining the informativeness required for their addressee to identity the target. These effects have been explained from both speaker- and hearer-oriented perspectives. For example, under an efficiency-based analysis (Koolen et al., ), a speaker does not need to work out precise distinctive features between a target referent and the many distracters in a cluttered display and instead takes a shortcut by mentioning the salient property of colour as a preferred attribute. Rationally, colour may help reduce the search space by eliminating candidate referents of a colour other than that mentioned in the chosen referring expression, and even if it doesn’t, or does so only marginally, the speaker has not conceded too much of a cost. On the other hand, addressee-oriented accounts centre around speakers enabling hearers to profit from the pop-out effect that colour affords (Gatt et al., ). Finally, and following the view of reference as a collaborative process (Clark & Wilkes-Gibbs, ), speaker-hearer accounts have also been proposed, promoting the view that overspecification is efficient in communication for reasons shared by the speaker and the addressee: what is salient for the speaker is also salient for the addressee (Rubio-Fernández, ).
.
.. Pulling together contextual constraints on referring In sum, accounting for reference and its variation requires an understanding of how multiple constraints in the context impact not only referential choice, but also its mediating influences like attention. A wide range of non-linguistic pressures are at play in addition to the multitude of linguistic constraints comprehensively studied by scholars working in the discourse tradition. Each of the factors we have discussed are deemed important at a functional level because they relate to inferences about the goals, knowledge states, and perceptions of the discourse participants. In addition, they affect the production and comprehension processes necessary to achieve successful reference, because they affect fundamental memory and attentional processes that are relevant to language use.
. F
.................................................................................................................................. Identifying constraints on referential choice is a relatively straightforward task. Measuring their interaction and relative impact is a more complex undertaking. Moreover, although variables such as discourse accessibility or presence in common ground can predict a speaker’s choice of expression to a certain extent, their ultimate forms remain probabilistic. One challenge for models of reference production is the considerable variability present in speakers’ choice of referring expressions, both between and within speakers. Different speakers choose different referring expressions in telling even a simple vignette (Castro Ferreira et al., ; Zerkle & Arnold, ), and the same speaker is unlikely to be consistent in her word choices when telling the same story on different occasions. Although the factors influencing referential choice and interpretation have been studied extensively from various perspectives (e.g. discourse, psycholinguistic, developmental, computational), we do not yet have a comprehensive account of reference. In order to develop such an account, several outstanding questions must be addressed. Here we list a few. • Individual differences. Novel considerations such as individual cognitive constraints, for example processing speed and memory capacity, coupled with innovative methods, for example computational modelling (Hendriks, ) may help the field progress towards a fuller understanding of variability in reference. • Goals vs. processes. The research we have reviewed suggests that a theory of reference production must grapple with two overlapping questions: (1) how do contextual constraints affect the speaker’s goals, and (2) how do contextual constraints modulate the cognitive processes used to implement those goals? For example, an unusually coloured object (a purple banana) might change the communicative goals, increasing the speaker’s wish to note the colour. At the same time, the salience of the colour may attract the speaker’s attention and modulate discourse cues to accessibility. • Linguistic vs. non-linguistic sources of information. How do physical, linguistic, and interpersonal aspects of the context interact? Bringing together these constraints has the
potential to further advance our understanding of referential variability. For example, how might the effect of display density interact with recency of mention and speaker goals? • Examination of all referential decisions. When researchers examine reference production, they typically examine just one dimension of linguistic form choice, such as pronoun vs. more explicit forms, modified vs. unmodified expressions, or acoustically prominent vs. reduced expressions. Yet speakers must weigh multiple options at once for each referring event, suggesting that a full model must account for the set of all choices at once.
. C
.................................................................................................................................. In reviewing the literature from the informativeness and discourse-based traditions, this chapter has brought two highly related yet previously distinct theoretical approaches together. We have described their major assumptions and concerns, for example that referring expressions are expected to reduce uncertainty about the identity of referents, and have reviewed the methods and data that have informed their development. We have seen that both the informativeness and discourse-based approaches provide a systematic account of referential choice. Empirically, the Gricean approach has focused on levels of informativeness and of felicity across a variety of visual contexts, and finds that although speakers frequently produce informative and felicitous referring expressions, deviations from expected forms are driven by complex pragmatic processes recruiting multimodal information. Unexpected forms can also have efficient pragmatic effects on the hearer, for example contrastive inference and faster reference resolution. The discourse tradition also takes a functionalist view of reference production, asking which form types are appropriate under different discourse conditions. Empirically these studies have aimed to characterize the linguistic contexts that affect reference form. They also examine the relation between the linguistic context and referential form choice, as well as the relation between the context and psychological mechanisms like attention and common ground. We have examined the constraints on referential choice, exploring how aspects of context influence speakers’ means of referring. A wide range of constraints are highlighted: those coming from the extra-linguistic context (the interlocutors and the referents that make up a communicative situation), from the discourse itself, and from the cognitive representations of that discourse. In reviewing theory and data on reference, we aim to raise awareness of the research questions, assumptions, methods, and findings from the two established traditions of discourse and informativeness. We look forward to collaborations investigating interactions between the two, with the ultimate aim of building a comprehensive model of reference production.
A J. Arnold was partially supported by NSF grant #.
......................................................................................................................
......................................................................................................................
. I
.................................................................................................................................. T prosody of an utterance, which includes the grouping of the words of the utterance, the relative prominences of the words and the tonal movements, makes tremendous and diverse contributions to the meaning of the utterance. Prosody can convey the emotional status of the speaker, such as anger, disgust, fear or sadness (e.g. Brosch et al., ). Prosody distinguishes between speech acts: for instance, the American English declarative sentence It’s raining is a request for confirmation when uttered with a rise at the end, but an assertion when uttered with a fall at the end (e.g. Gunlogson, ). Prosody can also influence whether an interrogative sentence like Do you want to do the laundry? is understood as seeking information, as an invitation or as a request (the illocutionary act), and how annoyed, authoritative, and polite the speaker is taken to be (the perlocutionary effect) (Jeong & Potts ). Prosody can even influence the truth conditions of utterances. Whether, for example, the relative clause who was on the balcony in utterances of Someone shot the servant of the actress who was on the balcony is taken to modify the actress or the servant depends on the prosody of the utterance (e.g. Speer et al., ). Prosody also influences the truth conditions of utterances with focus-sensitive operators like only (e.g. Jacobs, ; Von Stechow, ; Rooth, ; Beaver & Clark, ). Thus, an utterance of Kim only serves Sandy Johnnie Walker in which Johnnie Walker is the prosodically most prominent expression is taken to mean that Kim serves Sandy nothing but Johnnie Walker, whereas an utterance in which Sandy is the prosodically most prominent expression is taken to mean that Kim serves nobody but Sandy Johnnie Walker. This chapter examines the contributions of prosody to meaning by considering experimental research on a category of meaning that has been central to research on prosody and meaning, namely information-structural focus. The study of information structure is concerned with the question of how the (morphological, syntactic, semantic, prosodic, etc.) properties of an utterance reflect the organization of the discourse in which the utterance occurs (see, e.g., Lambrecht, ; Krifka, ; Féry & Krifka, ). In this chapter, ‘focus’ refers to an information-structural property of parts of utterances and is defined on the basis of questions: an expression is a focus of an utterance if the content of that expression answers the (possibly implicit) question that the utterance addresses
(see, e.g., Roberts, a; Beaver & Clark, ). In (), for instance, where B’s utterance addresses A’s interrogative utterance, the proper name Mary answers the question denoted by the interrogative utterance and hence is the focus of B’s utterance. In (), the focus of B’s utterance is the indefinite noun phrase a tornado.1 In this chapter, brackets subscripted with ‘F’ ([]F) identify focused expressions. This notation should not be taken to convey any commitments about the syntactic representation of focus (but see, e.g., Rooth, ). To avoid confusion, the term ‘focus’ is not used in this chapter to refer to a phonetic or phonological property of focused expressions (e.g. ‘prosodic focus’). () A: Who saw the tornado? B: [Mary]F saw the tornado. ()
(adapted from Roberts, b: )
A: What did Mary see? B: Mary saw [a tornado]F.
Question-answer pairs like () and () are the main diagnostic for information-structural focus, regardless of whether focus is marked prosodically, morphologically or syntactically in a given language.2 As these examples illustrate, foci can be expressions whose denotations are given, like the proper name Mary in (), or new, like the indefinite noun phrase a tornado in (). Some works, such as Katz & Selkirk (), reserve the term ‘focus’ for contrastively focused expressions, to the exclusion of expressions providing new information, as discussed later. This chapter proceeds as follows. Section . provides a brief introduction to prosody and how experimental investigations can meet challenges to exploring the contributions of prosody to meaning. Section . examines how expressions that are information-structural foci are prosodically realized, and section . how prosodically realized foci are perceived and interpreted. Both sections . and . discuss data from a variety of languages, the influence of context, how focus size is distinguished and the role of experimental methods in exploring the prosody of focus. The chapter concludes in section . with a brief discussion of the contributions of prosody to semantic/ pragmatic phenomena related to focus, such as contrastive topic and presupposition. Additional linguistic phenomena influenced by focus are discussed by Kim (Chapter in this volume).
1 What is referred to as focus here is referred to as ‘rheme’ in other parts of the literature (e.g. Halliday, ; Roberts, b). Other authors define ‘focus’ on the basis of sets of alternatives (e.g. Rooth, ), givenness (e.g. Schwarzschild, ) or a combination of givenness and providing an answer to the Question Under Discussion (QUD) (e.g. Büring, ). Different definitions of ‘focus’ are discussed in Kim’s Chapter , in this volume. 2 Some languages have been reported to not use prosody to mark information-structural focus, including Wolof (Rialland & Robert, ), Yukatek (Kügler & Skopeteas, , ), Chichewa, Chitumbuka, and Durban (Downing, ), NłeɁkepmxcin (Koch, ), Northern Sotho (Zerbian, ), and K’iche’ (Yasavul, ; Burdin et al., ). For discussion of prosodic, morphological, and syntactic marking of focus see, for example, Féry & Krifka, , and Zimmermann & Onea, . For the hypothesis that information-structural focus is marked prosodically in all languages, see Roberts, .
. P: A
.................................................................................................................................. This section is a brief introduction to prosody and to challenges to exploring prosody and meaning. For more detailed introductions to prosody from different perspectives see, for example, Pierrehumbert & Hirschberg, ; Hirst & Di Cristo, : chapter ; Gussenhoven, ; Grice & Baumann, ; Ladd, ; and Beckman & Venditti, . Any utterance has phonetic properties that can vary independently of the expressions uttered: for instance, in English, the sentence Mary saw a tornado can be uttered with prosodic prominence on tornado without changing the contribution the expression makes to the truth conditions of the sentence. Prosodic prominence on tornado can therefore be used to convey (what is referred to as) a post-lexical meaning. Crosslinguistically, the phonetic properties of utterances that can vary independently of the expressions uttered in order to convey post-lexical meanings include fundamental frequency, duration (of syllables or pauses) and intensity. Languages differ in whether a particular phonetic property is used to convey post-lexical meanings: for instance, a rise in fundamental frequency that is associated with the stressed syllable of an expression is a correlate of a post-lexical (high tone) pitch accent in English (e.g. tornado means ‘tornado’ regardless of whether it is realized with a high or a low tone pitch accent), but such a rise may be a correlate of a lexical pitch accent in Swedish (e.g. anden means ‘the duck’ or ‘the spirit’, depending on the pitch accent with which it is realized) or of tone in Mandarin Chinese (where, famously, ma means ‘mother’ or ‘hemp’ depending on whether it is realized with a high tone or a rising tone, respectively). Languages also differ in the phonetic properties that have become conventionalized bearers of post-lexical meaning and are thus part of the phonology of the language. The phonetic and phonological properties of a language that are used to convey post-lexical meaning constitute the prosody of the language. The prosody of a language can provide cues about the organization of the utterance and its relation to the context in which it occurs: these cues include phrasing (how the expressions of the utterance are grouped), prosodic prominence (the relative prominence of the expressions in the utterance), and tonal movements (the tune of the utterance). Examples of such information cued by prosody include (see section .) the attitude of the speaker, the speech act of the utterance, the syntactic and semantic relationships among the expressions of the utterance, and the information structure of the utterance. In English, for instance, whether the relative clause who was on the balcony modifies the servant or the actress can be indicated by marking either one of the two noun phrases as the right edge of a phrase, by a combination of lengthening and hyperarticulation of segments preceding the edge, pitch movements associated with the edge and a pause (for detailed discussion see, e.g., Wagner & Watson, ; and Speer et al., ). Thus, the study of prosody (or ‘intonation’, as it is also sometimes called) is not concerned with phonetic or phonological properties of an utterance that determine the lexical meanings of the expressions uttered, but with those properties that convey other types of meaning, under a broad use of the term ‘meaning’. For an introduction to the autosegmental-metrical framework of intonational phonology assumed by many of the works reviewed in this chapter see, for example, Beckman, ; Shattuck-Hufnagel & Turk, ; and Ladd, .
Given the contributions that prosody makes to natural language meaning, theories of meaning must capture how speakers use prosody to convey meaning and how listeners perceive and interpret the prosody of the utterances they hear. Studying the contributions that prosody makes to meaning is challenging for a number of reasons. First, since not all phonetic and phonological properties of an utterance contribute to post-lexical meaning, research on prosody must tease apart those that do from those that don’t. This task is complicated by the fact that the prosody of any language is complex, with multiple components, some of which have continuous expression in the speech signal (see, e.g., Wagner & Watson, , Watson, , on prominence in production and comprehension). Second, since the vocal folds and speech patterns of the speakers of a language vary along a number of dimensions, the phonetic properties that underlie the prosody of the language do not have any absolute value for fundamental frequency, duration or intensity in the language, but vary by age, sex, etc. (e.g. men’s speech is typically characterized by lower fundamental frequency values than women’s). There are thus no invariant phonetic correlates of information-structural notions such as focus (see, e.g., Féry, ) and even the prosody of utterances that convey the same meaning is highly variable, both within and across speakers of the same language (e.g. Baumann et al., ; Ito & Speer, ; Speer et al., ). In short, the mapping between the prosody of an utterance in any given language and the meaning of the utterance is not straightforward. Furthermore, there is to date no agreement about the nature of the interface between prosody and meaning (for discussion see, e.g., Prieto, ). Open questions include, for example, whether the interface is direct, or mediated by syntax (for a probabilistic perspective, see, e.g., Calhoun, ), and whether individual tonal movements or combinations thereof are the basic meaning bearing units of prosody (see, e.g., Gussenhoven, , for discussion and relevant references). And, finally, as noted earlier, there is cross-linguistic variation in the prosodic systems of languages, even closely related ones (see, e.g., Hirst & Di Cristo, ; Jun, , ). Experimental investigations can help meet these challenges to studying the contributions of prosody to meaning. Perception and comprehension experiments can, for instance, help tease apart those phonetic and phonological properties of utterances that influence postlexical meaning from those that don’t. Production experiments designed to explore the prosodic realization of utterances with particular meanings can serve to identify speakerinvariant prosodic cues to those meanings. Finally, cross-linguistic comparison is facilitated by running the same experiments with speakers of different languages. This chapter illustrates advances made by experimental investigations into the contributions of prosody to meaning, in particular the information structural notion of focus.
. T
.................................................................................................................................. This section is concerned with the production of prosodically realized focus. Section .. introduces prosodic cues to focus across languages, and argues that the prosody of focus in a given language may be independent of the typological characteristics of the prosodic system of the language. Section .. shows that contextual information about focus
influences the prosodic realization of focus, which has consequences for the question of whether the grammars of languages distinguish different types of focus, as discussed in section ... The idea that differences in focus size are prosodically distinguished is motivated in section ... Section .. concludes with a discussion of methods used in research that investigates how focus is prosodically realized.
.. Prosodic realization of information-structural focus across languages In languages that mark information-structural focus prosodically, focused expressions are generally more prosodically prominent than expressions that are not focused. Languages differ in the phonetic and phonological means to implement prosodic prominence. These means include the presence and type of pitch accents and boundary tones, duration of segments and pauses, fundamental frequency range, and the alignment of pitch targets to the segmental string (e.g. Jun, ; Ladd, ; Watson, ; Féry, ). In English, for instance, focused expressions are made relatively more prosodically prominent than nonfocused expressions in the same phrase by realizing the focused expression with a pitch accent, a longer duration, an expanded pitch range and greater intensity (e.g. Cooper et al., ; Eady & Cooper, ; Eady et al., ; Terken & Hirschberg, ; Xu & Xu, ; Breen et al., ). In Jun’s () typology of prosody, English is a head-prominence language since prosodic phrase-level prominence is marked on the phrase head, which is identified as the head of the phrase through pitch accenting. In Paraguayan Guaraní (Tupí-Guaraní), also a head-prominence language, focused expressions are distinguished from non-focused ones by the type of pitch accent realized, by the steepness of the slope of rising pitch accents and by duration (Clopper & Tonhauser, , ; Burdin et al., ; Turnbull et al., ). Korean, unlike English and Paraguayan Guaraní, is an edge-prominence language, which means that phrase-level prosodic prominence is marked on the left or right edge of the prosodic phrase through boundary tones (the left edge, in the case of Korean; see Jun, ; Kim, ). A Korean utterance is typically phrased into accentual phrases and (larger) intonation phrases. An accentual phrase, which is slightly larger than a word, is marked by a particular tonal pattern. A focused expression is made relatively more prosodically prominent than non-focused expressions by expanding its pitch range, lengthening its initial segment, and by reducing the pitch range on the following expressions and the duration of the expression before and after the focused expression. Optionally, a boundary tone may be inserted before the focused expression and the following words may be dephrased; that is, realized in the same phrase as the focused word (e.g. Jun, ; Jun & Lee, ; Jun & Kim, ). Jun (: ) proposed that the way in which information structure, including focus, is prosodically realized in a given language may depend to some extent on the prosodic structure of the language. This proposal was investigated on the basis of a study of the prosody of noun, adjective and noun phrase focus in American English, Paraguayan Guaraní, K’iche’ (Mayan), and Moroccan Arabic in Burdin et al. (). In contrast to American English and Paraguayan Guaraní, which are head-prominence languages,
Moroccan Arabic and K’iche’ were described in Burdin et al. () as head/edgeprominence languages, which means that they exhibit both head-prominence and edgeprominence marking (see Jun, ). Burdin and her colleagues found that, contrary to Jun’s proposal, ‘the prosodic realization of focus is orthogonal to the head-, head/edge-, and edge-prominence distinction’ (p. ): for instance, focused expressions were lengthened in American English, Paraguayan Guaraní, and Moroccan Arabic, but not in K’iche’. (See Burdin et al., : section ., for discussion of other literature that also suggests this conclusion.) In sum, languages that mark information-structural focus prosodically rely on a range of phonetic and phonological cues, and which cues a language relies on may be independent of the typological characteristics of the language.
.. The prosodic realization of focus is context-dependent Question-answer pairs are fundamental in research on information-structural focus: since interrogative utterances determine the foci of their answers, question-answer pairs allow researchers to explore how foci are (prosodically, morphologically, or syntactically) realized. It has long been observed that answers must be congruent with the interrogative utterances they address (e.g. Paul, , ; Von Stechow, ; Rooth, ), as shown on the basis of the question-answer pairs in (): J’s utterance in (a), with Turkish prosodically prominent (as indicated by the capital letters), is congruent with M’s question in (a), hence judged to be acceptable in response to that question, but not congruent with M’s question in (b), hence judged to be unacceptable in response to that question. The reverse pattern holds for J’s utterance in (b), in which David is prosodically prominent. ()
Mandy, Craige, and Judith are having lunch at a place that serves Turkish, Lebanese, and Irish coffee. a. M: What kind of coffee does David like? J: David likes [TURkish]F coffee. b. M: Who likes Turkish coffee? J: [DAvid]F likes Turkish coffee. (adapted from Tonhauser, : )
It follows from question-answer congruence that the focus marking of an utterance constrains the question addressed by the utterance, regardless of whether the question is realized by an interrogative utterance, as in the examples in (), or whether the question is implicit, as in many naturally occurring examples (see, e.g., Roberts, a; Beaver et al., ). The fact that what is focused in an utterance can be conveyed both by context (e.g. the interrogative utterances in ()) and by prosody suggests the information-theoretic hypothesis that the prosodic realization of focus is enhanced when the context does not identify which expression is focused. This hypothesis was investigated by Turnbull et al. (): in their production experiments, speakers of American English and Paraguayan Guaraní played an interactive game in which they instructed a confederate (of the same language) to place a series of tiles depicting different objects into numbered boxes on a
game board. The speakers’ utterances were either made in a context in which it was predictable from the other game pieces (visible to the speaker and the confederate) whether the adjective, the noun, or the entire noun phrase was focused, or in a context in which the focus was unpredictable from the other game pieces. For instance, when the English utterance Put the orange flower in box two was made in a context in which the other game pieces were flowers of a variety of colours, adjective focus was predictable from the visual context; when the same utterance was made in a context in which the other game pieces consisted of a variety of objects, including flowers, in a variety of colours, adjective focus was not predictable from the visual context. Turnbull and his colleagues found that the prosodic realization of focus was enhanced in both languages when the visual context did not predict which expression was focused. For instance, when adjective focus was not predictable from context, focused English adjectives were longer, realized with higher f peaks and more likely to be realized with a rising pitch accent than when adjective focus was predictable from context. In Paraguayan Guaraní, whether focus was predictable from context did not affect pitch-accent type or duration, but the f slope of accented adjectives was steeper (regardless of whether the adjective was focused) when focus was not predictable from context than when it was. Thus, the prosodic realization of focus is affected in both languages by the contextual predictability of focus, but the influence of contextual predictability on prosodic realization of focus appears to be language-specific. Context-dependence in the prosodic realization of focus was also illustrated in Beaver et al.’s () study of second occurrence focus, a repeated focus: in B’s utterance in (), vegetables associates with the focus-sensitive operator only, but vegetables is given. Second occurrence focus is marked with brackets subscripted with ‘SOF’. () A: Everyone already knew that Mary only eats [vegetables]F. B: If even [Paul]F knew that Mary only eats [vegetables]SOF, then he should have suggested adifferent restaurant. (Beaver et al., : ) Beaver et al. () found that second occurrence foci are not realized with a pitch accent, but distinguished from unfocused expressions by a longer duration and a higher energy. Thus, whether a focus is realized in American English with a pitch accent depends on the context in which the focus occurs. In sum, the prosodic realization of focus in a given language depends on the contextual cues to focus and the context in which the focused expression is realized.
.. Prosodic realization of other information-structural properties of foci Focused expressions denote answers to the (explicit or implicit) questions addressed by the utterances in which they occur, but they may differ along other information-structural dimensions. In the examples in (), for instance, the noun phrase a coffee is the focus of A’s utterances, but the denotation of the noun phrase differs along other informationstructural dimensions: the denotation is new information in (a), it is previously mentioned
and contrasted with the other member of a contextually-given set, namely tea, in (b), and it is used to correct an interlocutor’s assumption in (c). ()
a. Q: What would you like to drink? A: I’ll have [a coffee]F, please. b. Q: Would you like to have a coffee or a tea? A: I’ll have [a coffee]F, please. c. At the breakfast buffet, A may have coffee or tea, and has just been given a cup of tea. A: I don’t want tea, I want [a coffee]F.
Given that focus is not the only meaning conveyed prosodically (see section .), it is unsurprising that foci that differ in their information-structural properties may differ in their prosodic realization. The three studies discussed in the following illustrate this for foci that differ along the lines of the examples in (). In Ito et al. (), American English-speaking participants gave instructions to a confederate about how to decorate a tree, thereby producing sequences of utterances consisting of a colour adjective and an object-denoting noun, such as . . . blue ball . . . blue house. The noun of an adjective-noun expression was defined as contrastive if it was immediately preceded by an adjective-noun expression with the same adjective but a different noun, like house in . . . blue ball . . . blue house (likewise for adjectives). They found that adjectives and nouns whose denotations were contrastive (whether new or given) were more likely to be realized with a L+H* pitch accent3 than adjectives and nouns whose denotations were (non-contrastive and) new. In Breen et al.’s () experiments, speakers uttered answers and listeners had to identify the question they were answering: for example, an utterance of DAmon fried an omelet this morning answers Who fried an omelet this morning? but not What did Damon fry this morning? (recall that capital letters indicate prosodic prominence). An expression was defined as contrastive in the presence of an explicit contrast set: while Damon in the justmentioned congruent question-answer pair is not contrastive, it is taken to be contrastive when DAmon fried an omelet this morning is uttered in response to Did Harry fry an omelet this morning?. In an experiment in which speakers were given feedback about whether listeners were successful in identifying the question they were answering, speakers’ utterances distinguished between contrastive and non-contrastive subjects, verbs, and objects, with contrastive expressions realized with longer durations and higher intensity than (given or new) noncontrastive ones. Surprisingly, non-contrastively focused words were produced with higher f than contrastively focused ones, a finding which may be in contrast to that of Ito et al. (), given that L+H* pitch accents are typically realized with a higher f than H* pitch accents (for discussion and references, see, e.g., Ito & Speer, ). Katz & Selkirk () also found differences in the prosodic realization of contrastive and (non-contrastive) discourse-new expressions. Their production experiment compared the realization of expressions like that Modigliani and MoMA in utterances like So he would (only) offer that Modigliani to MoMA in which both expressions were discourse-new, or 3
The L+H* pitch accent is a high tone pitch accent that is preceded by a low tone. For an introduction to the Tones and Break Indices (ToBI) annotation system, see Beckman & Ayers ().
only one was discourse-new and the other one was contrastive. For Katz & Selkirk (), an expression was contrastive if it associated with a focus-sensitive expression like only. They found that contrastively focused expressions did not differ from discourse-new ones in terms of pitch accenting or phonological phrasing, but that contrastively focused constituents had a longer duration, a greater relative intensity and a greater f movement than discourse-new ones in the same position. (For relevant studies in other languages, see, e.g., Krahmer & Swerts, , on Dutch; D’Imperio, , on Neapolitan Italian; and Baumann et al., , and Braun, , on German.) These three studies suggest that prosody does not merely distinguish focused and nonfocused expressions (section ..), but also distinguishes whether a focused expression denotes new or contrastive information. It is important to note, however, that the term ‘contrastive’ was defined differently in the three studies reviewed here: its definition involved an expression in the immediately preceding utterance (Ito et al., ), a question that the utterance was correcting (Breen et al., ), and association with a focus-sensitive expression (Katz & Selkirk, ). The fact that the studies explored three different (though possibly related) notions of ‘contrastive’ focus may very well account for the differences observed in the three studies in how prosody distinguishes contrastively and noncontrastively focused expressions. These studies thus highlight the importance of unambiguously defining the meaning categories whose prosodic realizations are explored. (For discussion, see also Kim’s Chapter on focus, in this volume.) The findings of the three studies also bear on the question of whether foci that differ in other information-structural properties are grammatically distinguished. Katz & Selkirk (), for instance, argued that contrastive foci and discourse-new foci are represented differently in the syntax and thereby receive distinct interpretations. (For related proposals see, e.g., Kiss, ; Vallduví & Vilkuna, ; Selkirk, .) An alternative position (which is compatible with foci being syntactically represented or not) is one on which differences between foci are attributable to context and hence need not be grammatically represented (see, e.g., Rooth, ; Roberts, ; Büring, ).
.. Prosodic marking of focus size This section shows that prosody provides cues to the size of the focus of a given utterance. B’s utterances of Kim baked a cake in (a‒c), for instance, differ in whether the focus is the direct object, the verb phrase, or the entire sentence, respectively. ()
a. A: What did Kim bake today? B: Kim baked [a cake]F. b. A: What did Kim do today? B: Kim [baked a cake]F. c. A: What happened? B: [Kim baked a cake]F.
Since B’s utterances in (a‒c) can all be realized with pitch accents on cake, and no other pitch accents, some authors assume that prosodic marking of focus is systematically
ambiguous with respect to whether the direct object, the verb phrase, or the entire clause is focused (e.g. Gussenhoven, ; Selkirk, ). Accordingly, Selkirk’s () widely adopted Focus Projection Hypothesis predicts that a pitch accented direct object can ‘project’ focus from the direct object to the verb phrase and to the entire utterance. (See Ladd, : chapter , for a discussion of focus projection.) However, experimental research on the prosodic realization of focus has provided empirical evidence that the size of focus can be prosodically distinguished. For American English, Breen et al. () found that speakers produce subject-verb-object sentences differently depending on whether the entire sentence or only the object is focused. Specifically, they found that when only the object is focused, the object is produced with the highest maximum f, the longest duration and the maximum intensity compared to other words in the sentence, whereas the words of the sentence were more prosodically similar to one another when the entire sentence was focused. Baumann et al. () investigated the prosodic realization of German sentences with sentence focus, verb phrase focus, and object focus, and found that the narrower the focus was, the more likely the direct object was to be produced with a longer duration and with a higher f peak relative to previous f peaks, and the less likely the verb was to be realized with an accent. For European Portuguese, Frota () found that the object is realized with a longer duration and a different pitch accent when only the object is focused than when the entire sentence is focused. Similarly, Hanssen et al. () showed that, in Dutch, sentence and object focus are distinguished by the duration of the object and by the timing, scaling, and slope of f targets and movements. Focus size is also distinguished in languages genetically unrelated to European ones. For Korean, for instance, a verb-final language, Kim et al. () found that verb phrase focus was distinguished from sentence focus by inserting an intonational phrase boundary, by raising the pitch peak of the first word of the verb phrase, and by lengthening the first and second words in the verb phrase. Jun & Kim () observed that the object was relatively more prominent in object focus than in verb phrase focus: in object focus, the f peak of the object was higher and the object was longer than in verb phrase focus, and the verb phrase was more likely to be dephrased or produced with a reduced pitch range in object than in verb phrase focus. Finally, focus size is also marked within the noun phrase. As mentioned earlier, Burdin et al. () compared the prosodic realization of focus in utterances in which the adjective, the noun or the entire noun phrase was focused across four genetically unrelated languages. For American English, they found that noun focus is distinguished from noun phrase focus in that the adjective is less likely to be unaccented when the noun phrase is focused than when the noun is focused. (For a similar finding in Dutch, see Krahmer & Swerts, .) For Paraguayan Guaraní, in which the noun precedes the adjective, they found that adjectives were longer when the noun phrase was focused than when the adjective was focused. In Moroccan Arabic, the noun again precedes the adjective. Here, Burdin and her colleagues found that both the adjective and the noun were longer when the noun phrase was focused than when the adjective was focused. (K’iche’ did not show any prosodic marking of focus; see footnote .) In sum, genetically and typologically unrelated languages use a variety of prosodic means to lend prominence to particular parts of an utterance and to thereby identify the size of the focus of the utterance. These findings suggest that the assumption that underlies Selkirk’s
Focus Projection Hypothesis, that prosodic focus marking does not distinguish focus size, may not be empirically adequate, at least from the production perspective. Section .. addresses whether listeners attend to prosodic cues to focus size.
.. Experimental tasks to explore the prosodic realization of focus Experimental research on the prosodic realization of focus relies on a variety of tasks: recordings are made of individual participants reading scripts (e.g. Beaver et al., ; Katz & Selkirk, ), of pairs of participants reading scripts to one another (e.g. Clopper & Tonhauser, ), of participants completing a task with a confederate (e.g. Ito & Speer, ; Burdin et al., ) and of pairs of participants engaging in an interactive task (e.g. Watson et al., ; Breen et al., ). Compared to empirical claims about the prosodic realization of focus that are based on individual researchers’ intuitions, experimental investigations have the advantage of allowing researchers to quantify phonetic and phonological cues to focus, and to distinguish between speaker-invariant and speaker-specific prosodic cues to focus. The aforementioned tasks differ significantly in the extent to which the elicited utterances resemble conversational speech, with participants reading scripts being the least and participants engaging in interactive tasks being the most like conversational speech. An advantage of scripted language is that researchers have full control over the content, length, and structure of the produced utterances, thus facilitating the comparison of productions of pairs of utterances that differ only in the variables of interest. However, as discussed in Ito & Speer (), one of the worries of investigating prosody on the basis of productions of read, scripted language is that since such productions differ from conversational speech, the results of the investigation have limited generalizability. Ito & Speer therefore urge researchers to rely on tasks that examine ‘the relationship between produced intonation and the linguistic information conveyed in a discourse . . . without changing the nature of the speech people typically produce in conversation’ (p. ). See the references earlier in this section for examples of such tasks.
. T ..................................................................................................................................
This section examines how listeners perceive and interpret prosodically realized focus.4 Section .. examines the perception and interpretation of prosodically realized focus 4
The research reviewed in this section is limited to experimental research, to the exclusion of works that analyse the meaning of (mostly English) prosody by abstracting over findings in the experimental literature (e.g. Pierrehumbert & Hirschberg, ; Steedman, , , ; and Truckenbrodt, ). Specifically, the research reviewed in this section is limited to works based on perception tasks, which ask participants to make a judgement about the signal (e.g. prominence ratings), and interpretation tasks,
across languages, and .. shows that the perception and interpretation of prosodically realized focus is context-dependent. Section .. illustrates that information-structural differences between foci are perceived and interpreted, and section .. addresses the perception and interpretation of focus size. Section .. concludes with a discussion of methods.
.. The perception and interpretation of focus across languages Not only do speakers use prosody to signal what is focused in an utterance (section ..), listeners also attend to prosodic cues to focus in their interpretation of utterances. In an early study on American English, Most & Saltz () found that listeners rely on the location of pitch accents in utterances of sentences like The pitcher threw the ball in identifying whether the subject or the object is focused. In one of their experiments, participants listened to utterances of such sentences in which either the subject or the object was prosodically prominent and were asked to write a question to which the utterance would be an appropriate answer. Most & Saltz found that the expression that was prominent in the answer utterance corresponded, in per cent of cases, to the expression that was focused in the utterance, given the question that the participant had provided. For instance, the question Who threw the ball? was written for the utterance The pitcher threw the ball in which the subject was prominent. Birch & Clifton () also relied on question-answer congruence to explore whether listeners attend to prosody in identifying foci. Asking native speakers of American English to judge the appropriateness of congruent and incongruent questionanswer pairs, they found that listeners rated utterances in which a focused noun phrase was pitch-accented as more appropriate than utterances in which a focused noun phrase was not pitch-accented. Breen et al. () also found that listeners in their Experiment (where speakers and listeners received feedback about whether the listeners’ response was correct) were able to determine whether the subject, verb, or object was focused. In sum, these studies provide empirical evidence that listeners attend to prosodic cues in identifying which expression of an utterance is focused.5 Clopper & Tonhauser () investigated whether Paraguayan Guaraní listeners attended to prosodic cues to focus, as well as which prosodic cues they attended to. As mentioned in section .., Clopper & Tonhauser identified, on the basis of production experiments, that focused expressions in Guaraní are more likely than non-focused expressions to be realized with a rising pitch accent, with a longer duration and a steeper slope for rising pitch accents. The question-answer pairs collected in one of the production experiments served as stimuli for an interpretation and a perception task. In the interpretation task, Guaraní listeners were asked to identify which of two lexically identical answers which ask participants to make a judgement about meaning (e.g. whether an utterance is congruent with a question). Accent placement facilitates listeners’ utterance comprehension (e.g. Bock & Mazzella, ; Terken & Nooteboom, ; Ayers, ) and information about focus from accent placement is processed rapidly online (e.g. Dahan et al., ). For a more thorough discussion of the influence of prosodic prominence on processing and memory, see Kim’s Chapter in this volume. 5
that differed in their prosody was the preferred response to a question. They found that listeners attended to the pitch-accent type and the slope of rising pitch accents in identifying which expression in the utterance was focused. However, they also found that Guaraní listeners’ performance was significantly lower than chance on this task: that is, listeners were not able to reliably use prosodic cues to identify which expression in an utterance was focused, contrary to the findings of Breen et al. (). As discussed in Clopper & Tonhauser (: section .), this result may at least partially be attributed to the task design (listeners did not receive feedback about whether their answers were correct) and to the fact that the materials included highly variable natural Guaraní utterances that included incongruous prosodic cues to focus. In the prominence perception task, Clopper & Tonhauser asked Guaraní listeners to identify which word of the two-word answer utterances was prosodically more prominent. They found that listeners attended to the duration of the stressed syllable in identifying which expression was perceived as prosodically more prominent. As in the interpretation task, performance was not significantly above chance: listeners judged the focused expression to be more prosodically prominent than the non-focused one in per cent of cases. As discussed in Clopper & Tonhauser (: section .), the chance performance is consistent with research examining the perception of prosodic prominence by naïve listeners in other languages (e.g. Streefkerk et al., ; Mo et al., ). Taken together, the results from interpretation and perception tasks suggest that prosodic cues to focus constrain but do not determine what is focused in an utterance. In sum, there is evidence from genetically unrelated languages that listeners attend to prosodic cues in identifying what is focused in an utterance. Perception and interpretation experiments can also serve to identify which prosodic cues to focus listeners attend to, as well as how successful listeners are in determining which expression is focused from prosodic ones. Whether languages differ in the strength of their prosodic cues to focus is an open question.
.. Contextual influence on the perception and interpretation of prosodically realized focus Context not only has an effect on the production of prosodically realized focus (section ..), but also on its perception and interpretation. Bishop (), for instance, showed that American English listeners’ perception of prosodically realized focus is influenced by the question that the utterance addresses. In his study, participants listened to questionanswer pairs, like the three shown in (), and were asked to rate how ‘stressed’ each word in the answer utterances sounded. Crucially, the same recording was used for the answer in each question-answer pair, namely a recording of the answer with verb phrase focus. ()
A: What happened yesterday? / What did you do yesterday? / What did you buy yesterday? B: I bought a motorcycle.
Bishop found that the objects of the answer utterances were rated as more stressed when the answers were heard in response to an object focus question than in response to a verb
phrase or sentence focus question. Thus, ‘listeners’ judgements of prosodic prominence are significantly and independently affected by their interpretation of an utterance’s information structure’ (p. ): that is to say, by what is contextually identified as the focus of the utterance. Similarly, Turnbull et al. () found that American English expressions that were realized with L+H* pitch accents were more likely to be perceived as prosodically prominent when listeners were explicitly made aware of the contrastive context in which the utterances were made. The perception of prosodic prominence is also influenced by the prosodic context. In Krahmer & Swerts’s () study, Dutch listeners were asked to distinguish between words realized with (what the authors refer to as) contrastive and new accents. When presented with expressions realized with such accents in their sentential context, expressions with contrastive accents were rated as more prominent than expressions with new accents, but this difference in prominence rating was not observed when the expressions were presented in isolation. Thus, listeners attended to prosodic cues to the information-structural status of the expressions, but these cues were contextually enhanced. Context also influences the interpretation of prosodically realized focus, as shown by Kurumada et al. (). Their study investigated the interpretation of prosodically realized focus using sentences of the form It looks like a . . . , such as It looks like a zebra. Utterances of these sentences differ in their interpretation depending on whether the verb looks or the noun (e.g. zebra) is the focus: while the utterances typically favour an affirmative interpretation (‘it’s a zebra’) regardless of which expression is focused, the negative interpretation (‘it’s not a zebra’) is more likely to arise when the verb is focused than when the noun is focused. In one experiment, Kurumada et al. () showed that when listeners were made aware that speakers can also produce It’s a zebra, they were more likely to assign negative interpretations to It looks like a zebra, regardless of focus placement. This finding suggests that contextual information about the speaker’s intended meaning can weaken the contribution of prosodically realized focus to meaning. Another experiment reported in Kurumada et al. () suggests that information about a speaker’s use of prosody to convey meaning influences the extent to which listeners rely on prosodically realized focus in deriving the speaker’s intended meaning. In conclusion, the studies reviewed earlier show that the perception and interpretation of focus is context-dependent—at least in English and other well-studied European languages.
.. Interpreting prosodic cues to foci that differ in other information-structural properties The studies reviewed in this section show that listeners attend not just to prosodic cues to focus (section ..) but also to prosodic cues to other information-structural properties of foci. Breen et al. (), for instance, found that a few American English listeners were able to rely on prosodic cues to distinguish whether an utterance like Damon fried an omelet this morning was made in response to a question like Who fried an omelet this morning? (in which Damon is new information) or to a question like Did Harry fry an omelet this morning? (in which Damon is corrective and contrasted with Harry). They found that more listeners were able to identify which question the utterance was responding to when the
utterance was preceded by I heard that . . . , which suggests that ‘speakers are able to convey contrastiveness using words outside of the clause containing the contrastively focused element’ (p. ). Many studies have compared the interpretation of high tone pitch accents (H* in the ToBI annotation system, see footnote ) and L+H* accents. In one of Ito & Speer’s () eye-tracking experiments, listeners heard sequences like . . . green drum . . . blue drum and were found to fixate on the relevant objects (e.g. drums) earlier when blue was realized with a L+H* accent than with a H* accent. This finding suggests that listeners are more likely to interpret adjectives as contrastively focused when they are realized with a L+H* accent than when they are realized with a H* pitch accent. Likewise, Gotzner & Spalek () showed, using a truth-value judgement task, that German expressions were more likely to generate exhaustive inferences when realized with L+H* rather than H* accents. The question of whether H* marks foci whose denotation is new (as claimed in, e.g., Pierrehumbert & Hirschberg, ), was addressed in Watson et al.’s () eye-tracking study. While they did find that American English listeners were more likely to fixate on a contrastive referent than on a new one when they heard the L+H* accent, listeners were not more likely to fixate on a new referent than a contrastive one when they heard the H* accent (an expression was taken to be ‘contrastive’ if it contrasted with the denotation of an expression in an immediately prior utterance, similarly to Ito & Speer, ). Watson et al. () took these findings to show that L+H* conveys a contrastive meaning (in support of Pierrehumbert & Hirschberg, ) but that H* is compatible with new and contrastive foci. Although the aforementioned studies suggest that listeners interpret H* and L+H* differently, other research suggests that the two pitch accents do not correspond to two distinct perceptual categories. Bartels & Kingston (), for instance, synthesized a continuum of stimuli between H* and L+H* by varying the acoustic characteristics of the accent (peak height, peak alignment, peak timing, depth of the dip preceding the peak) and the context in which the sentences were presented. They asked American English listeners to rate whether the relevant expression received a contrastive or a non-contrastive interpretation. (Here, contrastiveness was defined as the explicit denial of a salient alternative, like apple in Amanda had a banana.—No, she had an apple.) Bartels & Kingston did not find clear evidence for a categorical boundary between L+H* and H*. What they did find was that, the higher the peak height was, the more likely a contrastive interpretation was. (But see Calhoun, , for a different finding.) The depth of the dip preceding the peak and the timing of the peak were secondary cues to a contrastive interpretation. Ladd & Morton () found that listeners categorically distinguished the two accents in an identification task but they did not identify a category boundary in discrimination tasks, which they took to mean that the two accents are categorically interpreted, but continuously perceived.
.. Perceiving and interpreting the size of prosodically realized focus Listeners can perceive and interpret prosodic cues to differences in focus size. D’Imperio (), for instance, found, using a prominence perception task, that Neapolitan Italian
listeners perceive utterances of subject-verb-object declarative sentences with sentence focus differently from utterances with object focus: listeners identified the object as most prominent in only per cent of sentence focus utterances but in per cent of object focus utterances. Can listeners interpret prosodic cues to differences in focus size? Gussenhoven () answered this question negatively. He found that participants were at chance in distinguishing whether an utterance of a subject-verb-object sentence, like We repair radios, that had been made in response to a question that elicited object focus (What is it you repair?) responded to that question or to a question about the verb phrase (What is the nature of your business?). Breen et al. (), on the other hand, found that American English listeners can distinguish between object and sentence focus: listeners in their Experiment were able to correctly identify utterances that were produced with (non-contrastive) object focus per cent of the time, perceiving such utterances as having sentence focus only per cent of the time. Similarly to Gussenhoven’s () task, listeners in their study had to identify which question an utterance responded to, but in contrast to Gussenhoven’s task, listeners in their study were seated in the same room as the speakers, and both the speakers and the listeners received feedback when a listener chose the wrong question: ‘This cue was included to ensure that speakers knew when their productions did not contain enough information for the listener to choose the correct answer’ (Breen et al., : ). Since the speakers’ utterances more clearly disambiguated focus size when such feedback was given, it is possible that the listeners’ ability to perceive distinctions in focus size depends on whether speakers are aware of the need to disambiguate their productions. As noted in Bishop (), which provides an overview of research into the perception and interpretation of prosodic cues to focus size, previous studies seem to suggest that prenuclear prominence (i.e. a pitch accent on the verb) may be ‘less relevant to identifying broad focus than to identifying narrow focus’ (p. ). The results of his investigation into the processing of prosodic cues to focus size, which involved a cross-modal priming task, suggest that prenuclear accents were unexpected when the object of subject-verb-object sentences was narrowly focused, but did not influence reaction times when the verb phrase of the sentence was focused. While Bishop’s findings are compatible with Selkirk’s () Focus Projection Hypothesis, studies like D’Imperio () or Breen et al. () challenge the assumption that an utterance with a prosodically prominent object noun phrase is ambiguous between object, verb phrase, and sentence focus.
.. Experimental tasks for the perception and interpretation of prosodically realized focus Although researchers have intuitions about how they perceive and interpret prosodically realized focus, the preceding sections have shown that experimental investigations are essential to identifying the phonetic and phonological properties of utterances that naïve listeners attend to. Experimental investigations can also identify inter-speaker variation in the prosodic cues attended to; for discussion see, for example, Breen et al. () and Bishop (). The preceding sections have also highlighted that research at the interface of prosody and meaning must carefully define the meanings that are investigated, including
‘focus’ or ‘contrastive focus’. Finally, research must consider that the perception and interpretation of prosodically realized meaning, such as focus, may be context-dependent and control for context: when context is not controlled for, participants’ interpretations may differ simply because participants imagine different contexts. Experimental research on the perception and interpretation of prosodically realized focus employs a wide range of tasks: participants are asked to identify important or prosodically prominent words (e.g. D’Imperio, ; Clopper & Tonhauser, ), to identify congruent question-answer pairs (e.g. Most & Saltz, ; Gussenhoven, ; Birch & Clifton, ; Breen et al., ; Clopper & Tonhauser, ), or to answer questions about what a particular utterance means (e.g. Kurumada et al., ), or whether the utterance is true (e.g. Terken & Nooteboom, ). Eye-tracking and priming experiments provide information about the processing of prosodically realized focus (e.g. Dahan et al., ; Ito & Speer, ; Bishop, ). Stimuli can be resynthesized utterances (e.g. Ladd & Morton, ; Calhoun, ) or utterances produced by trained speakers (e.g. Gussenhoven, ) or untrained speakers (e.g. Breen et al., ; Clopper & Tonhauser, ); they may also differ in whether they realize specific prosodies (e.g. D’Imperio, ; Ito & Speer, ) or are randomly selected from a production experiment (e.g. Breen et al., ; Clopper & Tonhauser, ). See Watson et al. () for arguments in favour of experimental tasks that can be used to explore both the production and comprehension of prosody, and tasks that allow for the integration of contextual information.
. C
.................................................................................................................................. In this chapter, an information-structural focus was defined as an expression that answers the question addressed by the utterance in which the expression is realized. As discussed in the preceding sections, the prosodic realization of utterances in many languages constrain what is focused in the utterance, and speakers of these languages can perceive and interpret prosodically realized foci. Experimental investigations are invaluable in identifying the phonetic and phonological cues to focus because such cues are subject to inter-speaker and inter-utterance variation, and because the phonetic and phonological properties of utterances are not only implicated in conveying focus. Although many open questions remain, the results of this research already have important implications for theories of meaning. For instance, theories of meaning must consider that the prosodic realization of focus is highly variable across utterances, speakers, and contexts. Similarly, although the prosodic realization of an utterance provides cues to focus, prosodically realized focus does not, typically, unambiguously determine the perception and interpretation of focus in such utterances, even when the prosodic cues to focus are consistent and speakers were aware of the ambiguity to be resolved. As a consequence, theories of meaning that start from the premise that the input to interpretation unambiguously marks what is focused in an utterance (e.g. via F-marking in the syntax) constitute an idealization of the complex interplay of prosody and focus. (Discrepancies between empirical findings and theoretical analyses in research on clefts are discussed in Onea’s Chapter in this volume.)
Information-structural focus is related to or implicated in other categories of meaning. As a consequence, studies of the production, perception, and interpretation of focus are relevant for a broad set of meanings beyond focus. Contrastive topics, for instance, have been defined as expressions that answer super-questions to the question addressed by the utterance in which the expression is realized; that is, as foci relative to a super-question (e.g. Jackendoff, ; Roberts, a; Büring, ). It is therefore not surprising that contrastive topics, too, can be prosodically realized, and that prosodically realized contrastive topics can be perceived and interpreted (e.g. Braun, ; Wagner et al., ). Contrastive topic-marking in turn influences the interpretation of sentences with scalar adjectives: De Marneffe & Tonhauser (to appear) found that American English utterances with scalar adjectives that were realized with the contrastive topic contour were more likely to give rise to a scalar implicature than when the utterance was realized with (what they referred to as) a ‘neutral’ contour. More generally, research on conversational implicatures has established, for a range of scalar expressions, that the relevant scalar implicature is more likely to arise when the scalar expression is prosodically prominent than when it is not (e.g. Chevallier et al., ; Thorward, ; Zondervan, ; Cummins & Rohde, ; Schwarz et al., to appear). Lastly, it has been hypothesized that information-structural focus is implicated in presupposition projection (e.g. Beaver, ; Beaver et al., ; Simons et al., ), and experimental research on presuppositions has shown that the prosodic realizations of utterances with presupposition triggers influences the projectivity of the presuppositions (Cummins & Rohde, ; Tonhauser, ; Stevens et al., ); see also the discussion in Schwarz’s Chapter on presuppositions, in this volume. In conclusion, as these few examples already show, experimental investigations of prosody are essential to developing a comprehensive perspective on natural language meaning.
A I thank Cynthia Clopper, Chris Cummins, Craige Roberts, and Jon Stevens for comments on earlier drafts of this chapter, and I gratefully acknowledge support from National Science Foundation grant BCS- in the writing of this chapter.
......................................................................................................................
......................................................................................................................
R on politeness has grown steadily over the past several decades and represents a truly multidisciplinary (if not interdisciplinary) field. Linguists, social psychologists, cognitive psychologists, communication scholars, sociolinguists, and others have all made important contributions to this endeavour. Experimental research on politeness, the focus of this chapter, represents a relatively small subset of this work. It has, however, made important contributions, particularly in terms of evaluating theoretical proposals. The purpose of this chapter is to provide a relatively broad overview of experimental research on linguistic politeness, describing both methodological techniques as well as some of the major findings and their theoretical implications. I provide first a brief overview of the major theoretical orientations to politeness. This is followed by a review of early empirical research on Brown & Levinson’s (B&L) politeness theory as well as the issues raised by this research. I then consider more recent research examining () the role of politeness in reasoning and the communication of uncertainty, and () the cognitive and neural processes involved in the processing of politeness.
. A
.................................................................................................................................. Although there is a long history of scholarly interest in politeness (e.g., see Watts, ), experimental work on politeness followed the appearance of maxim-based models in the late s and early s. Two of these models, those of R. T. Lakoff (, ) and Leech (), adopted the Gricean () view of communication, but expanded Grice’s set of conversational maxims to include a set of politeness maxims designed to account for why some linguistic forms were preferred over others. Both theorists proposed a Politeness Principle (PP) that worked in conjunction with Grice’s Cooperative Principle (CP). For Leech, the PP included the maxims of generosity, tact, approbation, modesty, agreement, and sympathy. Lakoff ’s () politeness maxims included: give options, don’t impose, and make the other feel good. Politeness occurs, according to these models, because speakers are constrained by these politeness maxims.
The most popular maxim-based theory, of course, was B&L’s politeness theory. Their theory, published first as a chapter in a volume on Questioning in , and then later as a stand-alone book in , continues to be the standard approach, against which all new theoretical developments are compared. The unique contribution of their theory was the inclusion of face-work as a universal motive for regulating politeness. Borrowing the concept of face from Goffman (), and merging it with Durkheim’s () positive and negative rites, B&L postulated the existence of two universal human desires: negative face (desire for autonomy) and positive face (desire for approach by others). Humans are assumed to be oriented to both positive and negative face, and politeness is the linguistic (and non-linguistic) means for symbolically attending to these two dimensions. Equally important in their theory was the inclusion of broad social variables (power, distance, and act imposition) designed to influence the perceived weightiness of an act, and hence the politeness with which that act is performed. Although popular and generative, numerous criticisms of the B&L model resulted in the development of discursive or post-modern approaches that began to gain currency in the s (Watts, , ; Eelen, ; Locher, ) and today remain a vibrant alternative to the B&L approach. Fundamental to these approaches is the distinction between politeness as a lay or folk concept, termed first-order politeness (politeness), and politeness as a technical, sociolinguistic variable, or second-order politeness (politeness). Researchers taking a discursive approach generally argue that there is a fundamental divergence between first- and second-order politeness, and that it is the former rather than the latter that should be the central concern for sociolinguists. On this view, there are no universal motives or mechanisms for politeness; rather, politeness is completely situated. More recently, there have been attempts to articulate a middle ground between an overarching politeness model (à la B&L) and situated discursive views of politeness. These middle-ground or interactional models (Terkourafi, ; Arundale, ; Haugh, ) view politeness less as a strategy and more as expected behaviour, but behaviour that nonetheless is partially a function of more general interactional constraints.
. T B & L’
.................................................................................................................................. Early experimental work on the B&L model focused primarily on two issues: the proposed ordering of politeness strategies and the role played by social variables in determining politeness levels. In terms of the former, B&L proposed the existence of four linguistic super-strategies that constituted a universal continuum of politeness. The ordering was based on the degree of face-threat associated with each of the following strategies; bald-onrecord (no face support) was the least polite (and hence most face-threatening), followed in ascending order by positive politeness (emphasize solidarity with hearer), negative politeness (indicate respect for hearer’s autonomy), and off-record (maxim violation with deniability) politeness (the least face-threatening). Experiments designed to test this ordering focused primarily on requests and the general research strategy was to have participants rate tokens of the four super-strategies in terms of politeness and related interpersonal
constructs (e.g. liking of the speaker). Partial support for the theory’s predicted ordering of the super-strategies was reported by several researchers (Fraser & Nolan, ; Hill et al., ; Blum-Kulka, ; Bauman, ; Holtgraves & Yang, ). More fine-grained politeness orderings were examined as well. Clark & Schunk (; see also Holtgraves & Yang, ) examined the perceived politeness of a set of negatively polite requests and found perceived politeness to vary as a function of the implied cost (i.e. threat) to the hearer. For example, ‘Could you x?’ was perceived as less costly, and hence more polite, than ‘Would you x?’ This research, however, raised several issues regarding this ordering. First, a consistent exception to the predicted ordering was that negatively polite forms were often ranked higher in politeness than off-record forms (Blum-Kulka, ; Holtgraves & Yang, ). Some researchers suggested that off-record forms carry a cost because the recipient must make an effort in order to infer the speaker’s meaning (Leech, ; Blum-Kulka, ). Others have argued that off-record forms may give the impression of manipulativeness on the part of the speaker (Lakoff, ). This issue is part of a larger question regarding the extent to which politeness should be equated with indirectness. If politeness is viewed as any deviation from maximally efficient communication (i.e. not in accord with all Gricean maxims) then there is a rough correspondence between politeness and indirectness. However, indirectness can occur for reasons other than politeness. Pinker and colleagues (Pinker et al., ; Lee & Pinker, ), for example, have argued that off-record forms are motivated by attempts to negotiate the nature of the relationship between interactants, and that the use of off-record forms accomplish this by providing speakers with a means of achieving plausible deniability. In this view, then, politeness and indirectness do not reside on the same scale. A related issue concerns the proposed ordering of negative and positive politeness strategies. B&L, following Durkheim () and Goffman (), argue that positive politeness (an approach-based strategy) is inherently less polite than negative politeness (an avoidance-based strategy), due to the presumption of closeness inherent in the former. However, some researchers argued that these forms are qualitatively different and hence cannot be ordered on a unidimensional continuum (Scollon & Scollon, ; Baxter, ; Tracy, ; Lim & Bowers, ). Still, for directives (threats to the hearer’s negative face) the proposed ordering makes sense theoretically (negative politeness grants the hearer greater autonomy than positive politeness) and is supported by empirical research (Holtgraves & Yang, ). On the other hand, for acts that threaten primarily the hearer’s positive face, research suggests that positive politeness may be perceived as more polite than negative politeness (Lim & Bowers, ; Holtgraves, ). It is possible that there may be a specificity principle at work here such that strategies orienting to the specific type of face threatened will be regarded as the most polite strategy. Thus, negatively polite strategies would be more polite for acts threatening the hearer’s negative face, and positively polite strategies would be more polite for acts threatening the hearer’s positive face. Finally, several researchers have argued that B&L’s set of super-strategies is incomplete due to its overemphasis on politeness at the expense of impoliteness (Craig et al., ; Culpeper, ). That is, the least polite strategy in the B&L model is bald-on-record, a strategy which is simply the absence of any face support. No doubt aggressive attacks on others’ face occur and are not handled well within the B&L model.
. S
.................................................................................................................................. One of the most important features of the B&L model is the specification of links between politeness and the major social dimensions of power and distance. Their theory assumes that increasing hearer power, relationship distance, and act imposition will increase the overall weightiness of the act (i.e. degree of face-threat), and increased weightiness is expected to result in increased politeness. Researchers have used relatively straightforward role-playing scenario techniques to manipulate power, distance, and imposition in order to examine their predicted impact on the perceptions and production of politeness. Fairly consistent support has been reported for act imposition, with increasing imposition associated with increasing levels of politeness, an effect that has been reported for requests (Brown & Gilman, ; Leichty & Applegate, ; Holtgraves & Yang, ), expressions of gratitude (Okamoto & Robinson, ), recommendations vs. reports (Lambert, ), and accounts (i.e. explanations for untoward behaviour; McLaughun et al., ; Gonzales et al., ), as well as other speech acts (Brown & Gilman, ; Leichty & Applegate, ). Null findings have been reported (Baxter, ) but are rare. The power variable has also received experimental support. In general, greater politeness occurs or is expected as a function of increasing power of the recipient relative to the speaker (and hence decreasing power of the speaker). This effect has been reported with requests (Holtgraves & Yang, , ; Leichty & Applegate, ; Lim & Bowers, ), including observational studies of actual requests (Blum-Kulka et al., ), messages conveying bad news (Ambady et al., ), teasing (Keltner et al., ), reminders and complaints (Leichty & Applegate, ), criticisms (Lim & Bowers, ), accounts (Gonzales, Pederson et al., ), and questions (Holtgraves, ). Some of these effects have been replicated cross-culturally (Holtgraves & Yang, ; Ambady et al., ). Finally, the results for the effects of relationship distance on politeness have been mixed. Consistent with the theory, some researchers have reported greater politeness as a function of increasing distance between interlocutors (Wood & Kroger, ; Holtgraves & Yang, ); others have reported the exact opposite (Baxter, ; Brown & Gilman, ). And some (e.g. Lambert, ) have reported no relationship between distance and politeness. Distance, of course, is a multifaceted variable and it has been measured and manipulated in a variety of ways. Slugoski & Turnbull () (see also Brown & Gilman, ) argued that researchers sometimes confounded distance (i.e. familiarity) and affect (i.e. liking). Higher levels of politeness have been found to be associated with greater interpersonal distance (i.e. interactants are more polite with people with whom they are less familiar) but also with greater liking (people are more polite with those whom they like). More recently, the relationship between distance and politeness has been investigated in terms of a popular social psychology theory termed Construal Level Theory (CLT; Trope & Liberman, ; Liberman & Trope, ). For CLT, something is psychologically distant (temporally, spatially, or socially) when it is not part of one’s direct experience. Psychologically distant stimuli are typically represented at a higher, more abstract level of construal while stimuli that are close are represented at a lower, more concrete level of construal. Stephan et al. () argued that higher level construal is generally more polite than lower
level construal. In eight experiments they found increasing politeness to occur as a function of increasing spatial and temporal distance, and increasing politeness to result in inferences of greater temporal and spatial distance. For example, in terms of temporal distance, participants produced more polite messages directed at someone in the distant future relative to someone in the near future. And conversely, the use of more polite forms led to judgements of greater temporal distance than the use of less polite forms. This represents an important extension of politeness theory because it both demonstrates the bidirectional relationship between distance and politeness, and expands the notion of distance to include spatial and temporal distance. One issue that has been raised regarding the B&L model is the manner in which power, distance, and imposition interact. The model (implicitly) assumes that their effects are additive. Empirical research suggests otherwise. Many researchers who have examined the simultaneous impact of these variables on politeness have reported interactions between them, including Power by Distance interactions (Blum-Kulka et al., ; Holtgraves & Yang, ; Lim & Bowers, ), Imposition by Distance interactions (Leichty & Applegate, ; Holtgraves & Yang, ), and Imposition by Power interactions (Gonzales et al., ; Holtgraves & Yang, ). The existence of these interactions simply means that as the effects of one of the interpersonal variables become very large, the effects of the other two variables become much smaller. For example, a person who makes an extremely large request (i.e. high imposition) will tend to be polite regardless of power and distance. A second issue is whether politeness is influenced by variables other than power, distance, and imposition. Obviously it is. It is important to note, however, that power, distance, and imposition are high-level, abstract variables that should subsume more specific variables. For example, ethnicity, gender, occupational differences, and so on are variables that feed into power and distance, and ultimately, politeness. Even mood states may be incorporated in the model in this way. For example, Forgas (a,b) demonstrated that people in sad moods prefer to use greater levels of politeness than people in happy moods, possibly because a person’s mood influences their perceptions of the interpersonal context (power, distance, and imposition). Hence, people in a sad mood may perceive themselves as being relatively low in power, or perceive an act as being relatively more imposing, and it is these perceptions that affect their level of politeness. It may also be possible to use the model to explore individual differences in politeness, an area that has seen relatively little research. Introverts, for example, may perceive relatively greater distance between themselves and others, and hence produce higher levels of politeness. Extraverts, on the other hand, may perceive relatively less distance and hence favour the use of relatively less polite but more approach-based strategies (i.e. positive politeness). On the other hand, Terkourafi (, ; Vergis & Terkourafi, ) has argued for a frame-based approach to politeness in which different situational contexts, over time, come to be associated with expected politeness forms (i.e. they become conventionalized). Although these expectancies can be overridden by the context, the default meanings of these terms become part of their lexical meanings and the terms do not intentionally convey (im)politeness. This alternative offers a more granular approach, one in which power, distance, and imposition can play a role in politeness, but not the overarching role theorized by B&L.
A third issue concerns the direction of the relationship between power, distance, and politeness. For B&L, relative power and distance are often viewed as determinants of politeness; that is, speakers’ estimation of these variables determines act weightiness and hence the politeness strategy to be used. However, as many have noted, this is a static view and it is likely that the relationship is bidirectional. Specifically, if the use of a particular linguistic form is affected by power and distance, it follows that the use of a particular linguistic form will be informative for observers (including the hearer) regarding the speaker’s perceived power and distance. For example, because high-status speakers use less polite forms than lower-status interactants, the use of less polite forms should result in perceptions of higher speaker status, other things being equal. In a cross-cultural study using participants from the United States and South Korea, Holtgraves & Yang () found that less polite request forms were associated with perceptions of greater speaker power. And there is a fairly substantial literature on what is termed powerful (vs. powerless) language. In general, this research suggests that the use of powerful language (essentially less polite language) results in perceptions of relatively greater power, credibility and persuasiveness (Erickson et al., ; Bradac & Mulac, ; Gibbons et al., ; Hosman & Wright, ; Burrell & Koper, ; Holtgraves & Lasky, ; Blankenship & Holtgraves, ), although there is some evidence that these effects may be moderated by speaker gender (Carli, ). The relationship between politeness levels and perceptions of the speaker can be quite complex. For example, when a high-status speaker is extremely polite to a subordinate it will often result in perceived sarcasm (Slugoski & Turnbull, ). Moreover, this bidirectional relationship allows people to strategically vary their politeness as a means of negotiating and/or altering the interpersonal context; it is, in effect, an important component of impression management (Goffman, ). So, a higher-power person (e.g. a boss) can move from negative politeness to positive politeness in an attempt to negotiate a closer relationship. Similarly, a person in an established relationship may begin to use less politeness as a means of negotiating higher power in the relationship, and so on. This possibility can obviously result in interpersonal misperceptions or misunderstandings. A speaker may assume his politeness level reflects one dimension (e.g. closeness), but his interlocutor may assume it reflects a different dimension (e.g. power). This negotiated nature of politeness-based person perception awaits further empirical investigation.
. P, ,
.................................................................................................................................. A new line of research has developed over the past decade that examines the role of politeness in reasoning and the communication of uncertainty (Bonnefon et al., ; Bonnefon, ; Juanchich et al., Chapter in this volume). This research has demonstrated how politeness can influence the interpretation of probability terms (e.g. possible), connectives (e.g. or, and), and quantifiers (e.g. some, all). Consider first probability terms. Their use can convey varying degrees of uncertainty (e.g. It’s possible you’ll flunk the
course). There are, however, other motivations for their use. Specifically, probability expressions can function as a politeness strategy (e.g. as a hedge on an assertion) and used as a means of managing face. Rather than saying ‘You’ll never finish it in time’—an expression that might threaten the recipient’s face—a speaker can hedge his opinion with ‘It’s possible you won’t finish it in time’. The existence of multiple motives for the use of probability terms can influence the manner in which they are interpreted. This was demonstrated by Bonnefon & Villejoubert () who asked participants to imagine that their family doctor told them they would ‘possibly’ develop either deafness or insomnia in the upcoming year. Participants judged the probability of the more severe disease (deafness) to be significantly higher than the probability of the less severe disease (insomnia). This occurred because participants judged the doctor to be using ‘possibly’ as a face-management device significantly more frequently when the condition was deafness than when it was insomnia, and they adjusted their estimates accordingly. In this experiment it was the hearer’s face that was being threatened, although it is possible for the speaker’s face to be threatened as well, with a similar impact on the interpretation of probability terms (Juanchich et al., ). This effect has been demonstrated also with certain scalar expressions. Scalar expressions can be ordered on a scale with respect to their strength (e.g. ) (Levinson, ) and it has been argued that the use of the weaker, more inclusive, term (e.g. some) implies that the stronger term (e.g. all) does not hold (Levinson, ; Horn, ). Hence, the scalar implicature for ‘some’ is ‘some but not all’. Bonnefon et al. () demonstrated that face management can influence the likelihood that a scalar implicature is generated, in much the same manner that it influences the interpretation of probability terms. When a situation is face-threatening, ‘some’ can be used as a hedge to politely indicate ‘all’. Consistent with this reasoning, Bonnefon et al. () found that estimates of ‘some’ in ‘Some people hated your party’ were higher than estimates of ‘some’ in ‘Some people enjoyed your party’; people were more likely to assume that ‘some’ was being used in the service of face management in the former situation, and they adjusted their interpretations accordingly. More recently, Bonnefon et al. () demonstrated that this effect varies as a function of the discourse context such that the effect is more pronounced when the scalar term is preceded by a brief silence, a dispreferred marker signalling that bad news is forthcoming (Holtgraves, ). Finally, just as scalar expressions can be ordered on a scale according to their strength, so too can connectives (e.g. ). And just as with scalar expressions, the interpretation of these terms can be influenced by politeness considerations. Feeney & Bonnefon () manipulated the content connected by ‘or’ such that it was either positive or negative (i.e. face-threatening). They found that exclusive interpretations (i.e. ‘one or the other but not both’) were significantly more likely for positive content than for negative content. There is a potential face management motive for the term when the content is negative, and because of this, people are less likely to generate the exclusive interpretation and instead assume that both interpretations are possible. This research demonstrates that politeness can influence the interpretation of connectives, probability expressions and scalar terms. Is the production of these terms similarly affected? It appears so. Juanchich & Sirota () asked participants to choose the probability expression that they would use in order to convey a particular outcome, and they
manipulated the speaker’s communicative goal such that they were instructed either to be informative, to avoid blame (i.e. speaker face management), or to avoid upsetting the hearer (i.e. hearer face management). As expected, when the outcome was negative, participants chose expressions that conveyed less certainty under face management instructions (both hearer-face and speaker-face) than under instructions to be informative, an effect that was reversed when the outcome was positive. In a follow-up study, Sirota & Juanchich () asked participants to choose probability expressions that they would use to convey negative information, and then assessed their communicative motives (rather than manipulating them) for choosing those expressions. Participants frequently (.%) indicated face management as their motive, and when they did so, they were more likely to choose an expression that conveyed a lower probability, relative to when they indicated that their motive was to be informative. More recently, Holtgraves & Perdew () examined both production and interpretation in the same study. Participants in these experiments were presented with situations that varied in terms of face-threat and were asked how they would communicate potentially threatening information. Both hearer threat (Experiment ) and speaker threat (Experiment ) were examined, and participants either chose from a pre-existing set of utterances or responded in an open-ended manner. In both experiments a second set of participants read these utterances and provided judgements as to the degree of uncertainty conveyed by the utterance. In both experiments, messages in the face-threatening condition conveyed greater uncertainty than messages in the non-face-threatening condition, and the probability estimates made by the second set of participants varied as a function of the conveyed uncertainty. This research demonstrates that when examining speakers and hearers together, severe events may be judged less likely (rather than more likely), because speakers tend to hedge the certainty with which they communicate the information.
. P
.................................................................................................................................. Variations in politeness can have a range of social and cognitive effects. But how, exactly, do these effects occur? Several different lines of research examining the processing of politeness have been pursued. One line of research has examined the extent to which people encode and retain the wording that conveys politeness. This is an important issue because it is relevant for the claim that at least some forms of politeness are expected, normative behaviour and hence non-salient (Watts, ). A frequent finding in the memory literature is that people typically forget the exact wording of an utterance but retain the gist of what was said (Sachs, ). One important exception to this is that wording will be remembered well when it has interpersonal implications (Keenan et al., ; MacWhinney et al., ). Politeness, of course, has clear interpersonal implications and research suggests that politeness wording is spontaneously encoded and retained. For example, Holtgraves (; see also Slugoski, ) examined incidental memory for wording that varied in politeness and found that people remembered politeness wording at levels exceeding chance. Interestingly, even when the specific wording was not remembered, people appear to have encoded the overall level of politeness and recalled wordings
consistent with that level of politeness. In other words, if participants heard an impolite request, when asked to recall that request they tended to recall an impolite (rather than polite) form, even if they could not remember the exact wording. It was also the case, however, that memory for politeness wording was greater for cases of politeness wording that were inconsistent with the social context. For example, participants in a psychology experiment were more likely to remember impolite forms if the speaker was low (rather than high) in status (a graduate student) and polite forms if the speaker was high (rather than low) in status (a faculty member). Such forms violate expectations and hence are remembered well. Memory studies have a potential limitation in that their results can be ambiguous in terms of actual on-line processing. More recently, researchers have begun to use on-line techniques in order to examine the processing of politeness in real time. One such approach is to use an eye-tracking methodology. For example, Raizen et al. () used an eyetracking procedure to examine the processing of taboo words (i.e. potential violations of positive face). They found that early fixations on taboo words varied as a function of an interaction between the speaker’s identity and the situational appropriateness of a taboo word. Specifically, taboo words in appropriate situations resulted in longer fixations when uttered by someone unlikely to use taboo words (‘saints’) than when uttered by someone expected to use taboo words (‘sinners’). Other researchers have used electrophysiological techniques to examine politeness processing. For example, Jiang and colleagues (Jiang et al., , ) recorded EventRelated Potentials (ERPs) as participants read conversations in which speaker status and pronoun type (respectful vs. disrespectful) were manipulated. Status-inconsistent pronouns (e.g. disrespectful pronouns from a lower-status speaker) resulted in an enhanced N—interpreted as an indicator of semantic integration effort—compared to statusconsistent pronouns, thereby suggesting that brain activity varies as a function of the pragmatic implications. The status-inconsistent pronoun also resulted in a later negativity (‒ milliseconds (ms)) but not late positivity (i.e. P). The authors interpret the failure to observe the latter as indicating support for a contextualized view of politeness (Terkourafi , ) rather than an intentionalist view (B&L). These researchers also examined individual differences in reactions to the politeness violation. Male participants, as well as those scoring higher on a measure of fantasy (ability to transpose oneself into the feelings and actions of fictitious others), demonstrated a significantly larger N in response to the status-inconsistent pronoun than females and those scoring lower on fantasy. Hoeks et al. () examined neurophysiological reactions to impolite utterances that were request refusals. Participants listened to dialogues in which one person asked another for a favour (either a request for an action to be performed or a question) and the other responded with a blunt ‘No’ or a ‘No’ that included an apology and reason for the refusal. These researchers found a significant P for the former relative to the latter. However, for the request for action, reactions to the refusal condition started earlier, were more broadly distributed, and of a larger magnitude relative to the question condition. The authors interpret this reaction as reflecting participants’ reorganization of their conversation model as a result of linguistic impoliteness. The possibility that this reorganization may reflect also an updating of representations of the speaker (i.e. that he’s rude) was noted.
Recently, Holtgraves & Kraus () used electrophysiological techniques to examine the role of politeness in the comprehension of several different scalar expressions. Prior research using electrophysiological techniques to examine scalar expressions has explored their use in logical contexts (e.g. Some people have lungs) (Noveck & Posada, ; Nieuwland et al., ; Hartshorne, Snedeker, Azar, & Kim, ; Skordos & Barner and Breheny, Chapters and , respectively, in this volume). In contrast, Holtgraves & Kraus examined their use in conversational contexts. They had participants read scenarios followed by a target utterance containing a scalar expression (e.g. some) in the first half of the utterance, with a second half continuation of the utterance containing either the pragmatic meaning (e.g. not all) or the semantic meaning (e.g. all). The semantic meaning resulted in a larger P than did the pragmatic meaning, and this difference was greater when the situation was face-threatening relative to when it was not face-threatening. This suggests that in conversational contexts, scalar implicatures are generated when the scalar term is encountered, especially when the situation is face-threatening. Deficit-based research is also relevant for understanding politeness processing. Consider, for example, recent research on language deficits in Parkinson’s disease (PD). Although PD is primarily associated with debilitating extrapyramidal motor dysfunction, it also affects thinking, reasoning, planning, and language functions, and in terms of the latter there is some evidence of pragmatic impairment in PD (e.g. Lewis et al., ; McNamara & Durso, ), including politeness. To investigate the latter, Holtgraves & McNamara () used a role-playing task and asked participants (those with PD and matched controls) to imagine being in situations in which they were to make a request of another person and to write out exactly what they would say in order to make each request. Overall, the PD participants were less polite than the control participants. More importantly, although increasing imposition was associated with increasing politeness for control participants, this did not occur for PD participants, suggesting a reduced sensitivity to the social context for those with PD. This reduced sensitivity also occurred for speaker power, but only for PD participants who were on higher doses of dopaminergic medication. What are the potential sources of the politeness impairment in PD? One possibility is that it reflects an overall decline in cognitive capabilities, especially executive cognitive functions (Owen et al., ; Lange et al., ; Taylor & Saint-Cyr, ; Troster & Woods, ; McNamara et al., ). Reduced cognitive resources in PD could result in an attentional deficit such that variations in request size are not noticed; because they are not noticed there is no corresponding change in politeness. Consistent with this possibility, researchers have demonstrated a clear connection between executive function deficits in PD and the ability to contribute meaningfully (e.g. being appropriately informative) to conversations (Holtgraves et al., ). Another possibility is that even when variations in the context are noticed, the cognitive capacities required to produce more polite (and cognitively complex) strategies are hindered in people with PD. In the Holtgraves & McNamara () research, participants on higher doses of dopaminergic medication did notice variations in recipient status (based on manipulation check items) but they failed to produce more polite strategies for a higher-power recipient. Researchers also have examined politeness processing deficits associated with damage to the right hemisphere (RHD). Pell (; Experiment ) examined politeness judgements of RHD and heathy controls in which both prosody and utterance type (direct, indirect, very
indirect) were manipulated. Past research has suggested RHD is associated with a prosody processing deficit. However, in this experiment it was the interaction of prosody and language, more specifically the occurrence of linguistic and prosodic discrepancies as a means of conveying sarcasm, that posed difficulties for RHD participants (but not for healthy controls). In addition, the distinctions made by RHD participants based on linguistic cues alone were not as fine as those made by healthy controls, a finding similar to that reported for PD individuals.
. C
.................................................................................................................................. Experimental studies of politeness continue to be a small but relatively critical component of politeness research. After a flurry of experimental activity in the s and s designed to test predictions derived from the B&L model, experimental studies of politeness have expanded into new areas as described in this chapter. Looking forward, I expect there will be a continuing focus on the processing of politeness. The use of electrophysiological techniques in this regard is only beginning and has the potential to provide relevant information regarding the processes involved in comprehending politeness. Experimental studies of the role of politeness in reasoning and the communication of uncertainty should continue as well. Hopefully, the success of this endeavour will prompt researchers to consider other areas for which the analysis of politeness processes may be relevant. There is, for example, no shortage of topics in social psychology (e.g. persuasion, person perception, relationship negotiation, etc.) for which politeness plays an important but as yet unexamined role. Other important areas for future experimental research include the relationship between politeness and indirectness, the manner in which non-verbal politeness interacts with and modifies verbal politeness, individual differences in politeness, and of course cross-cultural similarities and differences in politeness, all areas that could be examined experimentally. Experimental studies of politeness should continue to remain an important component of politeness research because they allow for the evaluation of theoretical predictions and tests of competing models. For example, the results of early empirical research on the B&L model suggested several ways in which that model needed to be modified, including the ordering of politeness strategies and the interactive effects of social variables. Experimental studies of the processing of politeness have much to offer in this regard. Or consider the claim that politeness is normative and expected and hence typically not noticed (Watts, ). Experimental memory research suggests that this may not be the case, and that people do encode some representation of a speaker’s level of politeness, even when it is expected (Holtgraves, ). Or consider the manner in which ERPs to status pronoun violations provide support for a contextualist rather than intentionalist view of politeness (Jiang et al., ). The downside of experimental studies of politeness is their potential artificiality and decontextualized nature. Trade-offs are involved, of course, and gains in experimental control are often paid for with a drop in realism. Still, it is possible to make experimental stimuli fairly realistic, by, for example, collecting actual discourse samples to then be used
in experimental research. And the development of new methodologies may eventually allow for the examination of real-time situated politeness via the use of electrophysiological techniques as individuals engage in (constrained) natural language use (see, e.g., Hoeks & Bouwer, ). The gain in precise experimental control, coupled with the back and forth between theory and data, can facilitate advances in our understanding of certain facets of politeness, an understanding that can contribute to and compliment advances made with non-experimental techniques. In short, experimental research on politeness should not be ending anytime soon.
A The writing of this chapter was supported by a grant from the National Science Foundation (BCS-).
......................................................................................................................
......................................................................................................................
-´
. I
.................................................................................................................................. Young children find desires easier to understand than beliefs and may understand desires with little or no understanding of beliefs. In this chapter, I review evidence and explanations for that lag. I conclude by offering a novel explanation, arguing that children’s conception of other people undergoes an important shift when they begin to engage in conversation. Initially young children conceive of other people primarily as agents with goals, but around three years of age they start to construe people as epistemic subjects capable of exchanging information for the formation and updating of beliefs. Hence, they shift from a desire psychology to a belief-desire psychology.
M has changed in social cognition research since Paul Harris opened his contribution to Theories of Theories of Mind with this passage, back in . However, as I will argue in this chapter, nothing in the last twenty years suggests that Harris was wrong to believe that children learn to reason about beliefs in conversation. This is why I will conclude that Theory of Mind has a lot to learn from experimental pragmatics.
. B : A , ?
.................................................................................................................................. Reasoning about other people’s beliefs lies at the heart of human Theory of Mind: our capacity to interpret other people’s behaviour by reference to their desires, intentions, and other mental states. The central role of belief reasoning in Theory of Mind research is most evident in the experimental record, where the large majority of studies in the last thirty years have investigated children’s and infants’ performance in false-belief tasks. Standard false-belief tasks were originally designed for preschoolers and elicit a response by asking the child a direct test question. For example, in the Sally-Anne task (Baron-Cohen et al., ), a character named Sally puts a marble in a box before leaving the scene, and in her absence, a second character named Anne moves the marble to a basket. Children are then
asked ‘When Sally comes back, where will she look for her marble?’ If children understand that Sally holds a false belief, they should anticipate that she will look for the marble in the container where she had left it, and not where Anne put it. Unlike standard false-belief tasks, gaze-monitoring versions adapted for younger children and infants (e.g. Clements & Perner, ; Onishi & Baillargeon, ) rely on indirect measures of belief understanding (e.g. anticipatory looking to the location where Sally will look for her marble). The results of hundreds of Theory of Mind studies have revealed that children under the age of tend to fail false-belief tasks by predicting that the agent will look for the object in its current location (for a metaanalysis, see Wellman et al., ). By contrast, infants under years of age are able to pass indirect versions of these tasks by anticipating that the agent will return to the original container (for a recent review, see Scott & Baillargeon, ). These two sets of findings have been the main focus of Theory of Mind research in the last three decades. In this chapter, I will argue that this increasingly narrow focus on false-belief tasks has limited the scope of this research field to laboratory tasks, while disregarding belief reasoning in everyday situations. Given their intense focus on false-belief tasks, it is perhaps surprising that Theory of Mind studies have not addressed the question of when children engage in belief reasoning in their daily lives. The design of false-belief tasks would suggest that they engage in belief reasoning whenever they have to predict the behaviour of a mistaken person. This ability may be important for everyday life, but does not seem so frequent as to deserve the disproportionate attention that false-belief tasks have received in Theory of Mind research. Here I will defend the idea that belief reasoning does indeed play a key role in our daily lives, but not in the kind of scenarios recreated in false-belief tasks. Instead, on an everyday basis, we engage in belief reasoning whenever we engage in communication (Geurts & Rubio-Fernández, ). The received view on the use of Theory of Mind in communication is that it is necessary for rather sophisticated pragmatic phenomena. As Peterson puts it, ‘throughout childhood, an understanding of Theory of Mind is fundamental to skilled communication and social interaction, especially in complex situations including reminiscence, sarcasm, deception, humor, interpersonal conflict and persuasive negotiation’ (: ). However, the role of Theory of Mind in communication is not only limited to figurative language, as in irony comprehension (Spotorno & Noveck, Chapter in this volume) or the recognition of the speaker’s referential intention in metonymy (Schumacher, Chapter in this volume); Theory of Mind use extends to literal language production and comprehension. If we take communication to serve the primary function of exchanging information among interlocutors, any informative contribution to a conversation must take into account what the interlocutors know and do not know (for a discussion of epistemic awareness in the derivation of scalar implicatures, see Breheny, Chapter in this volume). This means speakers must take into account their common ground with their listeners in order to successfully communicate. If we did not engage in this form of belief reasoning, conversations would be plagued with redundancies and repetitions, as well as ambiguities— just to mention a few pragmatic issues related to common ground (for a discussion of the role of common ground in referential communication, see Rohde and Davies & Arnold, Chapters and , respectively, in this volume). Consider two simple examples: if you could not remember your best friend’s birthday, you would probably call a mutual friend, but you would not ask a random passer-by on the
- ´
street (who is unlikely to know your best friend). On the other hand, if you were looking for an underground station in an unfamiliar city, you may ask a passer-by for directions (rather than calling one of your friends). The fact that we try to address our questions to knowledgeable interlocutors (or at least to interlocutors whom we hope will know the answer to our questions) is evidence that communication requires reasoning about other people’s epistemic states. These examples of belief reasoning in communication seem closer to our everyday experience of interacting with others than the kind of false-belief scenarios used in Theory of Mind tasks. In this respect, experimental pragmatics has a lot to offer to Theory of Mind research. More specifically, the more varied and complex evidence of belief attribution observed in experimental pragmatics studies (relative to the pass-or-fail results of falsebelief tasks) has the potential to reveal a much more nuanced picture of the development of children’s epistemological theories.
. T M, : R
.................................................................................................................................. Children’s performance in false-belief tasks has been shown to correlate with their language development (for a metaanalysis, see Milligan et al., ). However, these correlational studies have focused on syntax, semantics, or general language ability (which combines syntax and vocabulary measures) but have not included pragmatic measures of Theory of Mind development (e.g. children’s ability to distinguish between ignorant and knowledgeable interlocutors). Children’s pragmatic abilities have been extensively investigated in developmental studies looking at different communicative settings and speech acts (e.g. Matthews et al., ; Rakoczy & Tomasello, ; Grosse et al., ; Grosse & Tomasello, ), but the results of most of these studies have not been examined in connection with children’s performance in Theory of Mind tasks (cf. Dunn et al., ; Bernard & Deleau, ). Since Onishi & Baillargeon () published their seminal false-belief study with infants, the hottest debate in Theory of Mind research has revolved around the question of whether infants possess a Theory of Mind, which has unfortunately relegated the results of experimental pragmatics studies even further. There is, however, no consensus as to what abilities infants and toddlers display in looking studies, with some researchers arguing that infants only reveal a very rudimentary or implicit Theory of Mind (Butterfill & Apperly, ; Wellman, ; Rakoczy, ), while others go further in their scepticism and affirm that infants’ success in false-belief tasks does not require a Theory of Mind (Perner & Ruffman, ; Heyes, ; Ruffman, ). On the issue of nativism, an interesting parallel can be drawn between Theory of Mind and language acquisition research. In linguistics, there is still an ongoing debate about the extent to which language is innately specified (Chomsky, ; Goldberg, ; Lieven, ). However, even if researchers disagree on whether humans are endowed with a universal grammar from birth, there seems to be a general agreement that a full theory of language must explain not only its inception but also its development (Tomasello, ; Pinker, ). That is, even if language were largely innate, the question remains as to what
changes from the time a child is born to the point when they become able to communicate fluently. By contrast, experimental studies in Theory of Mind have investigated, almost exclusively in recent years, the question of whether the concept of belief is innate or early emerging, rather than looking at how that concept develops in the first years of life. While early studies in the s and s addressed a variety of complex research questions in Theory of Mind development (including the acquisition of mental state verbs or the development of perspective taking in referential communication), the last decade has seen an increasingly narrow interest in infants’ performance in false-belief tasks, at the expense of addressing other interesting developmental questions that span the preschool and early school years. In favour of adopting a broader developmental approach in Theory of Mind research, in this chapter, I will briefly review a number of experimental studies with a special focus on communication as a way to highlight the important contribution that experimental pragmatics could make to the study of children’s early epistemological theories.
. T--: D ?
.................................................................................................................................. Harris (, ) argued that conversation has a tutorial role in the development of belief reasoning because young children come to appreciate that interlocutors have different information at their disposal, which they exchange and update in communication. This view led Harris to predict that early references to think and know would have a conversational function (e.g. ‘I don’t know’), rather than referring to mental states as such (e.g. ‘How do you know that?’). Early studies of naturalistic conversations between young children and adults supported Harris’s prediction (Shatz et al., ; Bartsch & Wellman, ). However, in view of the positive results of infant false-belief studies, Harris and colleagues have recently come to challenge the conclusions of these early studies—and indirectly, their original prediction. Harris, Ronfard, & Bartz () and Harris, Yang, and Cui () investigated -yearolds’ understanding of knowledge by looking at various pragmatic measures, such as their use of the mental state verb know, their expressions of ignorance, their formulation of information-seeking questions and their empirical corroboration of statements. Harris et al.’s interest in investigating Theory of Mind development in communication is driven by their wanting to find the ‘missing developmental link’ between infants’ successful performance in indirect false-belief tasks and -year-olds’ success in standard false-belief tasks. Thus, Harris, Ronfard, & Bartz () and Harris, Yang, and Cui () hypothesized that early studies on the acquisition of mental state verbs may have underestimated -year-olds’ abilities and that less conservative measures may offer a more accurate picture of their understanding of knowledge. While I commend Harris, Ronfard, & Bartz () and Harris, Yang, & Cui () for looking for evidence of Theory of Mind development beyond false-belief tasks, I disagree with Harris and colleagues’ motivation for two reasons. First, as discussed earlier, the
- ´
interpretation of infants’ success in false-belief tasks is highly controversial (e.g. Heyes, ; Ruffman, ; Rakoczy, ). Second, and most importantly, even if infants were able to attribute false beliefs to others, the mindreading abilities demonstrated in looking studies need not be the same as those required in pragmatic tests of belief reasoning in communication (including the use of mental state verbs). Therefore, in my opinion, the positive results of infant false-belief studies should not fuel rich interpretations of early pragmatics studies. In the following, I will discuss some of these rich interpretations of young children’s behaviour. Shatz et al. () and Bartsch & Wellman () carried out a corpus analysis of the spontaneous utterances of a number of children between the ages of and years with a focus on their use of mental state verbs. In order to reliably evaluate young children’s abilities to talk about the mind, Shatz et al. and Bartsch & Wellman distinguished between conversational functions of mental state verbs (e.g. opening a conversation with the formulaic phrase ‘You know what?’) and genuine references to mental states (e.g. ‘Mommy can’t sing it. She doesn’t know it’). This distinction is important because, as Shatz and colleagues point out, ‘the mere occurrence of a word in a child’s speech does not necessarily indicate that its producer understands its semantic presuppositions’ (: ). The results of these studies suggest that children start using mental state verbs in the months after their second birthday, but these early uses have conversational functions (i.e. turns of phrase, repetitions, and idiomatic phrases). Genuine references to mental states appear around the third birthday, including expressions of contrast between the contents of mental states and the world (e.g. ‘I was teasing you; I was pretending, ’cept you didn’t know that’). The early appearance of contrastives at age suggests that children may start revealing false-belief reasoning in naturalistic conversation before they are able to pass standard false-belief tasks. If so, these data would support the view that standard false-belief tasks are unnecessarily hard for young children (e.g. because these tasks draw children’s attention to the target object with the test question, increasing the salience of the wrong response; Rubio-Fernández & Geurts, , ). Harris, Ronfard, & Bartz () and Harris, Yang, & Cui () challenged Shatz et al. () and Bartsch & Wellman () for discarding children’s uses of the phrase ‘I don’t know’ as potentially formulaic, arguing that if we examine this phrase in the context of the preceding and subsequent utterances that were produced by the child’s interlocutor, ( . . . ) we can ask if it was used appropriately, notably as a plausible expression of ignorance in response to a range of questions, or, alternatively, if it was used in a more inflexible fashion—for example, as a way to demur or withdraw from a conversation. (Harris, Yang, & Cui, : )
Harris, Ronfard, & Bartz () and Harris, Yang, & Cui () reanalysed the corpora used by Shatz et al. () and Bartsch & Wellman () and found that at age , the most frequent function of the verb know was to deny their own knowledge; that is, -year-olds typically used this verb in the phrase ‘I don’t know’. The data suggested that toddlers produced ‘I don’t know’ at appropriate points in the conversation, which led Harris and colleagues to conclude that -year-olds are able to make relevant contributions to the ongoing conversation and acknowledge their own ignorance (Harris, Ronfard, & Bartz, : ; Harris, Yang, & Cui, : ). However, it must be
noted that Harris and colleagues’ reanalysis did not address Shatz and colleagues’ and Bartsch & Wellman’s concerns when discarding such uses as potentially formulaic: unelaborated uses of the phrase ‘I don’t know’ may be a conventional negative response simply meaning ‘I can’t answer’. The fact that young children are able to produce such negative responses at appropriate points in a conversation cannot be taken as evidence that -year-olds are admitting their own ignorance in doing so. Harris, Ronfard, & Bartz () and Harris, Yang, & Cui () also found a clear asymmetry in their data: -year-olds would deny their own knowledge (rather than that of their interlocutors) but would ask information-seeking questions to their interlocutors (rather than to themselves). Harris and colleagues propose that such an asymmetry reflects a deepseated difference between children’s privileged access to their own mental states as opposed to those of their interlocutors—from which it follows that asking self-directed questions, for example, would be otiose (Harris, Ronfard, & Bartz, : ). This interpretation of young children’s uses of the verb know is again unnecessarily rich: a more parsimonious interpretation of this asymmetry would relate it to the frequency of such uses in child-directed speech. It is an empirical question, of course, but it seems plausible that adults would more often tell young children ‘I don’t know’ than ‘You don’t know that’, and likewise, that children would more often hear adults ask ‘Do you know where my toothbrush is?’, for example, than the selfdirected version ‘Do I know where my toothbrush is?’ It is therefore possible that children’s early knowledge denials and know questions simply reflect the frequency of such uses in the adult speech that they hear, rather than an appreciation of the differential accessibility of one’s and others’ knowledge. A similar hypothesis has been corroborated in the investigation of children’s acquisition of whquestions: whereas traditional accounts had posited that the order in which English speaking children acquire wh-questions is determined by their syntactic and semantic complexity, Rowland et al. () found that the acquisition order of wh-questions could be successfully predicted from the frequency with which particular wh-words and verbs occurred in the children’s input and, moreover, that syntactic and semantic complexity did not reliably predict acquisition once input frequency was taken into account. Therefore, the frequency of different functions of mental state verbs in caregiver speech needs to be investigated before we can derive any conclusions about young children’s understanding of other people’s mental states from their use of those verbs (see Dudley et al., under review).
. T--’ : W , ?
.................................................................................................................................. Some Theory of Mind researchers have interpreted infants’ success in false-belief tasks as evidence of an implicit awareness of mental states (e.g. Clements & Perner, ; Apperly & Butterfill, ). Contrary to such lean interpretations, Harris, Yang, & Cui () posited that a less conservative analysis of -year-olds’ uses of mental state verbs may actually reveal an explicit understanding of belief. However, in order to take -year-olds’ uses of mental state verbs as evidence of explicit Theory of Mind (comparable to the evidence gathered through standard false-belief tasks), we need to go beyond the mere production of
- ´
such forms, and investigate whether young children understand the factivity of mental state verbs. That is, we need to examine whether young children appreciate that know is used in contexts where the complement is taken to be true. For example, if I had just met Sally in the garden, I could claim ‘I know that Sally is outside’, whereas if I was not sure of her whereabouts, I could at most say ‘I think that Sally is outside’. Likewise, if I say ‘Sally doesn’t know that her marble is in the basket’, my statement presupposes that the marble is in the basket. Early studies suggest that children do not differentiate verbs like know and think until the age of (e.g. Macnamara et al., ; Johnson & Maratsos, ), or even later (Schulz, ; Léger, ). However, Dudley and colleagues () have recently argued that the tasks used in those studies were too demanding for young children, for reasons independent of the factivity of the verbs that were tested (e.g. children were asked to make metalinguistic judgements about utterances, which are known to be difficult for young children). Dudley and colleagues tried to better estimate -year-olds’ understanding of factivity by using a more appropriate task for that age group. In their paradigm (adapted from Moore & Davidge, ; and Moore et al., ), children had to find a toy hidden in one of two boxes using clues in the form of attitude reports containing think and know. Crucially, the experimenter (who knew where the object had been hidden) echoed such clues on behalf of a puppet (Lambchop) who whispered into her ear. Thus, children had to figure out whether the object was in the red or the blue box by using clues such as ‘Lambchop knows/thinks that it’s in the blue box’. Unlike previous research, the results of Dudley et al.’s study revealed that about half of their -year-olds’ were able to distinguish think and know, and appreciated that know presupposes the truth of its complement, while think does not. As in the early studies by Shatz et al. () and Bartsch & Wellman (), Dudley et al.’s () investigation of young children’s understanding of knowledge in a communicative task seems to reveal better performance than is usually observed in traditional Theory of Mind tasks. Thus, like Harris, Yang, & Cui (), Dudley et al. () concluded that their results support the view that there is continuity between infants’ success in indirect false-belief tasks and preschoolers’ performance in standard falsebelief tasks because -year-olds can be sensitive to their interlocutors’ epistemic states in a linguistic task. However, the individual performances observed by Dudley et al. () do not offer strong support to their conclusions about the continuity view of Theory of Mind development. In the critical condition, the children in their study had to understand that the clue ‘Lambchop doesn’t know that it’s in the blue box’ presupposed (or at least conveyed) that the object was actually in the blue box. Dudley and colleagues report that, of the twentyeight children in their study, per cent selected the correct box (corresponding with a factive interpretation of know). However, that figure results from grouping together those children who gave two or three correct responses (out of three trials). The most striking result when looking at the individual performances is that the preferred response was actually the non-factive one:1 when looking only at those children who gave the same
1
The number of children who gave three, two, one, and zero factive responses were the following: six, six, three, and thirteen—with the last group treating know as a non-factive verb in all three trials.
response in the trials, per cent systematically selected the non-factive response, whereas only per cent systematically selected the correct answer. Therefore, the results of Dudley et al. () are closer to the findings of previous studies than their statistical analyses suggest. Future studies should therefore investigate why the semantics of know might be initially non-factive, especially if one assumes that infants already possess an explicit Theory of Mind. However, young children’s use and comprehension of mental state verbs is not the only measure of their understanding of knowledge that can be gathered through communicative tasks. Other naturalistic situations where children need to engage in belief reasoning include updating their own knowledge when receiving reliable information in conversation, and relying on their common ground with their interlocutor in referential communication.
. E : E
.................................................................................................................................. In a pioneering study, O’Neill () asked whether -year-olds would adapt their referential strategies depending on their parents’ knowledge of the location of a toy. Her results showed that toddlers were more likely to refer to the toy’s new location via pointing gestures and vocalizations when the parents were absent at the time when the toy was placed out of reach on a high shelf than when the parents had witnessed the transfer. Older and younger -year-olds (mean age: months in Experiment and months in Experiment ) revealed comparable sensitivity to their parents’ knowledge states. However, O’Neill defended a non-mentalistic interpretation of her data, according to which toddlers would only need to appreciate, in a general sense, that the parents who missed the transfer of the object had disengaged from the critical events (rather than understanding that they were ignorant of the toy’s location). Then, in wanting to retrieve the object, -year-olds would be motivated to update their parents regarding the critical events. The results of a follow-up study by Dunham and colleagues () replicated O’Neill’s original findings and supported her lean interpretation of toddlers’ pragmatic abilities as a disengagement and updating heuristic. Dunham and colleagues hypothesized that one reason why the younger toddlers in O’Neill’s study might have pointed more towards the location of the toy in the Ignorant condition is that the parents had covered their eyes (rather than leaving the room), potentially suggesting that they were engaging in a hiding game. If the toddlers in that experiment had experience with such games, they might have simply produced more pointing gestures because that is what they would normally do in such situations. Confirming Dunham and colleagues’ hypothesis, the -month-olds in their study pointed as much in the Ignorant condition as they did in a Sham-Ignorant condition in which the parents first covered their eyes but then looked at the experimenter moving the object. By contrast, the -month-olds in their sample pointed similarly in the Sham-Ignorant
- ´
condition and in the Knowledgeable condition, showing an appreciation of the parents’ perceptual access to the critical event. However, Dunham et al. () encouraged conservative interpretations of their findings, arguing that older -year-olds may simply be able to make more accurate use of the same disengagement and updating heuristic by relying on more specific cues to the parents’ perceptual access to the events. Infants as young as months of age have been found to be sensitive to an interlocutor’s engagement with an object in a referential communication setting. Tomasello & Haberl () observed that, when infants interacted with two objects with one experimenter, and with a third object with another experimenter, they would reliably interpret the first experimenter’s expression of surprise on her return (‘Wow! Look at that one!’) as referring to the third object that they had not shared. Moll & Tomasello () and Moll et al. () further observed that -month-old infants had to become familiar with the objects in joint engagement with the experimenter in order to disambiguate the experimenter’s request for the target object in the test phase. Thus, witnessing the experimenter interact with the first two objects individually did not allow infants to pass this task until age months (Moll & Tomasello, ). Using the same paradigm, Moll et al. () investigated -year-olds’ pragmatic abilities in referential communication and observed that their reliance on the engagement heuristic led them to overestimate their interlocutor’s knowledge: when the experimenter remained present in the scene but could not see the third object because of a barrier, toddlers wrongly assumed that she was familiar with all three objects. Thus, -year-olds only showed a preference for the third object in the test phase if the experimenter had left the room before the third toy was introduced; otherwise, they wrongly assumed that all three objects were part of their common ground with the first experimenter. The results of referential communication studies with infants and toddlers suggest that, at a very early age, children rely on their interlocutor’s engagement (or disengagement) in a scene as a proxy for their knowledge (or ignorance) of relevant events. Interestingly, a similar effect has been observed with adults: Gorman et al. () asked participants to instruct another participant to select one of three objects in a series of trials. Critically, the objects were novel and the two participants had to learn their names prior to the instruction-selection phase. However, the participant who gave the instructions had to learn more names than the participant who had to follow the instructions. As expected, participants named the novel objects more often when the other participant knew the name for the novel object than when only they knew the name (in which case they opted for describing the novel object). Interestingly, participants were better able to discriminate shared and private names when they learned the shared names together with the other participant than when they learned them independently (but with mutual knowledge of their learning). Gorman and colleagues concluded that shared experience provides a strong memory cue to the common-ground status of names. The results of referential communication studies with infants, toddlers, and adults suggest that ordinary memory processes tracking our interlocutor’s engagement in an event may provide access to shared knowledge without necessarily involving metarepresentation of our interlocutor’s epistemic states. This basic ability may be a precursor of belief reasoning in early referential communication, and an efficient mechanism to establish common ground among adult communicators.
. P’ ’ : F
.................................................................................................................................. In an early study, Povinelli & DeBlois () observed that -year-old children consistently followed the pointing gesture of a knowledgeable informant (who had hidden a surprise) vs. an ignorant informant (who had left the scene before the surprise was hidden). In contrast, -year-olds were not able to discriminate between the knower and the guesser. Sodian and colleagues () argued that ignoring the pointing gesture of the ignorant adult might have posed high executive control demands on the younger children, and designed a follow-up study that aimed to provide a more accurate assessment of -yearolds’ understanding of knowing and guessing. Children in Sodian and colleagues’ study had to choose between two helpers, only one of whom had witnessed where a sticker had been hidden. Two-year-old children performed at chance in this task, and it was not until the age of months that they started to consistently discriminate between the knowledgeable and the ignorant helper when asked ‘Who can help you?’ Importantly, all the children in the study were at ceiling in a non-epistemic control task, where they had to understand that only the helper who had a key could open a treasure box. In view of their results (and those of Povinelli & DeBlois, ), Sodian and colleagues concluded that ‘the seeing-knowing relation is a genuine conceptual problem for children below the age of three years’ (: ). The results of Sodian et al. () might seem surprising given the results of social learning studies, which have repeatedly shown that young children prefer to learn from accurate speakers than from inaccurate ones (for a review, see Mills, ; cf. Heyes, ). However, preferring an informant who labels familiar objects correctly to one who does so incorrectly need not require appreciating the epistemic states of such informants: the mere normativity of the accurate speaker (or the other way around: the strange behaviour of the inaccurate speaker) might be enough to enable children to form the right preference (for discussion, see Lucas & Lewis, ). Therefore, rather than manipulating speaker accuracy, a more reliable evaluation of children’s understanding of a speaker’s mental states may be established by manipulating the quality of the information provided (i.e. whether it is relevant or sufficient vs. irrelevant or insufficient). Gillis & Nilsen () conducted such a study and asked children to seek help from one of two speakers: one who had provided enough information to find a sticker (e.g. ‘It’s under the black figure’, when there was only one black figure) and one who had not provided enough information (e.g. ‘It’s under the triangle figure’, when all figures were triangles). Both preschoolers (‒ years old) and schoolchildren (‒ years old) preferred to solicit information from the speaker who had provided sufficient information. However, a second experiment confirmed that the preschoolers (but not the schoolchildren) based their preference on their own success in finding the sticker when following the clues of the two speakers: when children were not allowed to look for the sticker after the clue, only the schoolchildren were able to evaluate the reliability of the two speakers (with the preschoolers performing at chance).
- ´
Gillis & Nilsen concluded that, compared with speaker accuracy (which children as young as years are able to use in similar paradigms), the ability to use information quality to assess speaker reliability seems to emerge only at a later age. Before they are able to evaluate speaker reliability in terms of information quality, children rely on their own success when following a speaker’s clues. It must be noted that, from a pragmatic perspective, the underinformative speaker in Gillis & Nilsen’s paradigm need not have been ignorant of the location of the sticker (and hence unable to provide more specific clues); alternatively, this speaker might have simply been uncooperative (i.e. unwilling to provide sufficient information). In either case, selecting the sufficiently informative speaker over the underinformative one would require, at a minimum, appreciating that the clues provided by the underinformative speaker were pragmatically infelicitous. The results of the studies reviewed in this section suggest that children’s epistemological theories continue to develop in the preschool years. An interesting question, which has not been directly addressed in the literature, is the relationship between preschoolers’ understanding of other people’s knowledge and their understanding of the verb know. Robinson et al. () investigated children’s ability to infer speaker reliability from a speaker’s visual access to the contents of a box. Both preschoolers ( years old) and schoolchildren ( years old) changed their opinion more often when a better-informed adult disagreed with their guess about the contents of a box, than when they knew better than the adult what was inside the box. However, when the same group of children were asked ‘Who knows better what is inside the box?’, the schoolchildren performed comparably in the two tasks whereas the preschoolers performed significantly worse with the know-better question. Moreover, performance on the two tasks (i.e. changing opinion in disagreement vs. who knows better?) was unrelated in both age groups. Robinson and colleagues explained the surprising difference between the two tasks as a difference in the salience of the adult’s beliefs (which were verbalized in the disagreement task, but not in the question task). An alternative explanation that was not considered by the authors has to do with preschoolchildren’s understanding of the verb know: preschoolers may have an implicit understanding of other people’s knowledge that allows them to evaluate the reliability of their opinions (and update their own accordingly), but does not allow them to verbally express such an evaluation in terms of ‘knowing better’.
. S
.................................................................................................................................. The results of this brief review highlight the relevance of experimental pragmatics studies for Theory of Mind research. More in line with early accounts of the development of belief reasoning in conversation (Harris, , ) and contrary to what the current debate on infant Theory of Mind might suggest, children’s epistemological theories undergo a steep development from infancy to the first school years, which need not be reflected in their performance in standard false-belief tasks. Early studies by Shatz et al. () and Bartsch & Wellman () suggest that children start producing contrastives (i.e. verbal expressions of the difference between the contents of mental states and the corresponding reality) at age years, which is earlier than they start passing standard false-belief tasks. Likewise, the results of Dudley et al. () revealed that some -year-olds understand that ‘Sally doesn’t
know that the marble is in the basket’ presupposes (or at least conveys) that the marble is in the basket—contrary to how children of that age tend to respond to the false-belief question in the Sally-Anne task. As Sodian and colleagues put it, ‘it is possible that an awareness of others’ mental states emerges in communicative exchanges before it becomes available to predictions and explanations of others’ behaviours made from a third-person perspective’ (: ; for a related point, see Harris, Ronfard, & Bartz, ; Harris, Yang, & Cui, ). While experimental pragmatics tasks may reveal early Theory of Mind abilities that are not observable in standard false-belief tasks, they may also reveal shortcomings that Theory of Mind tasks do not: infants, toddlers, and young children show limitations in their understanding of other people’s knowledge in communication, which could be taken to contradict rich interpretations of infants’ successful performance in indirect false-belief tasks. Thus, the results of the studies reviewed in this chapter (and many others in the early Theory of Mind literature) have implications for theoretical accounts of Theory of Mind development. The first implication is that children’s epistemological theories are more complex and slow-developing than the results of indirect false-belief tasks might suggest. Therefore, in order to understand children’s belief reasoning capacities, we need to combine developmental data from traditional Theory of Mind tasks with experimental pragmatics studies where children need to use their developing concept of belief in interaction with others. The second implication is for specific theories of Theory of Mind development, which do not account for the evidence observed in experimental pragmatics studies. For example, Apperly & Butterfill (; Butterfill & Apperly, ; Low et al., ) have recently argued that humans possess two separate systems for belief reasoning: a simpler, evolutionary, and ontogenetically ancient system that operates fast and independently from central cognitive resources, and a more sophisticated system that develops later and operates more slowly, making substantial demands on executive-control processes. Rather than operating on belief representations, the more efficient system operates on registrations, which Butterfill & Apperly (; see also Rakoczy, ) define as a relation between an agent, an object, and a location. These non-mentalistic representations could in principle explain the patterns of results observed in early referential communication studies (e.g. O’Neill, ; Dunham et al., ; Tomasello & Haberl, ; Moll & Tomasello, ; Moll et al., , ), in which infants and toddlers seem to keep track of their interlocutor’s engagement (and disengagement) in a communicative situation. Likewise, low-level memory representations or registrations could also explain the improved performance of adults in Gorman et al.’s () investigation of common ground in referential communication. However, like other Theory of Mind researchers, Apperly, Butterfill and colleagues focus on the results of falsebelief tasks, but do not offer a satisfactory analysis of the development of young children’s epistemic theories reported in the experimental pragmatics literature. Theory of Mind researchers defending rich interpretations of infants’ and toddlers’ success in false-belief tasks (e.g. Carruthers, ; Helming et al., ) also need to explain the limitations that toddlers and preschoolers reveal in experimental pragmatics studies if we are to assume that infants’ concept of belief is as sophisticated and flexible as some of them argue (e.g. Scott & Baillargeon, ). In particular, they would need to explain why infants under years of age would be able to infer that when an agent sees an object being placed in a box, she comes to believe that the object is in that box (as it is assumed in falsebelief tasks; e.g., Onishi & Baillargeon, ) but -year-olds cannot apply the same
- ´
rationale when they have to select a knowledgeable helper (over an ignorant helper) on the basis of his visual access to the contents of a box (Povinelli & DeBlois, ; Sodian et al., ; see also Robinson et al., ). In conclusion, a comprehensive theory of the development of children’s epistemological theories needs to account not only for infants’ and children’s performance in false-belief tasks, but also for their appreciation of sensory perception as a source of beliefs, and their understanding of their interlocutors’ knowledge in communication, including the use and comprehension of mental state verbs. Social cognition researchers might argue that communication falls outside the scope of Theory of Mind research, and that it should therefore be investigated by psycholinguists and pragmatists. The problem with passing the buck in this fashion is that, in everyday life, children need to assess the epistemic states of their interlocutors (e.g. when asking and responding to questions) more often than they need to predict the behaviour of an agent with a false belief (as in the Sally-Anne task). Consequently, there is a danger that the bulk of Theory of Mind data will continue to fall outside of the scope of the theories that are designed to explain them.
A This chapter is based on a paper presented at the Summer School ‘Understanding Communication and Understanding Minds: The Role of Metarepresentation’ (Central European University, Budapest, June‒July ). I would like to thank all the students and the rest of the faculty at the summer school for a very stimulating discussion of this work, which helped me shape this chapter. This research was supported by a Young Research Talent Grant from the Research Council of Norway (Ref. ). I also gratefully acknowledge this funding.
......................................................................................................................
- ......................................................................................................................
. .
I their informal verbal exchanges people tend to follow the ‘one speaker at the time’ rule (Schegloff, ). The use of the term ‘turn-taking’ to describe the process in which this rule operates in human conversation is relatively recent, and attributed to Yngve () and Goffman () by Duncan (). Especially since the famous paper by Harvey Sacks, Emanuel Schegloff, & Gail Jefferson in the journal Language, which marks the birth of the sociological discipline now called Conversation Analysis (CA), turn-taking in conversation has attracted attention from a variety of disciplines. In this chapter, I will briefly summarize the main theoretical approaches and controversies regarding turn-taking, followed by some reflections on different ways it can be studied experimentally.
. T R
.................................................................................................................................. With certain caveats to be discussed later, the ‘one speaker at the time’ principle appears to be a universal property of human conversation (Stivers et al., ). It has been suggested that this is because of cognitive constraints. In the words of Duncan (), ‘it is difficult to maintain adequate mutual comprehensibility when participants in a conversation are talking at the same time’. While this is plausible and could well be one of the reasons for the principle, another important reason is that in conversation, contributions are usually shaped and given their social meaning by the previous utterance to which they are a response. This sequential dependency of contributions is only possible when contributions are ordered in time, as one cannot respond to a contribution that hasn’t at least started yet.1 In informal conversation there are no pre-assigned orderings of speakers or preallocated speech durations. The turn-taking system is a ‘locally managed’ system, meaning that the allocation of speaker roles happens on a turn-by-turn basis (Sacks et al., : ). This observation is also the basis for the passive-aggressive joking complaint ‘Oh, I’m sorry . . . did the middle of my sentence interrupt the beginning of yours?’ 1
. .
Before going into The Rules of the turn-taking system, it is important to note that a ‘turn’ is not the same as what linguists know as a ‘sentence’. The fundamental unit, the ‘atom’ of turn-taking is the Turn Construction Unit (TCU), which can encompass ‘sentential, clausal, phrasal, and lexical constructions’ (Sacks et al., : ). TCUs consisting of only one word are in fact very common. Note that I will use the words ‘turn’ and TCU interchangeably in this chapter. A speaker who ‘has the floor’ is entitled to produce one TCU, and the completion of that unit marks a so-called ‘Transition Relevance Place’ or TRP. When a TRP is reached, The Rules for next-speaker selection in Sacks et al. () can be summarized as follows. a) If the current speaker has selected a next speaker during their just completed TCU, then that selected speaker can (and should) take the next turn. b) If the current speaker has not selected a next speaker during their just completed TCU, the other participants to the conversation can ‘self-select’, that is, take the next turn. The first potential speaker that starts a TCU under this rule becomes the next speaker. c) If no speaker self-selects, the current speaker may take the next turn, but if another speaker self-selects, the current speaker must give way (examples in Sacks et al., footnote 14). Obviously, once the new speaker’s TCU ends at a TRP, The Rules apply again, etc. etc. The order of The Rules is important, and this has consequences for the timing. Rule c) is only invoked if it turns out that no speaker has self-selected, and rule b) only when there has not been a speaker selected as specified in a). This means that the time between the end of the previous turn and the beginning of the next is predicted to generally be shortest when rule a) applies, and longest for rule c) (see also Sacks et al., : , fn , referred to in Levinson, : ). The rule system ensures that even though turns are not allocated in advance, and the allocation is locally managed by the participants themselves, the proportion of gaps (silences between successive turns) and overlaps (two or more participants talking simultaneously) is minimized.
. I
.................................................................................................................................. There are a number of issues regarding the turn-taking system that are the subject of longstanding controversies. I will discuss the most central of these and the primary evidence that has been brought to bear upon them.
.. The status of The Rules The existence of or general adherence to the turn-taking rules is sometimes questioned. Anecdotal uncles or specific language communities are claimed not to observe The Rules, hence providing evidence against their existence. It is important to distinguish two types of scepticism regarding The Rules. The first type is based on the observation that overlapping speech or gaps between turns occur frequently, hence The Rules appear not to have been
-
observed. The second type agrees that rules are followed, but that these rules are not The Rules. To start with the first type of critique, the fundamental observation that overlapping speech and gaps do occur is not necessarily evidence against the operation of The Rules. There are many types of overlapping speech that are not treated by participants as ‘interruptive’. I will give two examples. First, there is what Schegloff () called a continuer (e.g. ‘uh-huh’). Yngve () called these brief utterances ‘backchannels’, although these represent a much larger category than Schegloff ’s continuers (see Duncan, : ). These are often produced in acoustic overlap with the current turn, but function as the opposite of an interruption, in the sense that the producer of the continuer displays understanding that the current contribution by the speaker is not yet complete. A second example is the situation that arises when the current speaker did not select the next speaker, and someone self-selects but starts her unit before the current speaker has completed his. This situation can either arise because the next speaker expected the unit by the current speaker to end earlier than it actually did, or because the next speaker wanted to be the first under rule b) to get the next turn. A comprehensive and lucid overview is provided by Drew () who systematizes and expands earlier seminal work by Jefferson and Schegloff (Jefferson, , , ; Schegloff, ). Importantly, automatic procedures to detect overlap on the basis of the speech energy envelope (e.g. Jaffé & Feldstein, ) are not able to distinguish between these ‘benign’ types of overlap and the ‘interruptive’ type. But even if we find overlapping speech that is genuinely interruptive, that is, not one of the cooperative types discussed by the authors cited earlier, and not starting at a TRP, it does not necessarily undermine the relevance of The Rules. It is important to realize that The Rules are not like physical laws in the sense that any violation automatically means a disconfirmation (like with the law of conservation of energy), but instead have a normative character. This means, among other things, that if it can be shown by the participants’ behaviour that they treat a violation as such, this confirms rather than falsifies the existence of the rule. Second, it means that the violation of the rule ‘does’ something; that is, the violation has in itself an interactive motivation. For instance, a delay in responding, creating a noticeable gap between turns, can signal a so-called non-preferred response (Atkinson & Drew, ; Pomerantz, ). The second type of scepticism regarding The Rules claims that for some language area (or culture) The Rules do not apply. One of the clearest such claims was made by Philips () about the Indigenous American people of the Warm Springs reservation in Oregon. On this account, lengthy pauses between questions and responses (and other first and second parts of adjacency pairs) are the default and therefore do not signal nonpreferredness. The second example is Reisman (), discussed extensively in Sidnell (), who claimed that in some Caribbean societies participants generally do not care whether talk is produced in overlap or not. Following an extensive investigation of recordings of Guyanese Creole, Sidnell concludes that participants are treating interruptive overlapping speech as rule violating, and exploit overlapping speech to perform certain interactive functions, as predicted by The Rules. A more recently claimed counterexample for The Rules is due to Meyer (), who studied Wolof village square meetings. During these meetings, multiple conversations take place at the same time, and the tolerance for overlapping speech is much higher than The Rules (or Schegloff, ) would predict. While this claim is interesting and deserves more
. .
extensive investigations, it appears to be the case that the village square meetings described in Meyer () are a special interactive genre that is different from informal conversation, in which multiple participants compete for attention in a public situation. These meeting documented by Meyer seem to be more reminiscent of public political debates without moderation. In the words of the author, Even the usual hierarchy of rules applying for speaker turns seems to be inverted: selfselection and current speaker continuing seem to override the ‘current speaker selects next speaker’ rule. Actively passing over the turn to a selected next speaker by the current speaker is rare, as true questions are in general, too. (Meyer, : )
Crucial for Meyer’s claim that The Rules are not in operation in Wolof conversation is whether non-official and informal gatherings among the Wolof (e.g. dinner conversations) have the same properties as the village square meetings. As far as I’m aware, there has not been a rigorous and systematic study of informal conversation that has provided solid evidence against The Rules, while there is evidence that suggests that they are universal. Stivers et al. () found that in ten different languages from a variety of language families, the gaps and overlaps between questions and answers were as small as The Rules would predict.
.. The processing of The Rules In the ‘ideal’ case, the turn-taking rules would produce no overlapping speech or gaps between turns. In actual practice this is of course not the case, but the system generally does seem to operate very well. A variable whose distribution and summary statistics provide a good assessment of the effectiveness of turn-taking is the Floor Transfer Offset or FTO (De Ruiter et al., ), which is the time of the beginning of the second turn minus the time of the end of the first turn. This value simultaneously provides information about both overlaps (negative values) and gaps (positive values). Other names that authors have given this variable are ‘delay from utterance end’ (Wesseling & Van Son, ), ‘between speaker intervals’ (Heldner & Edlund, ), or ‘response offset’ (Stivers et al., ). Most authors agree that the FTO distribution has an impressive amount of mass around zero, and estimates of the mode of the distribution converge on a value of around milliseconds (Levinson, ). This value is usually larger for face to face conversation than it is for audio-only (Ten Bosch et al., ) which may be one reason why the mean FTO for Dutch in De Ruiter et al. (), which was based on audio-only conversation, is much closer to zero than that of the Dutch data in Stivers et al. (). To account for this phenomenon, two very different types of model have been proposed. Duncan and colleagues (Duncan, ; Duncan & Fiske, ) put forth a model based on the exchange of signals. In line with popular theorizing in the s (Mehrabian, ; Archer & Akert, ), many of these signals were assumed to be nonverbal, specifically in the visual modality. In this approach, turn-taking is regulated by the speaker (who has the current turn) transmitting up to six ‘turn-yielding signals’ during the production of the turn, and the more of these signals that are transmitted, the higher the probability that the listener will take over the turn. This applies unless the speaker also transmits an
-
‘attempt-suppressing signal’ by performing a hand gesture. This suppressing signal effectively counteracts the turn-yielding signals, effectively stopping listeners from trying to take over the turn at that point. (Duncan, : , figure ). Duncan () further elaborated the system by proposing that a listener could signal her intention to be the next speaker by sending a ‘speaker state signal’. After recognizing a turn-yielding signal, and in the absence of a suppression signal, the listener can either produce a backchannel (Yngve, ) or a speaker state signal. The latter signal is either a) a ‘shift away in head direction’ from being pointed at the speaker, b) an audible inhalation, c) the initiation of a gesture, or d) increased loudness. This signal ‘has the advantage of being independent of the verbal content of the response, thus simplifying the process of recognizing the character of the auditor’s response’ (Duncan, : ). The mechanism suggested in Sacks et al. () is quite different. I write ‘suggested’ here because Sacks et al. make it clear that the details of the cognitive mechanisms involved in the turn-taking system are not their primary concern (Sacks et al., : ). Instead of relying on signals transmitted outside of the verbal exchanges, Sacks et al. assume that the coordination is accomplished by listeners who ‘project’ (another word for ‘predict’) an upcoming TRP in the current turn, before it has been fully completed. Projection is performed using the content of the utterance. Which aspects of their content is used for projection is another controversy that will be discussed in section ... This places much of the cognitive burden of the turn-taking process on the listener, who has to a) listen to the turn, b) predict when the next TRP will be, and c) prepare a matching response, so that when the TRP arrives, a coherent response can be produced promptly. There are a number of reasons why the model by Sacks et al. is to be preferred as the more plausible one. First, the planning, production, recognition, and interpretation of signals are not instantaneous processes, but take time. A simple button-press reaction to an auditory stimulus takes already ms or more, and ms for visual signals (Kosinski, ), and the signals specified in the turn-taking framework by Duncan and colleagues are arguably more ambiguous and less discrete than the signals used in such studies (simple beeps or light flashes). The fact that multiple signals are involved will potentially add even more processing time. The path in figure in Duncan () towards a smooth turn transition involves the exchange of four signals (involving both planning and execution by the sender, and perception and interpretation by the receiver), and the detection of the absence (which also takes time) of two more. In addition, the speech production system will have to initiate articulation, which takes in the order of a few hundreds of milliseconds as well (see also Magyari & De Ruiter, , and references therein). It is hard to imagine how such a system could produce the frequently occurring turn-transitions within ms that have been reported by the researchers cited earlier, especially when the relevant signals occur close to the end of a turn. Finally, it is also problematic for the signal model that it relies heavily on visual signals. This would predict that turn-taking would be harder to manage in telephone conversations than in face-to-face interactions, but if anything, the opposite appears to be true (Ten Bosch et al., ). Heldner & Edlund () have tried to reanimate the respond-to-signals approach and accordingly criticized the assumptions in the Sacks et al. model. They argued that since many (around %) FTO values in their corpus analysis are larger than ms, conversationalists might just have enough time to respond to signals at the end of turns (e.g. a
. .
pause). However, as mentioned earlier, this is implausible given what we know about the timing of the processes involved (Magyari & De Ruiter, ; Levinson & Torreira, ), and moreover it would not explain how the majority of turn transitions that are shorter than ms would work (Riest et al., ). Riest et al. () directly compared the predictions of the anticipation model and the ‘reaction to signal’ model. They had participants perform the turn end anticipation task from De Ruiter et al. () while being informed in advance how the turn in the upcoming trial would end. In another condition, participants had to anticipate the end of turns with their word order randomly scrambled, which makes prediction on the basis of lexico-syntactic content impossible. The response distribution to the natural, unaltered stimuli without advance information was very similar to the distribution with advance information, whereas the distribution for the scrambled word condition was essentially the same as a condition with white noise, in which participants have no choice but to respond to the end. This provides additional evidence that the anticipation model is generally correct, although it cannot be excluded that responding to the perceived end of a turn is a ‘backup strategy’ for cases where anticipation has failed (Riest et al., : ).
.. Cues for predicting the end of turns Predicting the end of a turn, as the Sacks et al. () model requires, is not a trivial task. The main problem is that the duration of turns is highly variable and not known in advance; it can even be shown that statistical knowledge about the duration of previously processed turns is not giving us the information we would need to predict the duration of future turns. Figure . is a histogram of the durations of the , turns from the conversations recorded by De Ruiter et al. ().
0.20
0.15
0.10
0.05
0
2
4
.. Distribution of turn duration
6
8
10
-
The horizontal axis represents the turn duration in seconds, and the vertical axis the proportion of turns with that duration. The smooth black line represents an exponential distribution with its single parameter fitted to the data. Apart from overestimating the proportion of turns with a near-zero duration and underestimating the frequency of very short turns, this distribution fits the turn duration distribution very well. The exponential distribution is the maximum entropy distribution with known mean and positive domain, meaning that it is theoretically not possible for a distribution of this type to contain less information. This means that distributional information conversationalists sample during the course of their lives is not helpful in predicting the duration of a particular turn. In more practical terms, this means that no matter how long a turn has already been going on, the conditional probability that it will be finished in the next moment remains constant. So the previous duration of an ongoing turn gives us no information at all about its remaining duration. Given that the distributional properties of turn duration are not informative, there must be other sources of information that listeners exploit to predict an upcoming TRP. About the source of information for this prediction there are two main contending hypotheses. The first is, as suggested by Sacks et al. () and also later by Auer (, ) that the cues are to be found in the lexico-syntactic (symbolic) content in the turn. The second is that the predictive cues are in the intonation,2 a hypothesis that has widespread appeal (e.g. Beattie et al., ; Local, Wells, & Sebba, ; Cutler & Pearson, ; Local, Kelly, & Wells, ; Couper-Kuhlen & Selting, ; Wells & Peppé, ; Caspers, ; Bögels & Torreira, ). Corpus studies have shown that the ends of turns often coincide with specific intonational contours and syntactic completion points (Caspers, ), and also with pragmatic completion points (Ford & Thompson, ). The tempting conclusion from these studies that therefore all these sources of information are used is not warranted, because a correlation does not allow us to draw conclusions about causality. Hence experimental evidence is essential. Schaffer () took fragments from conversations and applied lowpass filtering to them, making them unintelligible but preserving their intonation, and asked participants to judge whether these fragments were turn-final or not. She found that her participants were more consistent in their responses when they had access to the lexicosyntactic information. Cutler & Pearson () used fragments read from dialogue scripts in a similar task, and found some evidence for rising or falling intonational contours indicating turn-finality, matching an earlier discovery that Margaret Thatcher was often interrupted because of people misinterpreting her sharp pitch drop as a signal that she had finished her turn (Beattie et al., ). These experiments were off-line, meaning that participants were not under pressure to respond immediately. This lowers the ecological validity (Brewer, ) because in real life conversationalists do not have the time to contemplate at leisure whether a turn has ended or not. Also, Cutler and colleagues used stimuli read from scripts. These are usually quite different in speech rate and prosody (including intonation; see Tonhauser, Chapter in At this point, it is important to make a careful distinction between the terms ‘intonation’ and ‘prosody’. I will follow the convention here that prosody is a general term encompassing phenomena like timing, intonation, stress, etc. whereas intonation refers only to the fundamental frequency (F, pitch) of the unfolding speech. For instance, a potential turn-taking cue that is prosodic and not intonational is final lengthening (Oiler, ; Cutler & Pearson, ). 2
. .
this volume) from the natural speech that conversationalists produce and hear. To address these challenges, De Ruiter et al. () used an on-line task, and natural stimuli culled from recorded informal audio-only conversations. The task for the participants was to indicate by pressing a button the moment at which they anticipated the presented turn would end. The turns were either natural (as is), low-pass filtered (removing lexicosyntactic information but leaving the intonation intact), pitch-flattened (removing the pitch information but leaving the lexico-syntactic information in), or pitch-flattened and low-pass filtered, leaving only rhythmic information in. They found that removing pitch did not influence the (impressively accurate) anticipation performance relative to the natural condition. The low-pass filtered fragments led to a strong deterioration of the predictive performance. These combined results suggest that intonation is not necessary for good predictive performance, whereas lexico-syntactic information is. But De Ruiter et al. () also found evidence in their response distributions that having access to the intonation did improve participants’ performance when they did not have access to the lexico-syntactic information. This is in accordance with a gating study by Grosjean & Hirt () who concluded that ‘It is only when higher-level information is no longer present . . . that prosodic information is made available or is called into play’ (Grosjean & Hirt, : ). De Ruiter et al. () also found that sometimes a steep rise or fall in the intonation of natural fragments could ‘fool’ the participants into responding too early (De Ruiter et al., : ). This is a phenomenon that has also been documented in the studies by Cutler and colleagues mentioned earlier, and has been confirmed experimentally by Bögels & Torreira () who showed that intonational cues can trigger or delay responses. Given this evidence, it appears that symbolic speech content is the primary source of the information used in the anticipation of turn endings. This raises the question of whether syntactic or semantic factors are more relevant. These two aspects of utterances are notoriously hard to separate, but Riest et al. () approximated this separation by lowpass filtering either the open or the closed class words in the presented fragments. The resulting response distributions suggested that both syntactic and semantic information are necessary, but that semantic information (operationalized as the open class words) was more important than the syntactic information in the closed class words. Finally, recent studies with children have shed light on the developmental aspect of the use of these predictive cues. Keitel et al. () used eye-tracking to record turn-taking anticipation by participants looking at the next speaker. This allowed them to compare children aged , , , and months, and adults. They used both natural and pitchflattened speech. They found that only the -year-olds and the adults could reliably anticipate speaker role transition, and while the adults were equally good at anticipating speaker changes with and without the intonation, the -year-olds were clearly better with the intonation, suggesting that their syntactic development was not yet complete and the intonation provided additional and (for them) non-redundant syntactic cues. Using a similar eye-tracking based method, Lammertink et al. () had .‑year‑olds and adult controls listen to and look at video recordings of dialogues in which the prosodic and syntactic completeness were independently varied. The found that the .‑year‑olds could anticipate speaker changes, and they were using both intonation (which they call prosody) and syntactic information, but relying more on the syntactic cues than on the intonational cues.
-
Concluding this section, the evidence collected so far suggests that lexico-syntactic content is the principal information source for anticipating turn ends, and that intonation can in some situations also be exploited as an additional predictive cue. The major predictive work, however, is done by exploiting the lexico-syntactic information in the speech. It is also this information that allows listeners to generate early guesses about the speech act or communicative intention of the unfolding utterance, which could improve the prediction of the end of turn itself, as well as helping the listener prepare an appropriate next turn (Gisladottir et al., ; De Ruiter & Cummins, ; Levinson, ).
. R -
.................................................................................................................................. Investigating turn-taking experimentally is not straightforward, and requires a careful consideration of priorities regarding different types of validity. In this section, I will discuss a number of methodologies (both experimental and non-experimental) that have been used in turn-taking research and address their relative weaknesses and strengths.
.. Conversation analysis Carefully analysing detailed transcripts of recordings of natural conversation provides a shared record of the central phenomena involved in turn-taking. Researchers of turntaking coming from (psycho-)linguistics or psychology are advised to become familiar with the vast literature available in the discipline of CA, and if possible to engage in CA ‘data sessions’ (Harris et al., ) to get an appreciation of the methodology of CA, as well as the underlying orderliness of natural conversation (see Levinson, , for an accessible introduction to the principles of CA). Whatever one’s research question regarding turntaking, it is always a good idea to look at the phenomenon of interest in natural data, to see it operating in its ‘natural environment’.
.. Corpus analysis There are an increasing number of conversational corpora available, many of them enriched with detailed orthographic annotations. Examples are the British National Corpus (Burnard, ), the Switchboard Corpus (Godfrey et al., ), and the Dutch ‘Corpus Gesproken Nederlands’ (Oostdijk, ). It is important to be aware of the social context of the recordings. For instance, some corpora that have been used to study turn-taking (e.g. Caspers, ) contain transcriptions of task-based dialogues such as the ‘Map Task’ (Anderson et al., ) which have known and unknown consequences for the vocabulary, the types of speech act performed, and also for the naturalness of the dialogues (see the discussion in Healey et al., : ). Usually, corpus analyses involve the systematic quantification of codings, which seems straightforward, but can give rise to some hairy theoretical problems (Schegloff, ) for which it is not easy to find satisfactory solutions.
. .
Corpus analysis can reveal statistical patterns (e.g. correlations, differences in frequencies) but the method is not suitable for testing causal explanations. This requires the use of random controlled experimentation (see De Ruiter, , for a more extensive discussion).
.. Off-line experimental methods Experiments that allow participants to ‘judge’ aspects of presented stimuli (e.g. turns) without time pressure are known as off-line experiments in psycholinguistics. While this methodology can sometimes be useful (Schaffer, ; Cutler & Pearson, ; Grosjean & Hirt, ; Magyari & De Ruiter, ), and allows for a large class of research questions to be investigated in true experimental designs, it should be borne in mind that, in real-life turn-taking, the listener in particular is under severe time pressure and obliged to perform multiple tasks at the same time. This means that results from offline experiments should be interpreted with caution. If participants judge a turn fragment to be finished, after thinking consciously about it for several seconds, that does not mean that they would make the same decision while automatically processing natural speech that is coming in at a rate of several words per second.
.. On-line experimental methods The potential disadvantages of off-line experimentation are partially avoided by so-called on-line experimental methods. Here participants are asked to respond immediately, and their recorded response takes place with a speed that is more comparable to real-life processing. Examples of this methodology are the turn-anticipation tasks based on button press responses (De Ruiter et al., ; Bögels & Torreira, ; Riest et al., ), vocal responses (Wesseling & Van Son, ) and the use of eye-tracking (Keitel et al., ; Lammertink et al., ). It is also possible to use neurocognitive approaches to investigate turn-taking processes by combining electrophysiological recordings with the vocal or button press anticipation task, either studying the readiness potential (Jansen et al., ) or using frequency spectrum analysis (Magyari et al., ). fMRI methods are currently not very useful for the study of timing, as they lack the necessary temporal resolution.
. V
.................................................................................................................................. In experimentation involving human participants, it is always difficult to find the optimal trade-off between internal validity and ecological validity. Internal validity is the degree to which the conclusions from a study logically follow from the observed findings. The most difficult part in obtaining high internal validity is excluding alternative explanations, meaning that as many potentially relevant variables as possible should be controlled. These variables should be kept equal, balanced, or sufficiently randomized. However, the more experimental control there is, the higher the risk is that the resulting experimental task is very different from the situation in real life that we ultimately want to study. This
-
threatens the ecological validity (see Brewer, ). So researchers interested in turn-taking usually have to choose between different evils. For example, a non-experimental study based on transcripts of natural data has a very high ecological validity and a very low internal validity. The reverse is true for the so-called ‘lexical decision task’, which is very popular in psycholinguistics. One can control almost everything in a lexical decision experiment, but in our daily lives we rarely, if ever, encounter the problem of having to establish with great urgency whether the word our interlocutor just uttered is actually an existent word in our shared native language. Finding the right trade-off between these two equally important forms of validity is especially challenging in studying human interaction (De Ruiter, ; De Ruiter & Albert, ). In turn‑taking research, this is especially relevant when it comes to choosing stimulus materials. The temptation is strong to construct materials which are then recorded from people reading or acting out the constructed fragments, as it allows for perfect control over at least the lexico-syntactic content. The problem, however, is that read speech is very different from spontaneous speech, precisely in those aspects that many turn-taking researchers want to study. In the case of intonation, as L. E. de Ruiter (: ) summarizes, read speech has been found to have a wider pitch range (e.g., Eskénazi, ), steeper slopes in F declination (Swerts & Heldner, ) and to contain fewer rising movements than spontaneous speech. (Blaauw, ; Silverman et al., )
Note that precisely these phenomena that are typical of read speech have been claimed by researchers to play an important role in turn anticipation (e.g. Bögels & Torreira, ). But using read speech as stimuli fragments will result in cues that are normally not present, much less frequent, or substantially less salient in natural conversation. On the other hand, there are situations, for instance in developmental turn-taking research (Keitel et al., ; Lammertink et al., ) where using natural speech fragments may be impractical, for instance because the young participants may not be able to fully process natural speech by adults.
. S
.................................................................................................................................. The turn-taking system represents the cognitive hub between social interaction and language processing. It is a locally managed and highly flexible system that enables sequentially ordered exchanges between multiple participants, even though the duration of turns is a priori highly variable and unpredictable. There are a number of longstanding controversies in the field. Are The Rules put forth by Sacks et al. () a universal property of human interaction? The evidence so far suggests that they may be. Is the turn-taking system controlled by the mutual exchange of signals, or by listeners’ anticipation of TRPs? There are strong theoretical and empirical arguments for the latter position. Which cues do listeners use for this anticipation? Primarily the symbolic content, but intonational and other prosodic cues play a role too. When are children able to anticipate turn ends at approximately adult levels of performance? Current estimates are between . and years.
. .
The experimental approach to turn-taking is essential for investigating causal hypotheses, but complicated by the fundamentally interactive context in which turn-taking normally takes place. Specifically, finding the right balance between internal validity (e.g. issues of control) and ecological validity can be very challenging. For this reason, it is of central importance to use natural materials whenever possible. This holds for both the study of the relevant phenomena by inspecting transcripts (e.g. using the methodology of CA), and for the materials used in experimentation. The fact that artificial systems (e.g. interactive agents) are still very bad at turn-taking3 is indicative of how little we know about it. We have barely begun to scratch the surface of this complex system, and the number of open research questions is large. Studying turntaking also requires strong interdisciplinary collaborations between sociology, cognitive psychology, linguistics, and, last but not least, experimental pragmatics.
3 Some systems can achieve good predictive turn-taking behaviour in highly limited domains, for example, Skantzen & Schlangen ().
R
..........................................
Abbott, Barbara (). The Indefiniteness of Definiteness. New York: Springer. Abkarian, G. G. (). Comprehension of deictic locatives: the object ‘behind’ it. Journal of Psycholinguistic Research, (): –. Abrusán, Márta (). Predicting the presuppositions of soft triggers. Linguistics and Philosophy, (): –. doi: ./s---y Abrusán, Márta (). Presupposition cancellation: Explaining the ‘soft-hard’ trigger distinction. Natural Language Semantics [Advance Access]. doi: ./s--- Abrusán, Márta and Szendrői, Kriszta (). Experimenting with the king of France: topics, verifiability, and definite descriptions. Semantics and Pragmatics, (): –. Abusch, Dorit (). Lexical alternatives as a source of pragmatic presuppositions. In Brendan Jackson (ed.), Proceedings of Semantics and Linguistic Theory (SALT), , –. Ithaca, NY: Cornell University. Abusch, Dorit (). Presupposition triggering from alternatives. Journal of Semantics, (): –. doi: ./jos/ffp Achimova, Asya, Syrett, Kristen, Musolino, Julien, and Déprez, Viviane (). Children’s developing knowledge of wh-/quantifier question-answer relations. Language Learning and Development, : –. Acredolo, Linda P. (). Development of spatial orientation in infancy. Developmental Psychology, (): –. doi: ./-... Acredolo, Linda P. (). Laboratory versus home: the effect of environment on the -month-old infant’s choice of spatial reference system. Developmental Psychology, (): –. doi: ./ -... Acredolo, Linda P. (). The familiarity factor in spatial research. New Directions for Child and Adolescent Development, : –. doi: ./cd. Acredolo, Linda P. and Evans, Debra (). Developmental changes in the effects of landmarks on infant spatial behavior. Developmental Psychology, : –. doi: ./-... Aguiar, Andréa and Baillargeon, Renée (). Eight-and-a-half-month-old infants’ reasoning about containment events. Child Development, (): –. doi: ./j.-..tb.x Akhtar, Nameera and Tomasello, Michael (). The social nature of words and word learning. In Roberta Michnick Golinkoff et al. (eds.), Becoming a Word Learner: A Debate on Lexical Acquisition. Oxford: Oxford University Press. doi: ./acprof:oso/.. Alais, David and Burr, David (). The ventriloquist effect results from near-optimal bimodal integration. Current Biology, (): –. doi: ./j.cub... Alexandropoulou, Stavroula (). Testing the nature of variation effects with modified numerals. In Eva Csipak and Hedde Zeijlstra (eds.), Proceedings of Sinn und Bedeutung . Gottingen: Universität Göttingen. –. Alexandropoulou, Stavroula, Dotlačil, Jakub, McNabb, Yaron, and Nouwen, Rick (). Pragmatic inferences with numeral modifiers: novel experimental data. In Sarah D’Antonio, Mary Moroney, and Carol-Rose Little (eds.), Proceedings of the th Semantics and Linguistic Theory Conference. Washington, DC: Linguistic Society of America. –. Alexandropoulou, Stavroula, Dotlačil, Jakub, and Nouwen, Rick (). At least ignorance inferences come at a processing cost: support from eye movements. In Mary Moroney, Carol-Rose Little, Jacob Collard, and Dan Burgdorf (eds.), Proceedings of the th Semantics and Linguistic Theory Conference. Ithaca, NY: CLC Publications. –.
Allen, Shanley, Özyürek, Azlı, Kita, Sotaro et al. (). Language-specific and universal influences in children’s syntactic packaging of manner and path: a comparison of English, Japanese, and Turkish. Cognition, : –. doi: ./j.cognition... Almor, Amit (). Noun-phrase anaphora and focus: the informational load hypothesis. Psychological Review, : –. doi: ./-X... Almor, Amit and Nair, Veena A. (). The form of referential expressions in discourse. Language and Linguistics Compass, : –. Aloni, Maria (). Free choice in modal contexts. In Matthias Weisgerber (ed.), Proceedings of Sinn und Bedeutung . Konstanz: University of Konstanz, Vol. of Arbeitspapiere des Fachbereichs Sprachwissenschaft. –. Aloni, Maria and Port, Angelika (). Epistemic indefinites and methods of identification. In Luis Alonso-Ovalle and Paula Menéndez-Benito (eds.), Epistemic Indefinites. Oxford: Oxford University Press. Alonso-Ovalle, Luis (). Counterfactuals, correlatives, and disjunction. Linguistics and Philosophy, : –. Altmann, Gerry T. M. (). Ambiguity in sentence processing. Trends in Cognitive Sciences, (): –. Altmann, Gerry T. M. and Kamide, Yuki (). Incremental interpretation at verbs: restricting the domain of subsequent reference. Cognition, (): –. Altmann, Gerry T. M. and Steedman, Mark (). Interaction with context during human sentence processing. Cognition, (): –. doi: ./–()– Altmann, Stuart A. (). The structure of primate social communication. In Stuart A. Altmann (ed.) Social Communication among Primates. Chicago, IL: University of Chicago Press. –. Alxatib, Sam (). Denying bivalence and the super/sub-valuationary divide. Talk delivered in the Vagueness and Similarity workshop, ENS Paris, – May. Alxatib, Sam and Pelletier, Francis Jeffry (a). The psychology of vagueness: borderline cases and contradictions. Mind and Language, : –. Alxatib, Sam and Pelletier, Jeff (b). On the psychology of truth-gaps. In Rick Nouwen, Robert van Rooij, Uli Sauerland, and Hans-Christian Schmitz (eds.), Vagueness in Communication. Heidelberg: Springer. –. Alxatib, Sam, Pagin, Peter, and Sauerland, Uli (). Acceptable contradictions: pragmatics or semantics? A reply to Cobreros et al. Journal of Philosophical Logic, : –. Amanzio, Martina, Geminiani, Giuliano, Leotta, Daniela, and Cappa, Stefano (). Metaphor comprehension in Alzheimer’s disease: novelty matters. Brain and Language, (): –. Amaral, Patrícia and Cummins, Chris (). A cross-linguistic study on information backgrounding and presupposition projection. In Florian Schwarz (ed.), Experimental Perspectives on Presuppositions. Cham: Springer International Publishing. –. Ambady, Nalini, Koo, Jasook, Lee, Fiona, and Rosenthal, Robert (). More than words: linguistic and nonlinguistic politeness in two cultures. Journal of Personality and Social Psychology, : –. Ames, Louise Bates and Learned, Janet (). The development of verbalized space in the young child. Journal of Genetic Psychology, : –. doi: ./.. AnderBois, Scott, Brasoveanu, Adrian, and Henderson, Robert (). The pragmatics of quantifier scope: a corpus study. In Ana Aguilar Guevara, Anna Chernilovskaya, and Rick Nouwen (eds.), Proceedings of Sinn und Bedeutung . Cambridge, MA: MIT Working Papers in Linguistics. –. Andersen, Elaine S. (). Lexical universals of body-part terminology. In Joseph H. Greenberg, Charles A. Ferguson, and Edith A. Moravcsik (eds.), Universals of Human Language, Vol. . Stanford, CA: Stanford University Press. –. Anderson, Anne, Garrod, Simon, and Sanford, Anthony I. (). The accessibility of pronominal antecedents as a function of episode shifts in narrative text. Quarterly Journal of Experimental Psychology, : –. doi: ./ Anderson, Anne H., Bader, Miles, Gurman Bard et al. (). The HCRC Map Task Corpus. Language and Speech, (): –. doi: ./
Anderson, Catherine (). The structure and real-time comprehension of quantifier scope ambiguity. PhD thesis. Evanston, IL: Northwestern University. Andonova, Elena, Tenbrink, Thora, and Coventry, Kenny R. (). Function and context affect spatial information packaging at multiple levels. Psychonomic Bulletin and Review, : –. doi: ./PBR... Annaz, Dagmara, Van Herwegen, Jo, Thomas, Michael et al. (). Comprehension of metaphor and metonymy in children with Williams syndrome. International Journal of Language and Communication Disorders, (): –. Antoniou, Kyriakos, Cummins, Chris, and Katsos, Napoleon (). Why only some adults reject under-informative utterances. Journal of Pragmatics, : –. Aone, Chinatsu (). Resolution of collective-distributive ambiguity using model-based reasoning. In Douglas E. Appelt (ed.), Proceedings of the th Annual Meeting of the Association for Computational Linguistics. –. Berkeley, CA. Aparicio, Helena, Xiang, Ming, and Kennedy, Christopher (). Processing gradable adjectives in context: A visual world study. In Sarah D’Antonio, Mary Moroney, and Carol Rose Little (eds.), Proceedings of the th Semantics and Linguistic Theory Conference. Washington, DC: Linguistic Society of America. Apperly, Ian A. and Butterfill, Stephen A. (). Do humans have two systems to track beliefs and belief-like states? Psychological Review, (): –. doi: ./a Apresjan, Juri D. (). Regular polysemy. Linguistics, : –. Archer, Dane and Akert, Robin M. (). Words and everything else: verbal and nonverbal cues to social interpretation. Journal of Personality and Social Psychology, : –. doi: ./... Ariel, Mira (). Accessing Noun-Phrase Antecedents. London: Routledge. Ariel, Mira (). Accessibility theory: an overview. In Sanders, Ted, Schliperoord, Joost, and Spooren, Wilbert (eds.), Text Representation. Amsterdam: John Benjamins. –. Ariel, Mira (). Most. Language, (): –. Arnold, Jennifer E. (). Reference form and discourse patterns. PhD thesis. Stanford University. Arnold, Jennifer E. (). The effect of thematic roles on pronoun use and frequency of reference continuation. Discourse Processes, (): –. doi: ./SDP_ Arnold, Jennifer E. (). How speakers refer: the role of accessibility. Language and Linguistics Compass, : –. Arnold, Jennifer E. (). Women and men have different discourse biases for pronoun interpretation. Discourse Processes, : –. doi: ./X.. Arnold, Jennifer E. and Griffin, Zenzi M. (). The effect of additional characters on choice of referring expression: everyone counts. Journal of Memory and Language, : –. doi: ./j. jml... Arnold, Jennifer E. and Lao, Shin-Yi C. (). Put in last position something previously unmentioned: word order effects on referential expectancy and reference comprehension. Language and Cognitive Processes, (): –. doi: ./ Arnold, Jennifer E. and Lao, Shin-Yi C. (). Effects of psychological attention on pronoun comprehension. Language, Cognition and Neuroscience, (), –. doi: ./.. Arnold, Jennifer E., Eisenband, Janet G., Brown-Schmidt, Sarah, and Trueswell, John C. (). The immediate use of gender information: eyetracking evidence of the time-course of pronoun resolution. Cognition, : B–B. Arnold, Jennifer E., Tanenhaus, Michael K., Altmann, Rebecca J., and Fagnano, Maria (). The old and thee, uh, new: disfluency and reference resolution. Psychological Science, (): –. doi: ./j.-.. Arnold, Jennifer E., Hudson Kam, Carla L., and Tanenhaus, Michael K. (). If you say thee uh you are describing something hard: the on-line attribution of disfluency during reference comprehension. Journal of Experimental Psychology: Learning, Memory, and Cognition, (): –. doi: ./-...
Arnold, Jennifer E., Bennetto, Loisa, and Diehl, Joshua J. (). Reference production in young speakers with and without autism: effects of discourse status and processing constraints. Cognition, (): –. doi: ./j.cognition... Arnold, Jennifer E., Kaiser, Elsi, Kahn, Jason M., and Kim, Lucy K. (). Information structure: linguistic, cognitive, and processing approaches. WIREs Cognitive Science, : –. doi: ./ wcs. Arts, Anja (). Overspecification in instructive texts. PhD thesis. Tilburg University. Wolf Publishers: Nijmegen. Arts, Anja, Maes, Alfons, Noordman, Leo G. M., and Jansen, Carel (a). Overspecification in written instruction. Linguistics, (): –. Arts, Anja, Maes, Alfons, Noordman, Leo G. M., and Jansen, Carel (b). Overspecification facilitates object identification. Journal of Pragmatics, , –. doi: ./j.pragma... Arundale, Robert B. (). Face as relational and interactional: a communication framework for research on face, facework, and politeness. Journal of Politeness Research, (): –. Arzouan, Yossi, Goldstein, Abraham, and Faust, Miriam (). Brainwaves are stethoscopes: ERP correlates of novel metaphor comprehension. Brain Research, : –. doi: ./j. brainres... Asch, Solomon E. and Nerlove, Harriet (). The development of double function terms in children: an exploratory investigation. In Bernard Kaplan and Seymour Wapner (eds.), Perspectives in Psychological Theory: Essays in Honor of Heinz Werner. New York: International Universities Press. –. Asher, Nicholas (). Reference to Abstract Objects in Discourse. Boston: Kluwer Academic. Asher, Nicholas and Lascarides, Alex (). Logics of Conversation. Cambridge: Cambridge University Press. Asher, Nicholas and Morreau, Michael (). What some generic sentences mean. In Greg Carlson and Francis J. Pelletier (eds.), The Generic Book. Chicago: Chicago University Press. –. Ashworth, Jennifer E. (). Metaphor and the logicians from Aristotle to Cajetan. Vivarium, : –. Athanasopoulos, Panos and Bylund, Emanuel (). Does grammatical aspect affect motion event cognition? A cross-linguistic comparison of English and Swedish speakers. Cognitive Science, (): –. doi: ./cogs. Athanasopoulos, Panos, Bylund, Emanuel, Montero-Melis, Guillermo, et al. (). Two languages, two minds: flexible cognitive processing driven by language of operation. Psychological Science, (): –. doi: ./ Atkinson, J. Maxwell and Drew, Paul (). Order in Court: The Organization of Verbal Behaviour in Judicial Settings. London: Macmillan. Atlas, Jay D. and Levinson, Stephen C. (). It-clefts, informativeness, and logical form. In Peter Cole (ed.), Radical Pragmatics. New York: Academic Press. –. Au, Terry Kit-fong (). A verb is worth a thousand words: the causes and consequences of interpersonal events implicit in language. Journal of Memory and Language, : –. Auer, Peter (). On the prosody and syntax of turn-continuations. In Elizabeth Couper-Kuhlen and Margaret Selting (eds.), Prosody in Conversation. Cambridge: Cambridge University Press. Auer, Peter (). Projection and minimalistic syntax in interaction. Discourse Processes, : –. Augurzky, Petra and Franke, Michael (). Investigating the time course of pragmatic expectations—ERP evidence on German implicature processing. Poster presented at AMLaP , Bilbao. Autry, Kevin S. and Levine, William H. (). Activation of negated and non-negated entities. Journal of Pragmatics, (): –. doi: ./j.pragma... Ayers, Gayle M. (). Nuclear accent types and prominence: some psycholinguistic experiments. PhD thesis. The Ohio State University. Bach, Kent (). Thought and Reference. Oxford: Clarendon Press.
Bacovcin, Hezekiah and Schwarz, Florian (). Local accommodation and presupposition trigger class: results from the covered box task. Poster presented at the meeting of the Linguistic Society of America. Bacovcin, Hezekiah, Zehr, Jérémy, and Schwarz, Florian (). To accommodate or to ignore? the presuppositions of ‘again’ and ‘continue’ across contexts. Ms., University of Pennsylvania. Bade, Nadine (). Obligatory implicatures and the presupposition of ‘too’. In Urtzi Etxeberria, Anamaria Fălăus, Aritz Irurtzun, and Bryan Leferman (eds.), Proceedings of Sinn und Bedeutung . Vitoria-Gasteiz: University of the Basque Country. –. Bade, Nadine (). Obligatory presupposition triggers in discourse—Empirical foundations of the theories maximize presupposition and obligatory implicatures. PhD thesis. University of Tübingen. Baggio, Giosuè, Choma, Travis, Van Lambalgen, Michiel, and Hagoort, Peter (). Coercion and compositionality. Journal of Cognitive Neuroscience, (): –. Baillargeon, Renée, Needham, Amy, and Devos, Julie (). The development of young infants’ intuitions about support. Early Development and Parenting, : –. doi: ./edp. Baker, C. Lee (). Double negatives, Linguistic Inquiry, : –. Baker, Mark C. (). Thematic roles and syntactic structure. In Liliane Haegeman (ed.), Elements of Grammar: Handbook in Generative Syntax. Springer Netherlands. –. Baldwin, Dare A. (). Infants’ ability to consult the speaker for clues to word reference. Journal of Child Language, : –. doi: ./S Bale, Alan Clinton (). A universal scale of comparison. Linguistics and Philosophy, : –. Bale, Alan Clinton (). Scales and comparison classes. Natural Language Semantics, : –. Bale, Alan and Barner, David (). Grammatical alternatives and pragmatic development. In Anamaria Fălăus (Ed.), Alternatives in Semantics. London: Palgrave Macmillan. –. Baltin, Mark R. (). Do antecedent-contained deletions exist? Linguistic Inquiry, : –. Bambini, Valentina and Resta, Donatella (). Metaphor and experimental pragmatics: when theory meets empirical investigation. Humana.Mente—Journal of Philosophical Studies, : –. Bambini, Valentina, Ghio, Marta, Moro, Andrea, and Schumacher, Petra B. (). Differentiating among pragmatic uses of words through timed sensicality judgments. Frontiers in Psychology, : . doi: ./fpsyg.. Bambini, Valentina, Resta, Donatella, and Grimaldi, Mirko (). A dataset of metaphors from the Italian literature: exploring psycholinguistic variables and the role of context. PLoS ONE, (). e. doi: ./journal.pone. Bambini, Valentina, Canal, Paolo, Resta, Donatella, and Grimaldi, Mirko (submitted). Time-course and neurophysiological underpinnings of metaphor in literary context. Bard, Ellen G., Robertson, Dan, and Sorace, Antonella (). Magnitude estimation of linguistic acceptability. Language, (): –. Barker, Chris (). Free choice permission as resource-sensitive reasoning. Semantics and Pragmatics, (): –. Barner, David and Bachrach, Asaf (). Inference and exact numerical representation in early language development. Cognitive Psychology, (): –. Barner, David and Snedeker, Jesse (). Compositionality and statistics in adjective acquisition: -year-olds interpret tall and short based on the size distributions of novel noun referents. Child Development, : –. Barner, David, Thalwitz, Dora, Wood, Justin, Yang, Shu-Yu, and Carey, Susan (). On the relation between the acquisition of singular-plural morpho-syntax and the conceptual distinction between one and more than one. Developmental Science, : –. Barner, David, Chow, Katherine, and Yang, Shu-Ju (). Finding one’s meaning: a test of the relation between quantifiers and integers in language development. Cognitive Psychology, : –. Barner, David, Brooks, Neon, and Bale, Alan (). Accessing the unsaid: the role of scalar alternatives in children’s pragmatic inferences. Cognition, (): –.
Barner, D., L. Hochstein, M. Rubenson, and A. Bale (). Four-year-old children compute scalar implicatures in absence of epistemic reasoning. In Kristen Syrett and Sudha Arunachalam (eds.), Semantics in Language Acquisition. Amsterdam: Benjamins. pp. –. Barnes, Jennifer L. (). Fiction, imagination, and social cognition: insights from autism. Poetics, : –. Baron-Cohen, Simon, Leslie, Alan M., and Frith, Uta (). Does the autistic child have a ‘theory of mind’? Cognition, (): –. Baron-Cohen, Simon, Wheelwright, Sally, Skinner, Richard, Martin, Joanne, and Cluble, Emma (). The autism-spectrum quotient (AQ): evidence from Asperger syndrome/high-functioning autism, males and females, scientists and mathematicians. Journal of Autism and Developmental Disorders, (): –. Barr, Dale J. (). Pragmatic expectations and linguistic evidence: listeners anticipate but do not integrate common ground. Cognition, (): –. Barr, D. and Keysar, B. (). Anchoring comprehension in linguistic precedents. Journal of Memory and Languages, : –. Barsalou, Lawrence W. (). Perceptual symbol systems. Behavioral and Brain Sciences, (): –. Bartels, Christine and Kingston, John (). Salient pitch cues in the perception of contrastive focus. In Peter Bosch and Rob van der Sandt (eds.), Focus and Natural Language Processing (IBM Working Papers on Logic and Linguistics, v.). Columbia: University of Missouri. –. Bartlett, Elsa J. (). Sizing things up: the acquisition of the meaning of dimensional adjectives. Journal of Child Language, (): –. Barton, Dagmar (). Generic noun phrases in German-French bilinguals: on the use of the definite article in adult heritage speakers. PhD thesis. University of Hamburg. Barton, Dagmar, Kolb, Nadine, and Kupisch, Tanja (). Definite article use with generic reference in German: an empirical study. Zeitschrift für Sprachwissenschaft, (): –. Bartsch, Karen and Wellman, Henry M. (). Children Talk About the Mind. Oxford: Oxford University Press. Bartsch, Renate and Vennemann, Theo (). Semantic Structures: A Study in the Relation Between Syntax and Semantics. Frankfurt: Athenäum Verlag. Barwise, Jon and Cooper, Robin (). Generalized quantifiers and natural language. Linguistics and Philosophy, : –. Bauman, Isabelle (). The representational validity of a theory of politeness. Paper presented at the International Communication Association Convention, New Orleans, LA. Baumann, Stefan and Hadelich, Kerstin (). Accent type and givenness: an experiment with auditory and visual priming. In Maria-Josep Solé, Daniel Recasens, and Joaquin Romero (eds.), Proceedings of the th International Congress of Phonetic Sciences. Barcelona: ICPhS. –. Baumann, Stefan, Grice, Martine, and Steindamm, Susanne (). Prosodic marking of focus domains: categorical or gradient? Proceedings of Speech Prosody . Dresden: Speech Prosody. – Baxter, Leslie A. (). An investigation of compliance-gaining as politeness. Human Communication Research, (): –. doi: ./j.-..tb.x Bayer, Samuel (). The size of events. In Teresa Galloway and Justin Spence (eds.), Proceedings of Semantics and Linguistic Theory . Ithaca, NY: Cornell University Press. –. doi: ./salt.vi. Beattie, Geoffrey W., Cutler, Anne, and Pearson, Mark (). Why is Mrs. Thatcher interrupted so often? Nature, : –. doi: ./a Beaver, David (). When variables don’t vary enough. In Mandy Harvey and Lynn Santelmann (eds.), Proceedings of Semantics and Linguistic Theory IV. Ithaca, NY: CLC Publications. –. Beaver, David (). Presupposition and Assertion in Dynamic Semantics. Stanford, CA: CSLI Publications. Beaver, David (). Have you noticed that your belly button lint colour is related to the colour of your clothing? In Rainer Bäuerle, Uwe Reyle, and Thomas Ede Zimmerman (eds.), Presuppositions and Discourse: Essays Offered to Hans Kamp. Leiden: Brill.
Beaver, David and Clark, Brady (). Sense and Sensitivity: How Focus Determines Meaning. Oxford: Wiley-Blackwell. Beaver, David and Geurts, Bart (). Presuppositions. In Claudia Maienborn, Klaus von Heusinger, and Paul Portner (eds.), Semantics: An International Handbook of Natural Language Meaning, vol. . Berlin: De Gruyter. –. Beaver, David and Krahmer, Emiel (). Presupposition and partiality: back to the future. Journal of Logic, Language and Information, (): –. Beaver, David and Onea, Edgar (). Cleftomania. Talk presented at the workshop Questions in Pragmatics, University of Amsterdam, December. Beaver, David and Tonhauser, Judith (). Projective meaning: shadows on the wall. Talk given at XPRAG , Chicago. Beaver, David and Zeevat, Henk (). Accommodation. In Gillian Ramchand and Charles Reiss (eds.), Oxford Handbook of Linguistic Interfaces. Oxford: Oxford University Press. –. Beaver, David, Clark, Brady, Flemming, Edward, Jaeger, T. Florian, and Wolters, Maria (). When semantics meets phonetics: acoustical studies of second occurrence focus. Language, : –. Beaver, David, Roberts, Craige, Simons, Mandy, and Tonhauser, Judith (). Questions under discussion: where information structure meets projective content. Annual Review of Linguistics, : –. Beavers, John and Koontz-Garboden, Andrew (). Manner and result in the roots of verbal meaning. Linguistic Inquiry, (): –. doi: ./LING_a_ Beck, Sarah R. and Riggs, Kevin J. (). The development of counterfactual reasoning. In Henry Markovits (ed.), The Developmental Psychology of Reasoning and Decision Making. London: Psychology Press. Beck, Sarah R. and Riggs, Kevin J. (). Developing thoughts about what might have been. Child Development Perspectives, (): –. Beck, Sigrid and Gergel, Remus (). The diachronic semantics of English again. Natural Language Semantics, : –. Beckman, Mary E. (). The parsing of prosody. Language and Cognitive Processes, : –. Beckman, Mary E. and Ayers, Gayle M. (). Guidelines for ToBI labelling, version .. Ohio State University. Beckman, Mary E. and Venditti, Jennifer J. (). Intonation. In John Goldsmith, Jason Riggle, and Alan C. L. Yu (eds.), Handbook of Phonological Theory. Malden MA: Blackwell. –. Begeer, Sander, El Bouk, Saloua, Boussaid, Wafaa, Terwogt, Mark Meerum, and Koot, Hans M. (). Underdiagnosis and referral bias of autism in ethnic minorities. Journal of Autism and Developmental Disorders, : –. doi: ./s--- Behne, Tanya, Carpenter, Malinda, and Tomasello, Michael (). One-year-olds comprehend the communicative intentions behind gestures in a hiding game. Developmental Science, (): –. Behr, Rafael (). Theresa May: a one-nation Tory in a one-party state, July. https://www. theguardian.com/commentisfree//jul//theresa-may-tory-primeminister-england Behrend, Douglas A. (). The development of verb concepts: children’s use of verbs to label familiar and novel events. Child Development, (): –. doi: ./j.-..tb.x Behrens, Leila (). Typological Parameters of Genericity. Arbeitspapier (Neue Folge). Institut für Sprachwissenschaft, University of Köln. Behrens, Leila (). Genericity from a cross-linguistic perspective. Linguistics, (): –. Belke, Eva (). Visual determinants of preferred adjective order. Visual Cognition, (): –. doi: ./ Benatar, Ashley and Clifton, Charles, Jr. (). Newness, givenness and discourse updating: evidence from eye movements. Journal of Memory and Language, : –. doi: ./j.jml... Benn, Gottfried (). Prose, Essays, Poems, ed. Volkmar Sander. New York: Continuum Publishing Company. Bentin, Shlomo, McCarthy, Gregory, and Wood, Charles C. (). Event-related potentials, lexical decision and semantic priming. Electroencephalography and Clinical Neurophysiology, (): –.
Benz, Anton and Gotzner, Nicole (). Embedded implicatures revisited: issues with the TruthValue Judgment paradigm. In Judith Degen, Michael Franke, and Noah D. Goodman (eds.), Proceedings of the Formal and Experimental Pragmatics Workshop. Tübingen: FDSL. Benz, Anton and Van Rooij, Robert (). Optimal assertions and what they implicate. Topoi, (): –. Benz, Anton, Bombi Ferrer, Carla, and Gotzner, Nicole (). Scalar diversity and negative strengthening. Presentation at Sinn und Bedeutung . Berent, Gerald P., Kelly, Ronald R., and Schueler-Choukairi, Tanya (). Economy in the acquisition of English universal quantifier sentences: the interpretations of deaf and hearing students and second language learners at the college level. Applied Psycholinguistics, (): –. Berent, Iris, Pinker, Steven, Tzelgov, Joseph, Bibi, Uri, and Goldfarb, Liat (). Computation of semantic number from morphological information. Journal of Memory and Language, : –. Bergen, Leon and Goodman, Noah D. (). The strategic use of noise in pragmatic reasoning. Topics in Cognitive Science, : –. doi: ./tops. Bergen, Leon and Grodner, Daniel J. (). Speaker knowledge influences the comprehension of pragmatic inferences. Journal of Experimental Psychology: Learning, Memory, and Cognition, (): –. doi: ./a Bergen, Leon, Goodman, Noah D., and Levy, Roger (). That’s what she (could have) said: how alternative utterances affect language use. In Naomi Miyake, David Peebles, and Richard P. Cooper (eds.), Proceedings of the th Annual Conference of the Cognitive Science Society. Austin, TX: Cognitive Science Society. –. Bergen, Leon, Levy, Roger, and Goodman, Noah D. (). Pragmatic reasoning through semantic inference. Semantics and Pragmatics, (): –. doi: ./sp.. Berger, Frauke and Höhle, Barbara (). Restrictions on addition: children’s interpretation of the focus particles auch (also) and nur (only) in German. Journal of Child Language, : –. Berko, Jean (). The child’s learning of English morphology. Word, : –. Berlin, Bren and Kay, Paul (). Basic Color Terms: Their Universality and Evolution. Berkeley, CA: University of California Press. Berman, Ruth A. and Slobin, Dan I. (eds.) (). Relating Events in Narrative: A Cross-Linguistic Developmental Study. Hillsdale, NJ: Erlbaum. Bernard, Stéphane and Deleau, Michel (). Conversational perspective-taking and false belief attribution: a longitudinal study. British Journal of Developmental Psychology, (): –. doi: ./X Bernicot, Josie, Laval, Virginie, and Chaminaud, Stéphanie (). Nonliteral language forms in children: in what order are they acquired in pragmatics and metapragmatics? Journal of Pragmatics, : –. Berwick, Robert C. (). The Acquisition of Syntactic Knowledge. Cambridge, MA: MIT Press. Bever, Thomas G. (). The cognitive basis for linguistic structures. In John R. Hayes (ed.), Cognition and the Development of Language. New York: Wiley. –. Beyth-Marom, Ruth (). How probable is probable? A numerical translation of verbal probability expressions. Journal of Forecasting, : –. doi: ./for. Bill, Cory, Romoli, Jacopo, Schwarz, Florian, and Crain, Stephen (). Indirect scalar implicatures are neither scalar implicatures nor presuppositions (or both). Poster presented at CUNY . Birch, Stacy and Clifton, Charles, Jr. (). Focus, accent, and argument structure: effects on language comprehension. Language and Speech, : –. Birch, Stacy and Clifton, Charles, Jr. (). Effects of varying focus and accenting of adjuncts on the comprehension of utterances. Journal of Memory and Language, : –. doi: ./SX()- Birch, Stacy L. and Garnsey, Susan M. (). The effect of focus on memory for words in sentences. Journal of Memory and Language, : –. doi: ./jmla.. Birch, Stacy and Rayner, Keith (). Effects of syntactic prominence on eye movements during reading. Memory and Cognition, (): –.
Birch, Stacy L., Albrecht, Jason E., and Myers, Jerome L. (). Syntactic focusing structures influence discourse processing. Discourse Processes, (): –. doi: ./Sdp_ Bishop, Jason (). Information structural expectations in the perception of prosodic prominence. In Gorka Elordieta and Pilar Prieto (eds.), Prosody and Meaning. Berlin: De Gruyter. –. Bishop, Jason (). Focus projection and prenuclear accents: evidence from lexical processing. Language, Cognition and Neuroscience, : –. Blankenship, Kevin L. and Holtgraves, Thomas M. (). The role of different markers of linguistic powerlessness in persuasion. Journal of Language and Social Psychology, (): –. doi: ./ X Blasko, Dawn G. and Briihl, Deborah S. (). Reading and recall of metaphorical sentences: effects of familiarity and context. Metaphor and Symbol, (): –. doi: ./sms_ Blasko, Dawn G. and Connine, Cynthia M. (). Effects of familiarity and aptness on metaphor processing. Journal of Experimental Psychology: Learning, Memory, and Cognition, (): –. doi: ./-... Bloom, Paul (). How Children Learn the Meanings of Words. Cambridge, MA: MIT Press. Blum-Kulka, Shoshana (). Indirectness and politeness in requests: same or different? Journal of Pragmatics, (): –. Blum-Kulka, Shoshana, Danet, Brenda, and Gherson, Rimona (). The language of requesting in Israeli society. In Joseph P. Forgas (ed.), Language and Social Situations. New York: Springer. –. Bock, Kathryn and Eberhard, Kathleen M. (). Meaning, sound and syntax in English number agreement. Language and Cognitive Processes, (): –. doi: ./ Bock, Kathryn J. and Mazzella, Joanne R. (). Intonational marking of given and new information: some consequences for comprehension. Memory and Language, : –. Bock, Kathryn and Miller, Carol A. (). Broken agreement. Cognitive Psychology, (): –. doi: ./-()- Boell, Anna-Christina (). Exhaustivity effect of focus-particles in German it-clefts: empirical and corpus-based insights. In Marisa Köllner and Ramon Ziai (eds.), Proceedings of the ESSLLI Student Session. Bolzano: Bozen. –. Bogal-Allbritten, Elizabeth (). Slightly coerced: processing evidence for adjectival coercion by minimizers. Proceedings of the th Meeting of the Chicago Linguistic Society. Chicago: Chicago Linguistic Society. –. Bögels, Sara and Torreira, Francisco (). Listeners use intonational phrase boundaries to project turn ends in spoken interaction. Journal of Phonetics, : –. doi: ./j.wocn... Bolinger, Dwight (). A look at equations and cleft sentences. In Evelyn Firchow (ed.), Studies for Einar Haugen. Den Haag: De Gruyter. –. Bonini, Nicolao, Osherson, Daniel, Viale, Riccardo, and Williamson, Timothy (). On the psychology of vague predicates. Mind and Language, : –. Bonnefon, Jean-François (). Politeness and reasoning: face, connectives, and quantifiers. In Thomas M. Holtgraves (ed.), Oxford Handbook of Language and Social Psychology. New York: Oxford University Press. –. Bonnefon, Jean-François and Villejoubert, Gaëlle (). Tactful, or doubtful? Expectations of politeness explain the severity bias in the interpretation of probability phrases. Psychological Science, : –. doi: ./j.-...x Bonnefon, Jean-François, Feeney, Aidan, and Villejoubert, Gaëlle (). When some is actually all: scalar inferences in face-threatening contexts. Cognition, (), –. doi: ./ j.cognition... Bonnefon, Jean-François, de Neys, Wim, and Feeney, Aidan (). Processing scalar inferences in face-threatening contexts. In Laura Carlson, Christoph Hölscher, and Thomas F. Shipley (eds.), Proceedings of the rd Annual Conference of the Cognitive Science Society. Austin, TX: Cognitive Science Society. –. Bonnefon, Jean-François, Feeney, Aidan, and de Neys, Wim (). The risk of polite misunderstandings. Current Directions in Psychological Science, : –. doi: ./
Bonnefon, Jean-François, Dahl, Ethan, and Holtgraves, Thomas M. (). Some but not all dispreferred turn markers help to interpret scalar terms in polite contexts. Thinking and Reasoning, : –. Booij, Elbert J. and Sassoon, Galit Weidman (). Big differences: the standard for ‘big’ as used by adults and children. In Nurit Melnik (ed.), Proceedings of IATL . Cambridge, MA: MIT Working Papers in Linguistics. –. Borer, Hagit (a). Structuring Sense. Oxford: Oxford University Press. Borer, Hagit (b). Structuring Sense II: The Normal Course of Events. Oxford: Oxford University Press. Bornkessel-Schlesewsky, Ina and Schlesewsky, Matthias (). An alternative perspective on ‘semantic P’ effects in language comprehension. Brain Research Reviews, (): –. doi: ./j.brainresrev... Bosch, Peter (). ‘Vagueness’ is context-dependence: a solution to the Sorites paradox. In Thomas T. Ballmer and Manfred Pinkal (eds.), Approaching Vagueness. Amsterdam: North Holland. –. Bosch, Peter and Umbach, Carla (). Reference determination for demonstrative pronouns. ZAS Papers in Linguistics, : –. Bott, Lewis and Noveck, Ira A. (). Some utterances are underinformative: the onset and time course of scalar inferences. Journal of Memory and Language, (): –. doi: ./j. jml... Bott, Lewis, Bailey, Todd M., and Grodner, Daniel (). Distinguishing speed from accuracy in scalar implicatures. Journal of Memory and Language, (): –. doi: ./j.jml... Bott, Oliver (). The Processing of Events. Amsterdam: John Benjamins. Bott, Oliver and Radó, Janina (). How to provide exactly one interpretation for every sentence, or what eye movements reveal about quantifier scope. In Sam Featherston and Susanne Winkler (eds.), The Fruits of Empirical Linguistics, Vol. . Berlin, New York: De Gruyter. –. Bott, Oliver and Schlotterbeck, Fabian (). The processing domain of scope interaction. Journal of Semantics, (): –. Bouton, Lawrence (). Antecedent contained pro-forms. In Mary Ann Campbell, James Lindholm, Alice Davison et al. (eds.), Papers from the Sixth Regional Meeting of the Chicago Linguistic Society. Chicago: University of Chicago, Chicago Linguistic Society. –. Bowdle, Brian F. and Gentner, Dedre (). The career of metaphor. Psychological Review, : –. Bower, Gordon H. and Morrow, Daniel G. (). Mental models in narrative comprehension. Science, (): –. Bower, Gordon H. and Rinck, Mike (). Goals as generators of activation in narrative understanding. In Susan R. Goldman, Arthur C. Graesser, and Paul van den Broek (eds.), Narrative Comprehension, Causality and Coherence: Essays in Honor of Tom Trabasso. Mahwah, NJ: Erlbaum. –. Bowerman, Melissa (). Learning how to structure space for language: a cross-linguistic perspective. In Paul Bloom, Mary A. Peterson, Lynn Nadel, and Merrill F. Garrett (eds.), Language and Space. Cambridge, MA: MIT Press. –. Bowerman, Melissa and Choi, Soonja (). Shaping meanings for language: universal and languagespecific in the acquisition of spatial semantic categories. In Melissa Bowerman and Stephen C. Levinson (eds.), Language Acquisition and Conceptual Development. Cambridge: Cambridge University Press. –. doi: ./CBO. Bowerman, Melissa, de León, Lourdes, and Choi, Soonja (). Verbs, particles, and spatial semantics: learning to talk about spatial actions in typologically different languages. In Eve V. Clark (ed.), Proceedings of the th Annual Child Language Research Forum. Stanford CA: CSLI. –. Bradac, James J. and Mulac, Anthony (). A molecular view of powerful and powerless speech styles: attributional consequences of specific language features and communicator intentions. Communications Monographs, (): –.
Braine, Martin D. S. and O’Brien, David P. (eds.) (). Mental Logic. Mahwah, NJ: Erlbaum. Braine, Martin D. S. and Rumain, Barbara (). Children’s comprehension of ‘or’: evidence for a sequence of competencies. Journal of Experimental Child Psychology, : –. Braine, Martin D. S. and Rumain, Barbara (). Logical reasoning. In Paul Mussen (ed.), Handbook of Child Psychology, Vol. : Cognitive Development. New York: Wiley. –. Bransford, John D., Barclay, J. Richard, and Franks, Jeffery J. (). Sentence memory: a constructive versus interpretive approach. Cognitive Psychology, (): –. Brasoveanu, Adrian and Dotlačil, Jakub (a). Sentence-internal ‘same’ and its quantificational licensors: a new window into the processing of inverse scope. Semantics and Pragmatics, (): –. doi: ./sp.. Brasoveanu, Adrian and Dotlačil, Jakub (b). Strategies for scope-taking. Natural Language Semantics, : –. Braun, Bettina (). Phonetics and phonology of thematic contrast in German. Language and Speech, : –. Braun, Bettina and Tagliapietra, Lara (). The role of contrastive intonation contours in the retrieval of contextual alternatives. Language and Cognitive Processes, (): –. Braze, David, Shankweiler, Donald, Ni, Weijia, and Palumbo, Laura Conway (). Readers’ eye movements distinguish anomalies of form and content. Journal of Psycholinguistic Research, (): –. Bredart, Serge and Modolo, Karin (). Moses strikes again: focalization effect on a semantic illusion. Acta Psychologica, :–. doi: ./-()- Breen, Mara, Fedorenko, Evelina, Wagner, Michael, and Gibson, Edward (). Acoustic correlates of information structure. Language and Cognitive Processes, : –. Breheny, Richard (). Pro-active focus. UCL Working Papers in Linguistics, . Breheny, Richard (). Interface economy and focus. In Villy Rouchota and Andreas H. Jucker (eds.), Current Issues in Relevance Theory. Amsterdam: John Benjamins. –. Breheny, Richard (). A new look at the semantics and pragmatics of numerically quantified noun phrases. Journal of Semantics, (): –. doi: ./jos/ffm Breheny, Richard (). Experimentation-based pragmatics. In Wolfram Bublitz and Neal R. Norrick (eds.), Handbook of Pragmatics, Vol. : Foundations of Pragmatics. Berlin: De Gruyter. Breheny, Richard (). Language processing, relevance and questions. In Kate Scott, Billy Clark, and Robyn Carston (eds.), Relevance: Pragmatics and Interpretation. Cambridge: Cambridge University Press. Breheny, Richard, Katsos, Napoleon, and Williams, John N. (). Are generalised scalar implicatures generated by default? An on-line investigation into the role of context in generating pragmatic inferences. Cognition, (): –. Breheny, Richard, Ferguson, Heather J., and Katsos, Napoleon (). Investigating the timecourse of accessing conversational implicatures during incremental sentence interpretation. Language and Cognitive Processes, (): –. doi: ./.. Breheny, Richard, Ferguson, Heather J., and Katsos, Napoleon (). Taking the epistemic step: toward a model of on-line access to conversational implicatures. Cognition, (): –. doi: ./j.cognition... Breheny, Richard, Klinedinst, Nathan, Romoli, Jacopo, and Sudo, Yasutada (). The symmetry problem: current theories and prospects. Natural Language Semantics. doi: ./s---z Bremner, J. Gavin (). Egocentric versus allocentric spatial coding in nine-month-old infants: fFactors influencing the choice of code. Developmental Psychology, (): –. doi: ./ -... Bremner, J. Gavin and Bryant, Peter E. (). Place versus response as the basis of spatial errors made by young infants. Journal of Experimental Child Psychology, (): –. doi: ./()- Brennan, Jonathan and Pylkkänen, Liina (). Processing events: behavioral and neuromagnetic correlates of aspectual coercion. Brain and Language, (): –.
Brennan, Susan E. (). Centering attention in discourse. Language and Cognitive Processes, : –. Brennan, Susan E. and Clark, Herbert H. (). Conceptual pacts and lexical choice in conversation. Journal of Experimental Psychology: Learning, Memory and Cognition, (): –. Brennan, Susan E., Friedman, Marilyn W., and Pollard, Carl J. (). A centering approach to pronouns. Proceedings of the th Annual Meeting of the Association for Computational Linguistics. Stroudsberg, PA: Association for Computational Linguistics. Brewer, Marilynn B. (). Research design and issues of validity. In Harry T. Reis and Charles M. Judd (eds.), Handbook of Research Methods in Social and Personality Psychology. Cambridge: Cambridge University Press. –. Brisson, Christine (). Plurals, all, and the nonuniformity of collective predication. Linguistics and Philosophy, : –. Brock, Jon (). Language abilities in Williams syndrome: a critical review. Development and Psychopathology, : –. Bromme, Rainer, Brummernhenrich, Benjamin, Becker, Bettina-Maria, and Jucks, Regina (). The effects of politeness-related instruction on medical tutoring. Communication Education, : –. doi: ./.. Brooks, Neon, Audet, Jennifer, and Barner, David (). Pragmatic inference, not semantic competence, guides -year-olds’ interpretation of unknown number words. Developmental Psychology, (): –. Brooks, Patricia J. and Braine, Martin D. S. (). What do children know about the universal quantifiers all and each? Cognition, (): –. Brooks, Patricia J. and Sekerina, Irina (/). Shortcuts to quantifier interpretation in children and adults. Language Acquisition, : –. Brooks, Patricia J. and Sekerina, Irina A. (). Shallow processing of universal quantification: a comparison of monolingual and bilingual adults. In Ron Sun (ed.), Proceedings of the th Annual Meeting of the Cognitive Science Society. Mahwah, NJ: Erlbaum. –. Brooks, Patricia J., Jia, Xiangdong, Braine, Martin D. S., and da Graca Dias, Marcia (). A crosslinguistic study of children’s comprehension of universal quantifiers: a comparison of Mandarin Chinese, Portuguese and English. First Language, (): –. Brooks, Patricia J., Braine, Martin D. S., Jia, Gisela, and da Graca Dias, Marcia (). Early representations for all, each, and their counterparts in Mandarin Chinese and Portuguese. In Melissa Bowerman and Stephen C. Levinson (eds.), Language Acquisition and Conceptual Development. Cambridge: Cambridge University Press. –. Brosch, Tobias, Pourtois, Gilles, and Sander, David (). The perception and categorisation of emotional stimuli: a review. Cognition and Emotion, : –. Brouwer, Harm, Fitz, Hartmut, and Hoeks, John (). Getting real about semantic illusions: rethinking the functional role of the P in language comprehension. Brain Research, : –. Brown, R. (). A First Language: The Early Stages. London: Allen & Unwin Ltd. Brown, Penelope and Levinson, Stephen C. (). Politeness: Some Universals in Language Use. Cambridge: Cambridge University Press. Brown, Penelope and Levinson, Stephen C. (). Linguistic and nonlinguistic coding of spatial arrays: explorations in Mayan cognition. Working paper no. , Cognitive Anthropology Research Group, Max Planck Institute for Psycholinguistics, Nijmegen. Brown, Penelope and Levinson, Stephen C. (). Frames of spatial reference and their acquisition in Tenejapan Tzeltal. In Larry P. Nucci, Geoffrey B. Saxe, and Elliot Turiel (eds.), Culture, Thought, and Development. Mahwah, NJ: Lawrence Erlbaum Associates. –. Brown, Roger (). A First Language: The Early Stages. Cambridge, MA: Harvard University Press. Brown, Roger and Fish, Deborah (). The psychological causality implicit in language. Cognition, (): –. Brown, Roger and Gilman, Albert (). Politeness theory and Shakespeare’s four major tragedies. Language in Society, (), –.
Brown-Schmidt, Sarah (). Partner-specific interpretation of maintained referential precedents during interactive dialog. Journal of Memory and Language, (): –. doi: ./j. jml... Brown-Schmidt, Sarah and Hanna, Joy E. (). Talking in another person’s shoes: incremental perspective-taking in language processing. Dialogue and Discourse, : –. Brown-Schmidt, Sarah and Konopka, Agnieszka E. (). Experimental approaches to referential domains and the on-line processing of referring expressions in unscripted conversation. Information, : –. Brown-Schmidt, Sarah and Tanenhaus, Michael K. (). Watching the eyes when talking about size: an investigation of message formulation and utterance planning. Journal of Memory and Language, : –. Brown-Schmidt, Sarah and Tanenhaus, Michael K. (). Real-time investigation of referential domains in unscripted conversation: a targeted language game approach. Cognitive Science, : –. Brown-Schmidt, Sarah, Byron, Donna K., and Tanenhaus, Michael K. (). Beyond salience: interpretation of personal and demonstrative pronouns. Journal of Memory and Language, : –. Brown-Schmidt, Sarah, Gunlogson, Christine, and Tanenhaus, Michael K. (). Addressees distinguish shared from private information when interpreting questions during interactive conversation. Cognition, (): –. doi: ./j.cognition... Brummernhenrich, Benjamin and Jucks, Regina (). Managing face threats and instructions in online tutoring. Journal of Experimental Psychology, : –. doi: ./a Brun, Wibecke and Teigen, Karl H. (). Verbal probabilities: ambiguous, context dependent, or both? Organizational Behavior and Human Decision Processes, : –. doi: ./()- Bucci, Wilma (). The interpretation of universal affirmative propositions. Cognition, (): –. Buccola, Brian and Spector, Benjamin (). Modified numerals and maximality. Linguistics and Philosophy, : –. doi: ./s--- Budescu, David V. and Wallsten, Thomas S. (). On two complementary approaches to the study of verbal probabilities. In Wibecke Brun, Gideon Keren, Geir Kirkeben, and Henry Montgomery (eds.), Perspectives on Thinking, Judging, and Decision Making. Oslo: Universitetsforlaget. –. Budescu, David V., Karelitz, Tzur M., and Wallsten, Thomas S. (). Predicting the directionality of probability words from their membership functions. Journal of Behavioral Decision Making, : –. doi: ./bdm. Budescu, David V., Por, Han-Hui, and Broomell, Stephen B. (). Effective communication of uncertainty in the IPCC Reports. Climatic Change, : –. doi: ./s--- Budescu, David V., Por, Han-Hui, Broomell, Stephen B., and Smithson, Michael (). The interpretation of IPCC probabilistic statements around the world. Nature Climate Change, : –. doi: ./nclimate Bunger, Ann, Trueswell, John C., and Papafragou, Anna (). The relation between event apprehension and utterance formulation in children: evidence from linguistic omissions. Cognition, (): –. doi: ./j.cognition... Burdin, Rachel Steindel, Phillips-Bourass, Sara, Turnbull, Rory, Yasavul, Murat, Clopper, Cynthia G. and Tonhauser, Judith (). Variation in the prosody of focus in head- and head/edgeprominence languages. Lingua, : –. Burgess, Neil (). Spatial cognition and the brain. Annals of the New York Academy of Sciences, (): –. doi: ./annals.. Büring, Daniel (). On D-trees, beans and B-accents. Linguistics and Philosophy, : –. Büring, Daniel (). Intonation, semantics and information structure. In Gillian Ramchand and Charles Reiss (eds.), Oxford Handbook of Linguistic Interfaces. Oxford: Oxford University Press. –. Büring, Daniel (). The least at least can do. In Charles B. Chang and Hannah J. Haynie (eds.), Proceedings of WCCFL . Somerville, MA: Cascadilla Press. –. Büring, Daniel (). Conditional exhaustivity presuppositions in clefts (and definites). Ms. UCLA.
Büring, Daniel and Manuel Križ (). It’s that, and that’s it! Exhaustivity and homogeneity presuppositions in clefts (and definites). Semantics and Pragmatics, (): –. Burkhardt, Petra (). Inferential bridging relations reveal distinct neural mechanisms: Evidence from event-related brain potentials. Brain and Language, (): –. Burnard, Lou (). Reference Guide for the British National Corpus (World Edition). Oxford: Oxford University Computing Services. Burnett, Heather (). Signalling games, sociolinguistic variation and the construction of style. Linguistics and Philosophy. Burrell, Nancy A. and Koper, Randal J. (). The efficacy of powerful/powerless language on attitudes and source credibility. In Mike Allen and Raymond W. Preiss (eds.), Persuasion: Advances Through Meta-Analysis. Cresskill, NJ: Hampton Press. –. Buttelmann, David, Carpenter, Malinda, and Tomasello, Michael (). Eighteen-month-old infants show false belief understanding in an active helping paradigm. Cognition, : –. Butterfill, Stephen A. and Apperly, Ian A. (). How to construct a minimal Theory of Mind. Mind and Language, (): –. Bylinina, Lisa (). The grammar of standards: judge-dependence, purpose-relativity, and comparison classes in degree constructions. PhD thesis. Utrecht University. Byrne, Ruth M. J. (). Deductive reasoning. In Robert J. Sternberg (ed.), Encyclopedia of Intelligence. New York: Macmillan. Byrne, Ruth M. J. (). Cognitive processes in counterfactual thinking about what might have been. In Douglas L. Medin (ed.), The Psychology of Learning and Motivation. San Diego, CA: Academic Press. –. Byrne, Ruth M. J. (). Mental models and counterfactual thoughts about what might have been. Trends in Cognitive Sciences, (): –. Byrne, Ruth M. J. (). The Rational Imagination: How People Create Alternatives to Reality. Cambridge, MA: MIT Press. Byrne, Ruth M. J. (). Counterfactual thought. Annual Review of Psychology, : –. Byrne, Ruth M. J. and Tasso, Alessandra (). Deductive reasoning with factual, possible, and counterfactual conditionals. Memory and Cognition, (): –. Byrne, Ruth M. J., Segura, Susana, Culhane, Ronan, Tasso, Alessandra, and Berrocal, Pablo (). The temporality effect in counterfactual thinking about what might have been. Memory and Cognition, : –. Cacciari, Cristina and Glucksberg, Sam (). Understanding figurative language. In Morton Ann Gernsbacher (ed.), Handbook of Psycholinguistics. New York: Academic Press. –. Cacciari, Cristina and Padovani, Roberto (). The development of figurative language. In Michael Spivey, Mark Joanisse, and Ken McRae (eds.), Cambridge Handbook of Psycholinguistics. Cambridge: Cambridge University Press. –. Calhoun, Sasha (). Phonetic dimensions of intonational categories: the case of L+H* and H*. In Bernard Bel and Isabelle Marlien (eds.), Proceedings of Speech Prosody . Nara: Speech Prosody. Calhoun, Sasha (). The centrality of metrical structure in signaling information structure: a probabilistic perspective. Language, : –. Call, Josep and Tomasello, Michael (). What do chimpanzees know about seeing revisited: an explanation of the third kind. In Naomi Eilan, Christop Hoerl, Teresa McCormack, and Johannes Roessler (eds.), Issues in Joint Attention. Oxford: Oxford University Press. –. Call, Josep and Tomasello, Michael (). Does the chimpanzee have a theory of mind? years later. Trends in Cognitive Science, : –. Caramazza, Alfonso and Grober, Ellen (). Polysemy and the structure of the subjective lexicon. In Clea Rameh (ed.), Semantics: Theory and Application. Washington, DC: Georgetown University Press. –. Caramazza, Alfonso, Grober, Ellen, Garvey, Catherine, and Yates, Jack (). Comprehension of anaphoric pronouns. Journal of Verbal Learning and Verbal Behavior, (): –. doi: ./ S–()–
Carbary, Kathleen M. and Tanenhaus, Michael K. (). Syntactic priming in an unscripted dialogue task. Poster presented at the th CUNY Conference on Human Sentence Processing, UCSD Center for Research in Language, La Jolla, CA. Carey, Susan (). Conceptual Change in Childhood. Cambridge, MA: MIT Press. Carey, Susan (). The Origin of Concepts. Oxford: Oxford University Press. Carli, Linda L. (). Gender, language, and influence. Journal of Personality and Social Psychology, (): –. doi: ./-... Carlson, Gregory N. (). Reference to kinds in English. PhD thesis. University of Massachusetts, Amherst. Carlson, Gregory (). Same and different: some consequences for syntax and semantics. Linguistics and Philosophy, (): –. doi: ./BF Carlson, Gregory N. (). The truth conditions of generic sentences: two contrasting views. In Greg Carlson and Francis J. Pelletier (eds.), The Generic Book. Chicago: Chicago University Press. –. Carlson, Gregory N. (). Genericity. In Klaus von Heusinger, Claudia Maienborn, and Paul Portner (eds.), Semantics: An International Handbook of Natural Language Meaning. Berlin: De Gruyter. –. Carlson, Gregory N. and Tanenhaus, Michael, K. (). Thematic roles and language comprehension. In Wendy Wilkens (ed.), Syntax and Semantics, Vol. : Thematic Relations. New York: Academic Press. –. Carlson, Laura A. and Covey, Eric S. (). How far is near? Inferring distance from spatial descriptions. Language and Cognitive Processes, : –. doi: ./ Carlson, Laura A. and Kenny, Ryan (). Interpreting spatial terms involves simulating interactions. Psychonomic Bulletin and Review, : –. doi: ./BF Carminati, Maria Nella (). The processing of Italian subject pronouns. PhD thesis. University of Massachusetts at Amherst. Carnap, Rudolf, (). Logical Foundations of Probability. Chicago: University of Chicago Press. Carpenter, Patricia A. and Just, Marcel Adam (). Sentence comprehension: a psycholinguistic processing model of verification. Psychological Review, (): –. Carpenter, Patricia A., Just, Marcel Adam, Keller, Timothy A., Eddy, William F., and Thulborn, Keith R. (). Time course of fMRI-activation in language and spatial networks during sentence comprehension. NeuroImage, (): –. doi: ./nimg.. Carruthers, Peter (). Mindreading in infancy. Mind and Language, (): –. Carston, Robyn (). Quantity maxims and generalised implicature. Lingua, : –. Carston, Robyn (). Thoughts and Utterances: The Pragmatics of Explicit Communication. Oxford: Blackwell. Carston, Robyn (). Metaphor: ad hoc concepts, literal meaning and mental images. Proceedings of the Aristotelian Society, (): –. Carston, Robyn and Wearing, Catherine (). Hyperbolic language and its relation to metaphor and irony. Journal of Pragmatics, : –. Casasola, Marianella (). Can language do the driving? The effect of linguistic input on infants’ categorization of support spatial relations. Developmental Psychology, (): –. doi: ./ –... Casasola, Marianella (). The development of infants’ spatial categories. Current Directions in Psychological Science, (): –. doi: ./j.–...x Casasola, Marianella and Bhagwat, Jui (). Do novel words facilitate -month-olds’ spatial categorization? Child Development, (): –. doi: ./j.–...x Casasola, Marianella and Cohen, Leslie B. (). Infant categorization of containment, support and tight-fit spatial relationships. Developmental Science, (): –. doi: ./–. Casasola, Marianella, Cohen, Leslie B., and Chiarello, Elizabeth (). Six-month-old infants’ categorization of containment spatial relations. Child Development, (): –. doi: ./ –.
Caspers, Johanneke (). Local speech melody as a limiting factor in the turn-taking system in Dutch. Journal of Phonetics, : –. Castro Ferreira, Thiago, Krahmer, Emiel, and Wubben, Sander (). Individual variation in the choice of referential form. Proceedings of the Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. San Diego, CA: Association for Computational Linguistics. –. Cazden, Courtney B. (). The acquisition of noun and verb inflections. Child Development, : –. Cecchetto, Carlo (). Explaining the locality conditions of QR: consequences for the theory of phases. Natural Language Semantics, : –. Chafe, Wallace L. (). Givenness, contrastiveness, definiteness, subjects, topics, and point of view. In Charles N. Li (ed.), Subject and Topic. New York: Academic Press. –. Chafe, Wallace L. (). Discourse, Consciousness, and Time. Chicago: Chicago University Press. Chambers, Craig G. and San Juan, Valerie (). Accommodation and the interpretation of presupposition during referential processing. Poster presented at the th CUNY Sentence Processing Conference. Chambers, Craig G. and San Juan, Valerie (). Perception and presupposition in real-time language comprehension: insights from anticipatory processing. Cognition, (): –. doi: ./j.cognition... Chambers, Craig G. and Smyth, Ron (). Structural parallelism and discourse coherence: a test of centering theory. Journal of Memory and Language, : –. Chambers, Craig G., Tanenhaus, Michael K., and Magnuson, James S. (). Actions and affordances in syntactic ambiguity resolution. Journal of Experimental Psychology: Learning, Memory, and Cognition, (): –. Chang, Frederick R. (). Active memory processes in visual sentence comprehension: clause effects and pronominal reference. Memory and Cognition, : –. Chao, Sun and Breheny, Richard (under review). What would a compostional listener do? Another look at the timecourse of scalar implicatures. Chapman, Loren J. and Chapman, Jean P. (). Atmosphere effect re-examined. Journal of Experimental Psychology, (): –. doi: ./h Charlow, Simon (). ‘Strong’ predicative presuppositional objects. Proceedings of New Directions in the Theory of Presupposition. Bordeaux: ESSLLI. Chatzikonstantinou, Anastasios (). Semantic and prosodic processing of the negative polarity items in Greek. PhD thesis. University of Chicago. Chemla, Emmanuel (). An epistemic step for anti-presuppositions. Journal of Semantics, : –. Chemla, Emmanuel (a). Similarity: towards a unified account of scalar implicatures, free choice permission and presupposition projection. Ms. ENS Paris. Chemla, Emmanuel (b). Presuppositions of quantified sentences: experimental data. Natural Language Semantics, (): –. doi: ./s--- Chemla, Emmanuel and Bott, Lewis (). Processing presuppositions: dynamic semantics vs pragmatic enrichment. Language and Cognitive Processes, (): –. Chemla, Emmanuel and Bott, Lewis (). Processing inferences at the semantics/pragmatics frontier: disjunctions and free choice. Cognition, (): –. doi: ./j.cognition... Chemla, Emmanuel and Schlenker, Philippe (). Incremental vs. symmetric accounts of presupposition projection: an experimental approach. Natural Language Semantics, (): –. doi: ./s--- Chemla, Emmanuel and Singh, Raj (). Remarks on the experimental turn in the study of scalar implicature, Part II. Language and Linguistics Compass, (): –. doi: ./lnc. Chen, Lijing and Yang, Yufang (). Emphasizing the only character: emphasis, attention and contrast. Cognition, : –. Chen, Lijing, Li, Xingshan, and Yang, Yufang (). Focus, newness and their combination: processing of information structure in discourse. PLoS ONE, ():e.
Chevallier, Coralie, Noveck, Ira A., Nazir, Tatjana, Bott, Lewis, Lanzetti, Valentina, and Sperber, Dan (). Making disjunctions exclusive. Quarterly Journal of Experimental Psychology, (): –. doi: ./ Chevallier, Coralie, Wilson, Deirdre, Happé, Francesca, and Noveck, Ira A. (). Scalar inferences in autism spectrum disorders. Journal of Autism and Developmental Disorders, (): –. Chiappe, Dan L. and Kennedy, John M. (). Aptness predicts preference for metaphors or similes, as well as recall bias. Psychonomic Bulletin and Review, (): –. Chiappe, Dan, Kennedy, John M., and Smykowski, Tim (). Reversibility, aptness and the conventionality of metaphors and similes. Metaphor and Symbol, : –. doi: ./ SMS_ Chiarello, Christine (). Interpretation of word meanings in the cerebral hemispheres: one is not enough. In Paula J. Schwanenflugel (ed.), The Psychology of Word Meanings. Hillsdale, NJ: Erlbaum. –. Chierchia, Gennaro (). Dynamics of Meaning: Anaphora, Presupposition and the Theory of Grammar. Chicago and London: University of Chicago Press. Chierchia, Gennaro (). Reference to kinds across languages. Natural Language Semantics, (): –. Chierchia, Gennaro (). Scalar implicatures, polarity phenomena and the syntax/pragmatics interface. In Adriana Belletti (ed.), Structures and Beyond. Oxford: Oxford University Press. –. Chierchia, Gennaro (). Broaden your views: implicatures of domain widening and the ‘logicality’ of language. Linguistic Inquiry, (): –. Chierchia, Gennaro (). On the explanatory power of dynamic semantics. Talk given at Stanford University. Chierchia, Gennaro (). Logic in Grammar: Polarity, Free Choice, and Intervention. Oxford: Oxford University Press. Chierchia, Gennaro and McConnell-Ginet, Sally (). Meaning and Grammar: An Introduction to Semantics. Cambridge, MA: MIT Press. Chierchia, Gennaro, Crain, Stephen, Guasti, Maria Teresa, Gualmini, Andrea, and Meroni, Luisa (). The acquisition of disjunction: evidence for a grammatical view of scalar implicatures. In Anna H.-J. Do, Laura Domínguez, and Aimee Johansen (eds.), Proceedings of the th Annual Boston University Conference on Language Development. Somerville, MA: Cascadilla Press. –. Chierchia, Gennaro, Crain, Stephen, Guasti, Maria Teresa, and Thornton, Rosalind (). ‘Some’ and ‘or’: a study on the emergence of logical form. In S. Catherine Howell, Sarah A. Fish, and Thea Keith-Lucas (eds.), Proceedings of the th Boston University Conference on Language Development. Somerville, MA: Cascadilla Press. –. Chierchia, Gennaro, Fox, Danny, and Spector, Benjamin (). Hurford’s Constraint and the Theory of Scalar Implicatures. In Paul Egré and Giorgio Magri (eds.), Presuppositions and Implicatures: Proceedings of the MIT-Paris Workshop. Cambridge, MA: MIT Working Papers in Linguistics. –. Chierchia, Gennaro, Fox, Danny, and Spector, Benjamin (). The grammatical view of scalar implicatures and the relationship between semantics and pragmatics. In Claudia Maienborn, Klaus von Heusinger, and Paul Portner (eds.), Semantics: An International Handbook of Natural Language Meaning, Vol. : Berlin: De Gruyter. –. Choi, Soonja (). Influence of language-specific input on spatial cognition: categories of containment. First Language, (): –. doi: ./ Choi, Soonja and Bowerman, Melissa (). Learning to express motion events in English and Korean: the influence of language-specific lexicalization patterns. Cognition, : –. doi: ./–()–Z Chomsky, Noam (). Aspects of the Theory of Syntax. Cambridge, MA: MIT Press. Chomsky, Noam (). New Horizons in the Study of Language and Mind. Cambridge: Cambridge University Press.
Chow, Wing-Yee and Phillips, Colin (). No semantic illusions in the ‘Semantic P’ phenomenon: ERP evidence from Mandarin Chinese. Brain Research, : –. Cimpian, Andrei, Brandone, Amanda C., and Gelman, Susan A. (). Generic statements require little evidence for acceptance but have powerful implications. Cognitive Science, (): –. Cimpian, Andrei, Gelman, Susan A., and Brandone, Amanda C. (). Theory-based considerations influence the interpretation of generic sentences. Language and Cognitive Processes, (): –. Clark, Eve V. (a). Non-linguistic strategies and the acquisition of word meanings. Cognition, (): –. doi: ./–()– Clark, Eve V. (b). What’s in a word? On the child’s acquisition of semantics in his first language. In Timothy E. Moore (ed.), Cognitive Development and the Acquisition of Language. New York: Academic Press. –. Clark, Eve V. (). First language acquisition. In John Morton and John C. Marshall (eds.), Psycholinguistics I: Development and Pathology. London: Paul Elek/Ithaca, NY: Cornell University Press. –. Clark, Eve V. (). Here’s the top: nonlinguistic strategies in the acquisition of orientational terms. Child Development, : –. doi: ./ Clark, Eve V. (). The principle of contrast: a constraint on language acquisition. In Brian MacWhinney (ed.), Mechanisms of Language Acquisition. Hillsdale, NJ: Lawrence Erlbaum Associates. –. Clark, Eve V. (). On the pragmatics of contrast. Journal of Child Language, : –. Clark, Eve V. and Amaral, Patrícia M. (). Children build on pragmatic information in language acquisition. Language and Linguistics Compass, : –. doi: ./j.-X...x Clark, Eve V. and Clark, Herbert H. (). When nouns surface as verbs. Language, (): –. Clark, Herbert H. (). Linguistic processes in deductive reasoning. Psychological Review, : –. Clark, Herbert H. (). Space, time semantics, and the child. In Timothy E. Moore (ed.), Cognitive Development and the Acquisition of Language. New York, NY: Academic Press. –. Clark, Herbert H. (). Bridging. In Roger C. Schank and Bonnie L. Nash-Webber (eds.), Theoretical Issues in Natural Language Processing, New York: Association for Computing Machinery. Clark, Herbert H. and Chase, William G. (). On the process of comparing sentences against pictures. Cognitive Psychology, (): –. doi: ./-()- Clark, Herbert H. and Clark, Eve V. (). The Psychology of Language: An Introduction to Psycholinguistics. New York: Harcourt, Brace. Clark, Herbert H. and Gerrig, Richard J. (). Understanding old words with new meanings. Journal of Verbal Learning and Verbal Behavior, : –. Clark, Herbert H. and Gerrig, Richard J. (). On the pretense theory of irony. Journal of Experimental Psychology: General, (): –. Clark, Herbert H. and Krych, Meredyth A. (). Speaking while monitoring addressees for understanding. Journal of Memory and Language, (): –. doi: ./j.jml... Clark, Herbert H. and Lucy, Peter (). Understanding what is said from what is meant: A study on conversationally conveyed requests. Journal of Verbal Learning and Verbal Behavior, : –. Clark, Herbert H. and Marshall, Catherine (). Definite reference and mutual knowledge. In Aravind K. Joshi, Bonnie L. Webber, and Ivan A. Sag (eds.), Elements of Discourse Understanding. Cambridge: Cambridge University Press. –. Clark, Herbert H. and Murphy, Gregory L. (). Audience design in meaning and reference. Advances in Psychology, (C): –. doi: ./S-()- Clark, Herbert H. and Schaefer, Edward F. (). Collaborating on contributions to conversation. Language and Cognitive Processes, : –. Clark, Herbert H. and Schunk, Dale H. (). Polite responses to polite requests. Cognition, (): –. Clark, Herbert H. and Sengul, C. J. (). In search of referents for nouns and pronouns. Memory and Cognition, : –.
Clark, Herbert H. and Wilkes-Gibbs, Deanna (). Referring as a collaborative process. Cognition, : –. doi: ./-()- Clark, Herbert H., Schreuder, Robert, and Buttrick, Samuel (). Common ground and the understanding of demonstrative reference. Journal of Verbal Learning and Verbal Behavior, : –. doi: ./S-()- Clark, Dominic A. (). Verbal uncertainty expressions: a critical review of two decades of research. Current Psychology: Research and Reviews, : –. doi: ./BF Clark, Andy (). Predictive brains, situated agents, and the future of cognitive science. Behavioral and Brain Sciences, (): –. doi: ./SX Clarke, Valerie A., Ruffin, Coral L., Hill, David J., and Beamen, Arthur L. (). Ratings of orally presented verbal expressions of probability by a heterogeneous sample. Journal of Applied Social Psychology, : –. doi: ./j.-..tb.x Clarke, Alasdair D. F., Elsner, Micha, and Rohde, Hannah (). Where’s Wally: the influence of visual salience on referring expression generation. Frontiers in Psychology, :. doi: ./ fpsyg.. Clements, Wendy A. and Perner, Josef (). Implicit understanding of belief. Cognitive Development, (): –. Clifton, Charles, Jr. (). Situational context affects definiteness preferences: accommodation of presuppositions. Journal of Experimental Psychology: Learning, Memory, and Cognition, (): –. Clifton, Charles, Jr. and Frazier, Lyn (). Discourse integration guided by the ‘question under discussion’. Cognitive Psychology, : –. Clifton, Charles, Jr. and Frazier, Lyn (). Accommodation to an unlikely episodic state. Journal of Memory and Language, : –. doi: ./j.jml... Clifton, Charles, Jr., Fanselow, Gisbert, and Frazier, Lyn (). Amnestying superiority violations: processing multiple questions. Linguistic Inquiry, : –. Clopper, Cynthia G. and Tonhauser, Judith (). On the prosodic coding of focus in Paraguayan Guaraní. In Mary Byram Washburn, Katherine McKinney-Bock, Erika Varis, Ann Sawyer, and Barbara Tomaszewicz (eds.), Proceedings of the th West Coast Conference on Formal Linguistics. Somerville, MA: Cascadilla Proceedings Project. –. Clopper, Cynthia G. and Tonhauser, Judith (). The prosody of focus in Paraguayan Guaraní. International Journal of American Linguistics, : –. Cobreros, Pablo, Egré, Paul, Ripley, David, and Van Rooij, Robert (). Tolerant, classical, strict. Journal of Philosophical Logic, : –. Cobreros, Pablo, Egré, Paul, Ripley, David, and Van Rooij, Robert (). Pragmatic interpretations of vague expressions: strongest meaning and nonmonotonic consequence. Journal of Philosophical Logic, : –. Cohen, Ariel (). Generics, frequency adverbs, and probability. Linguistics and Philosophy, : –. Cohen, Ariel (). Relative readings of many, often, and generics. Natural Language Semantics, (): –. doi: ./A: Cohen, Ariel (). Generics and mental representation. Linguistics and Philosophy, : –. Cohen, Ariel and Krifka, Manfred (). Superlative quantifiers and meta-speech acts. Linguistics and Philosophy, (): –. Cohn, Lawrence D., Vázquez, Miguel E. Corts, and Alvarez, Alfredo (). Quantifying risk: verbal probability expressions in Spanish and English. American Journal of Health Behavior, (): –. doi: ./AJHB... Coll-Florit, Marta and Gennari, Silvia P. (). Time in language: event duration in language comprehension. Cognitive Psychology, (): –. Collins, John (). Generics sans Gen. Ms. Distributed at The Generic Workshop: Genericity in Linguistics and Philosophy, Harvard University. Colston, Herbert L. (). On necessary conditions for verbal irony comprehension. Pragmatics and Cognition, (): –.
Colston, Herbert L. (). Contrast and assimilation in verbal irony. Journal of Pragmatics, (): –. doi: ./S-()-X Colston, Herbert L. and O’Brien, Jennifer (). Contrast and pragmatics in figurative language: anything understatement can do, irony can do better. Journal of Pragmatics, (): –. Cometa, Michael S. and Eson, Morris E. (). Logical operations and metaphor interpretation: a piagetian model. Child Development, : –. doi: ./ Condry, Kirsten F. and Spelke, Elizabeth S. (). The development of language and abstract concepts: the case of natural number. Journal of Experimental Psychology: General, (): –. doi: ./-... Conroy, Anastasia, Lidz, Jeffrey, and Musolino, Julien (). The fleeting Isomorphism Effect. Language Acquisition, (): –. Conroy, Anastasia, Takahashi, Eri, Lidz, Jeffrey, and Phillips, Colin (). Equal treatment for all antecedents: how children succeed with Principle B. Linguistic Inquiry, (): –. Cooper, Roger M. (). The control of eye fixation by the meaning of spoken language: a new methodology for the real-time investigation of speech perception, memory, and language processing. Cognitive Psychology, (): –. doi: ./-()-X Cooper, William E., Eady, Stephen, and Mueller, Pamela R. (). Acoustical aspects of contrastive stress in question-answer contexts. Journal of the Acoustical Society of America, : –. Copestake, Ann and Briscoe, Ted (). Semi-productive polysemy and sense extension. Journal of Semantics, (): –. doi: ./jos/.. Coppock, Elizabeth and Brochhagen, Thomas (a). Raising and resolving issues with scalar modifiers. Semantics and Pragmatics, (): –. Coppock, Elizabeth and Brochhagen, Thomas (b). Diagnosing truth, interactive sincerity, and depictive sincerity. In Todd Snider (ed.), Proceedings of Semantics and Linguistic Theory . Santa Cruz, CA: Salt. –. Corkum, Valerie and Moore, Chris (). The origins of joint visual attention in infants. Developmental Psychology, (): –. Cornish, Elizabeth R. and Wason, Peter C. (). The recall of affirmative and negative sentences in an incidental learning task. Quarterly Journal of Experimental Psychology, (): –. Coulson, Seana and Lovett, Christopher (). Comprehension of non-conventional indirect requests: an event-related brain potential study. Italian Journal of Linguistics, (): –. Couper-Kuhlen, Elizabeth and Selting, Margaret (). Towards an interactional perspective on prosody and a prosodic perspective on interaction. In Elizabeth Couper-Kuhlen and Margaret Selting (eds.), Prosody in Conversation. Cambridge: Cambridge University Press. Coventry, Kenny R. and Garrod, Simon C. (). Saying, Seeing and Acting: The Psychological Semantics of Spatial Prepositions. Hove: Psychology Press. Coventry, Kenny R., Carmichael, Richard, and Garrod, Simon C. (). Spatial prepositions, objectspecific function, and task requirements. Journal of Semantics, (): –. doi: ./jos/ .. Coventry, Kenny R., Tenbrink, Thora, and Bateman, John (eds.) (). Spatial Language and Dialogue. Oxford: Oxford University Press. Cowart, Wayne (). Experimental syntax: applying objective methods to sentence judgements. Thousand Oaks, CA: Sage Publications. Cowles, H. Wind and Ferreira, Victor S. (). The influence of topic status on written and spoken sentence production. Discourse Processes, : –. Cowles, H. Wind, Kluender, Robert, Kutas, Marta, and Polinsky, Maria (). Violations of information structure: an electrophysiological study of answers to wh-questions. Brain and Language, : –. doi: ./j.bandl... Cowles, H. Wind, Walenski, Matthew, and Kluender, Robert (). Linguistic and cognitive prominence in anaphor resolution: topic, contrastive focus and pronouns. Topoi, : –. doi: ./s--- Cowper, Elizabeth (). Grammatical aspect in English. Linguistic Review, : –.
Craig, Robert T., Tracy, Karen, and Spisak, Frances (). The discourse of requests. Human Communication Research, (): –. Crain, Stephen and McKee, Cecile (). The acquisition of structural restrictions on anaphora. In Stephen Berman, Jae Woong Choe, and Joyce McDonough (eds.), Proceedings of the th Annual Meeting of the North Eastern Linguistics Society. Amherst, MA: University of Massachusetts, Graduate Linguistic Student Association. –. Crain, Stephen and Steedman, Mark (). On not being led up the garden path: the use of context by the psychological syntax processor. In Lauri Karttunen, David Dowty, and Arnold Zwicky (eds.), Natural Language Parsing: Psychological, Computational and Theoretical Perspectives. Cambridge: Cambridge University Press. –. Crain, Stephen and Thornton, Rosalind (). Investigations in Universal Grammar: A Guide to Experiments on the Acquisition of Syntax and Semantics. Cambridge, MA: MIT Press. Crain, Stephen, Ni, Weijia, and Conway, Laura (). Learning, parsing, and modularity. In Charles Clifton Jr., Lyn Frazier, and Keith Rayner (eds.), Perspectives on Sentence Processing. Hillsdale, NJ: Lawrence Erlbaum. –. Crain, Stephen, Thornton, Rosalind, Boster, Carole, Conway, Laura, Lillo-Martin, Diane, and Woodams, Elaine (). Quantification without qualification. Language Acquisition, : –. Crawley, Rosalind A., Stevenson, Rosemary J., and Kleinman, David (). The use of heuristic strategies in the interpretation of pronouns. Journal of Psycholinguistic Research, : –. doi: ./BF Crinean, Marcelle and Garnham, Alan (). Implicit causality, implicit consequentiality and semantic roles. Language and Cognitive Processes, : –. doi: ./ Croft, William (). The role of domains in the interpretation of metaphors and metonymies. Cognitive Linguistics, (): –. Crozier, Raymond and Dimmock, Patricia S. (). Name-calling and nicknames in a sample of primary school children. British Journal of Educational Psychology, (): –. Cruse, D. Alan (). Lexical Semantics. Cambridge: Cambridge University Press. Csibra, Gergely (). Recognizing communicative intentions in infancy. Mind and Language, (): –. Csibra, Gergely and Gergely, György (). Natural pedagogy as evolutionary adaptation. Philosophical Transactions of the Royal Society B, : –. Culpeper, Jonathan (). Impoliteness. Cambridge: Cambridge University Press. Cummins, Chris (). Constraints on Numerical Expressions. Oxford: Oxford University Press. Cummins, Chris and Katsos, Napoleon (). Comparative and superlative quantifiers: pragmatic effects of comparison type. Journal of Semantics, (): –. Cummins, Chris and Rohde, Hannah (). Evoking context with contrastive stress: effects on pragmatic enrichment. Frontiers in Psychology, : . doi: ./fpsyg.. Cummins, Chris, Sauerland, Uli, and Solt, Stephanie (). Granularity and scalar implicature in numerical expressions. Linguistics and Philosophy, : –. Cummins, Chris, Amaral, Patrícia, and Katsos, Napoleon (). Backgrounding and accommodation of presupposition: an experimental approach. In Emmanuel Chemla, Vincent Homer, and Gregoire Winterstein (eds.), Proceedings of Sinn und Bedeutung . Paris: ENS. –. Cutler, Anne (). Phoneme monitoring reaction time as a function of preceding intonation contour. Perception and Psychophysics, : –. Cutler, Anne and Fodor, Jerry A. (). Semantic focus and sentence comprehension. Cognition, : –. doi: ./-()- Cutler, Anne and Foss, Donald J. (). On the role of sentence stress on sentence processing. Language and Speech, : –. Cutler, Anne and Pearson, Mark (). On the analysis of prosodic turn-taking cues. In Catherine Johns-Lewis (ed.), Intonation and Discourse. London: Croom Helm. – D’Imperio, Mariapaola (). Breadth of focus, modality and prominence perception in Neapolitan Italian. Ohio State University Working Papers in Linguistics, : –.
D’Imperio, Mariapaola (). Focus and tonal structure in Neapolitan Italian. Speech Communication, : –. Dahan, Delphine, Tanenhaus, Michael K., and Chambers, Craig G. (). Accent and reference resolution in spoken-language comprehension. Journal of Memory and Language, : –. Dahl, Östen (). On generics. In Edward Keenan (ed.), Formal Semantics of Natural Language. Cambridge: Cambridge University Press. –. Dahl, Östen (). The marking of the episodic/generic distinction in tense-aspect systems. In Greg Carlson and Francis J. Pelletier (eds.), The Generic Book. Chicago: Chicago University Press. –. Dale, Robert and Duran, Nicholas D. (). The cognitive dynamics of negated sentence verification. Cognitive Science, (): –. Dale, Robert and Reiter, Ehud (). Computational interpretations of the Gricean maxims in the generation of referring expressions. Cognitive Science, : –. Dalrymple, Mary, Kanazawa, Makoto, Kim, Yookyung, Mchombo, Sam, and Peters, Stanley (). Reciprocal expressions and the concept of reciprocity. Linguistics and Philosophy, (): –. Davidson, Donald (). The logical form of action sentences. In Nicholas Rescher (ed.), The Logic of Decision and Action. University of Pittsburgh Press. –. Davies, Benjamin, Rattanasone, Nan Xu, and Demuth, Katherine (). Two-year-olds’ sensitivity to inflectional plural morphology: allomorphic effects. Language Learning and Development, : –. Davies, Catherine and Katsos, Napoleon (). Are interlocutors as sensitive to over-informativeness as they are to under-informativeness? Proceedings of the Workshop on Production of Referring Expressions: Bridging Computational and Psycholinguistic Approaches. Amsterdam: CSLI. Davies, Catherine and Katsos, Napoleon (). Over-informative children: production/comprehension asymmetry or tolerance to pragmatic violations? Lingua, (): –. Davies, Catherine and Katsos, Napoleon (). Are speakers and listeners ‘only moderately Gricean’? An empirical response to Engelhardt et al. (). Journal of Pragmatics, (): –. Davies, Catherine and Kreysa, Helene (). Looking at a contrast object before speaking boosts referential informativeness, but is not essential. Acta Psychologica, : –. Davies, Catherine, Andrés-Roqueta, Clara, and Norbury, Courtenay F. (). Referring expressions and structural language abilities in children with specific language impairment: a pragmatic tolerance account. Journal of Experimental Child Psychology, : –. doi: ./j.jecp... Dayal, Veneeta (). Number marking and (in)definiteness in kind terms. Linguistics and Philosophy, : –. De Cat, Cécile (). The cognitive underpinnings of referential abilities. In Ludovica Serratrice and Shanley E. M. Allen (eds.), The Acquisition of Reference. Amsterdam: John Benjamins. –. De Grauwe, Sophie, Swain, Abigail, Holcomb, Phillip, Ditman, Tali, and Kuperberg, Gina R. (). Electrophysiological insights into the processing of nominal metaphors. Neuropsychologia, (): –. doi: ./j.neuropsychologia... De León, Lourdes (). Exploration in the acquisition of geocentric location by Tzotzil children. Linguistics, (–): –. doi: ./ling...-. De Marneffe, Marie-Catherine and Tonhauser, Judith (). On the role of context and prosody in the generation of scalar implicatures of adjectives. Presentation at XPRAG.de Workshop, Formal and experimental pragmatics: Methodological issues of a nascent liaison, Berlin. De Marneffe, Marie-Catherine and Tonhauser, Judith (to appear). Inferring meaning from indirect answers to polar questions: the contribution of the rise-fall-rise contour. In Edgar Onea, Malte Zimmermann, and Klaus von Heusinger (eds.), Questions in Discourse. Leiden: Brill. De Neys, Wim and Schaeken, Walter (). When people are more logical under cognitive load: dual task impact on scalar implicature. Experimental Psychology, (): –. De Ruiter, Jan P. (). Methodological paradigms in interaction research. In Ipke Wachsmuth, Jan P. de Ruiter, Stefan Kopp, and Petra Jaecks (eds.), Alignment in Communication: Towards a New Theory of Communication. Amsterdam: John Benjamins.
De Ruiter, Jan P. and Albert, Saul (). An appeal for a methodological fusion of conversation analysis and experimental psychology. Research on Language and Social Interaction, (): –. De Ruiter, Jan P. and Cummins, Chris (). Forward modelling requires intention recognition and non-impoverished predictions. Behavioral and Brain Sciences, (): . De Ruiter, Jan P., Mitterer, Holger, and Enfield, Nick J. (). Projecting the end of a speaker’s turn: a cognitive cornerstone of conversation. Language, (): –. De Ruiter, Laura E. (). Information status marking in spontaneous vs. read speech in story-telling tasks: evidence from intonation analysis using GToBI. Journal of Phonetics, : –. De Vega, Manuel (). Levels of embodied meaning: from pointing to counterfactuals. In Manuel de Vega, Arthur Glenberg, and Arthur Graesser (eds.), Symbols and Embodiment: Debates on Meaning and Cognition. New York: Oxford University Press. –. De Vega, Manuel and Urrutia, Mabel (). Counterfactual sentences activate embodied meaning: an action-sentence compatibility effect study. Journal of Cognitive Psychology, : –. doi: ./.. De Vega, Manuel and Urrutia, Mabel (). Discourse updating in texts with a counterfactual event. Psicológica, : –. De Vega, Manuel, Urrutia, Mabel, and Riffo, Bernardo (). Cancelling updating in the comprehension of counterfactuals embedded in narratives. Memory and Cognition, : –. Deamer, Felicity and Pouscoulous, Nausicaa (submitted) Non-literal children: pre-schoolers’ understanding of figurative language. Deamer, Felicity, Breheny, Richard, and Pouscoulous, Nausicaa (). A contrastive look at the processing of metaphor and hyperbole. UCL Working Papers in Linguistics, : –. Deane, Paul D. (). Polysemy and cognition. Lingua, (): –. Declerck, Renaat (). Restrictive when-clauses. Linguistics and Philosophy, : –. Declerck, Renaat (). The origins of genericity, Linguistics, (): –. Degen, Judith (). Alternatives in pragmatic reasoning. PhD thesis. University of Rochester. Degen, Judith (). Investigating the distribution of ‘some’ (but not ‘all’) implicatures using corpora and web-based methods. Semantics and Pragmatics, (), –. doi: ./sp.. Degen, Judith and Goodman, Noah D. (). Lost your marbles? The puzzle of dependent measures in experimental pragmatics. In Paul Bello, Marcello Guarini, Marjorie McShane, and Brian Scassellati (eds.), Proceedings of the th Annual Conference of the Cognitive Science Society. Austin, TX: Cognitive Science Society. –. Degen, Judith and Tanenhaus, Michael K. (). Processing scalar implicature: a constraint-based approach. Cognitive Science, (): –. doi: ./cogs. Degen, Judith and Tanenhaus, Michael K. (). Availability of alternatives and the processing of scalar implicatures: a visual world eye-tracking study. Cognitive Science, (): –. doi: ./ cogs. Degen, Judith, Franke, Michael, and Jäger, Gerhard (). Cost-based pragmatic inference about referential expressions. In Markus Knauff, Michael Pauen, Natalie Sebanz, and Ipke Wachsmuth (eds.), Proceedings of the th Annual Conference of the Cognitive Science Society. Austin, TX: Cognitive Science Society. –. Degen, Judith, Franke, Michael, and Goodman, Noah D. (eds.) (). Proceedings of the ESSLLI Formal and Experimental Pragmatics Workshop. Tübingen: ESSLLI. Degen, Judith, Tessler, Michael H., and Goodman, Noah D. (). Wonky worlds: listeners revise world knowledge when utterances are odd. In David C. Noelle, Rick Dale, Anne Warlaumont et al. (eds.), Proceedings of the th Annual Conference of the Cognitive Science Society, : –. Degen, Judith, Kreiss, Elisa, Hawkins, Robert X. D., and Goodman, Noah D. (). Mentioning atypical properties of objects is communicatively efficient. Presentation at CUNY. Dehaene, Stanislas and Mehler, Jacques (). Cross-linguistic regularities in the frequency of number words. Cognition, : –. Del Pinal, Guillermo (). Prototypes as compositional components of concepts. Synthese, : –.
Del Pinal, Guillermo, Madva, Alex, and Reuter, Kevin (). Stereotypes, conceptual centrality and gender bias: an empirical investigation. Ratio, (): –. Delin, Judy (). A multi-level account of cleft constructions in discourse. In Hans Karlgren (ed.), COLING-: Proceedings of the th International Conference on Computational Linguistics. Helsinki: University of Helsinki. –. Delin, Judy (). Properties of it-cleft presupposition. Journal of Semantics, : –. Delin, Judy and Oberlander, Jon (). Syntactic constraints on discourse structure: the case of it-clefts. Linguistics, : –. DeLong, Katherine A., Urbach, Thomas P., and Kutas, Marta (). Probabilistic word preactivation during language comprehension inferred from electrical brain activity. Nature Neuroscience, (): –. DeLong, Katherine A., Quante, Laura, and Kutas, Marta (). Predictability, plausibility, and two late ERP positivities during written sentence comprehension. Neuropsychologia, : –. Deo, Ashwini and Madiman, Mokshay (). Scale structure and genericity: understanding generics using stochastic comparison. http://ling.auf.net/lingbuzz/ Dessalegn, Banchiamlack and Landau, Barbara (). More than meets the eye: the role of language in binding and maintaining feature conjunctions. Psychological Science, (): –. doi: ./ j.–...x Dessalegn, Banchiamlack and Landau, Barbara (). Interaction between language and vision: it’s momentary, abstract, and it develops. Cognition, (): –. doi: ./j.cognition... Destruel, Emilie (). The French c’est-cleft: empirical studies of its meaning and use. PhD thesis. University of Texas at Austin. Destruel, Emilie and Velleman, Leah (). Refining constrast: empirical evidence from the English it-cleft. In Christopher Piñon (ed.), Empirical Issues in Syntax and Semantics . Paris. –. Destruel, Emilie, Velleman, Daniel, Onea, Edgar, Bumford, Dylan, Xue, Jingyang, and Beaver, David (). A cross-linguistic study of the non-at-issueness of exhaustive inferences. In Florian Schwarz (ed.), Experimental Perspectives on Presuppositions. Cham: Springer. –. Destruel, Emilie, Beaver, David, and Coppock, Elizabeth (). Clefts: quite the contrary! Presentation at Sinn und Bedeutung , Edinburgh. Deutsch, Werner and Pechmann, Thomas (). Social-interaction and the development of definite descriptions. Cognition, (): –. doi: ./-()- DeVeaugh-Geiss, Joseph P., Zimmermann, Malte, Onea, Edgar, and Boell, Anna-Christina (). Contradicting (not-)at-issueness in exclusives and clefts: an empirical study. In Sarah D’Antonio, Mary Moroney, and Carol Rose Little (eds.), Proceedings of the th Semantics and Linguistic Theory Conference. Washington, DC: Linguistic Society of America. –. DeVeaugh-Geiss, Joseph P., Tönnis, Swantje, Onea, Edgar, and Zimmermann, Malte (). That’s not quite it: an experimental investigation of (non-) exhaustivity in it-clefts. Presentation at Sinn und Bedeutung , Edinburgh. Dews, Shelly and Winner, Ellen (). Obligatory processing of literal and nonliteral meanings in verbal irony. Journal of Pragmatics, : –. doi: ./S-()- Dhami, Mandeep K. (). On measuring quantitative interpretations of reasonable doubt. Journal of Experimental Psychology: Applied, : –. doi: ./a Dhami, Mandeep K., Mandel, David R., Mellers, Barbara A., and Tetlock, Philip E. (). Improving intelligence analysis with decision science. Perspectives on Psychological Science, : –. Di Paola, Simona, Canal, Paolo, Ricci, Irene, Bertini, Chiara, Bertinetto, Pier Marco, and Bambini, Valentina (). Metaphor out of school: electrophysiological correlates of metaphor processing in lower and higher literates. Poster presented at the th Annual CUNY Conference on Human Sentence Processing. Gainsville, University of Florida. Dickey, Michael Walsh and Bunger, Ann C. (). Comprehension of elided structure: evidence from sluicing. Language and Cognitive Processes, (): –. doi: ./ Diesendruck, Gil (). The principles of conventionality and contrast in word learning: an empirical examination. Developmental Psychology, : –. doi: ./-...
Diesendruck, Gil and Markson, Lori (). Children’s avoidance of lexical overlap: a pragmatic account. Developmental Psychology, : –. doi: ./-... Dieussaert, Kristien, Verkerk, Suzanne, Gillard, Ellen, and Schaeken, Walter (). Some effort for some: further evidence that scalar implicatures are effortful. Quarterly Journal of Experimental Psychology (), (): –. doi: ./.. Dillon, Brian, Mishler, Alan, Sloggett, Shayne, and Phillips, Colin (). Contrasting intrusion profiles for agreement and anaphora: experimental and modeling evidence. Journal of Memory and Language, (): –. Domaneschi, Filippo, Carrea, Elena, Penco, Carlo, and Greco, Alberto (). The cognitive load of presupposition triggers: mandatory and optional repairs in presupposition failure. Language and Cognitive Processes, (): –. doi: ./.. Donaldson, Margaret (). Children’s Minds. London: Fontana. Donaldson, Margaret and Lloyd, Peter (). Sentences and situations: children’s judgments of match and mismatch. In François Bresson (ed.), Problèmes Actuels en Psycholinguistique, Colloques Internationaux du Centre National de Recherche Scientifique. Paris: Centre National de Recherche Scientifique. –. Donaldson, Margaret and McGarrigle, James (). Some clues to the nature of semantic development. Journal of Child Language, (): –. Donaldson, Margaret and Wales, Roger (). On the acquisition of some relational terms. In John R. Hayes (ed.), Cognition and the Development of Language. New York: Wiley. Doran, Ryan, Baker, Rachel E., McNabb, Yaron, Larson, Meredith, and Ward, Gregory (). On the non-unified nature of scalar implicature: an empirical investigation. International Review of Pragmatics, : –. Doran, Ryan, Ward, Gregory, Larson, Meredith, McNabb, Yaron, and Baker, Rachel E. (). A novel paradigm for distinguishing between what is said and what is implicated. Language, : –. Dotlačil, Jakub and Brasoveanu, Adrian (). The manner and time course of updating quantifier scope representations in discourse. Language, Cognition and Neuroscience, (): –. Doupnik, Timothy S. and Richter, Martin (). The impact of culture on the interpretation of ‘in context’ verbal probability expressions. Journal of International Accounting Research, : –. doi: ./jiar.... Downing, Laura (). Focus and prominence in Chichewa, Chitumbuka and Durban Zulu. ZAS Papers in Linguistics, : –. Dowty, David R. (). On the syntax and semantics of the atomic predicate CAUSE. In Paul M. Peranteau, Judith N. Levi, and Gloria C. Phares (eds.), Proceedings of the Eighth Regional Meeting of the Chicago Linguistic Society. Chicago: Chicago Linguistic Society. –. Dowty, David R. (). Word Meaning and Montague Grammar: The Semantics of Verbs and Times in Generative Semantics and in Montague’s PTO. Dordrecht: Reidel. Dowty, David R. (). Collective predicates, distributive predicates, and all. In Fred Marshall (ed.), Proceedings of the rd Eastern States Conference on Linguistics. Columbus, OH: The Ohio State University. –. Dowty, David R. (). Thematic proto-roles and argument selection. Language, (): –. Drayton, Stefane, Turley-Ames, Kandi J., and Guajardo, Nicole R. (). Counterfactual thinking and false belief: the role of executive function. Journal of Experimental Child Psychology, : –. doi: ./j.jecp... Drenhaus, Heiner, Saddy, Douglas, and Frisch, Stefan (). Processing negative polarity items: when negation comes through the backdoor. In Stephan Kepser and Marga Reis (eds.), Linguistic Evidence: Empirical, Theoretical, and Computational Perspectives. Berlin: De Gruyter Mouton. –. Drenhaus, Heiner, beim Graben, Peter, Saddy, Douglas, and Frisch, Stefan (). Diagnosis and repair of negative polarity constructions in the light of symbolic resonance analysis. Brain and Language, (): –.
Drenhaus, Heiner, Blaszczak, Joanna, and Schütte, Juliane (). Some psycholinguistic comments on NPI licensing. In Estela Puig-Waldmüller (ed.), Proceedings of Sinn und Bedeutung . Barcelona: Universitat Pompeu Fabra. –. Drenhaus, Heiner, Zimmermann, Malte, and Vasishth, Shravan (). Exhaustiveness effects in clefts are not truth-functional. Journal of Neurolinguistics, : –. Dressler, Wolfgang U., Mayerthaler, Willi, Panagl, Oswald, and Wurzel, Wolfgang U. (). Leitmotifs in Natural Morphology. Philadelphia: Benjamins. Drew, Paul (). ‘Quit talking while I’m interrupting’: a comparison between positions of overlap onset in conversation. In Markku Haakana, Minna Laakso, and Jan Lindström (eds.), Talk in Interaction: Comparative Dimensions. Helsinki: Finnish Literature Society. Drozd, Kenneth F. (). Children’s weak interpretation of universally quantified sentences. In Melissa Bowerman and Stephen C. Levinson (eds.), Language Acquisition and Conceptual Development. Cambridge: Cambridge University Press. –. Drozd, Kenneth F. and Philip, William (). Event quantification in preschoolers’ comprehension of negation. In Eve V. Clark (ed.), Proceedings of the th Annual Child Language Research Forum. Stanford, CA: Center for the Study of Language and Information. –. Drozd, Kenneth F. and Van Loosbroek, Erik (). Weak quantification, plausible dissent, and the development of children’s pragmatic competence. In Annabel Greenhill, Heather Littlefield, and Cheryl Tano (eds.), Proceedings of the d Annual Boston University Conference on Language Development. Somerville, MA: Cascadilla Press. –. Drozd, Kenneth F. and Van Loosbroek, Erik (). The effect of context on children’s interpretations of universally quantified sentences. In Veerle van Geenhoven (ed.) Semantics in Acquisition. Dordrecht: Springer. –. Druzdzel, Marek J. (). Verbal Uncertainty Expressions: Literature Review. Pittsburgh, PA: Carnegie Mellon University, Department of Engineering and Public Policy. Dryer, Matthew S. (). Focus, pragmatic presupposition and activated propositions. Journal of Pragmatics, : –. Dudley, Rachel, Orita, Naho, Hacquard, Valentine, and Lidz, Jeffrey (). Three-year-olds’ understanding of know and think. In Florian Schwarz (ed.), Experimental Perspectives on Presuppositions. Cham: Springer International Publishing. –. Dudley, Rachel, Rowe, Meredith, Hacquard, Valentine, and Lidz, Jeffrey (under review). Acquiring the factivity of know. Duffy, Susan A., Morris, Robin K., and Rayner, Keith (). Lexical ambiguity and fixation times in reading. Journal of Memory and Language, (): –. Dulcinati, Giulio, Mazzarella, Diana, Pouscoulous, Nausicaa, and Rodd, Jennifer (). Processing metaphor: the role of conventionality, familiarity and dominance. UCL Working Papers in Linguistics, : –. Duncan, Starkey (). Some signals and rules for taking speaking turns in conversations. Journal of Personality and Social Psychology, (): –. doi: ./h Duncan, Starkey (). On the structure of speaker-auditor interaction during speaking turns. Language in Society, (): –. Duncan, Starkey and Fiske, Donald W. (). Face-to-Face Interaction: Research, Methods and Theory. Hillsdale, NJ: Lawrence Erlbaum. Dunham, Philip, Dunham, Frances, and O’Keefe, Colleen (). Two-year-olds’ sensitivity to a parent’s knowledge state: mind reading or contextual cues? British Journal of Developmental Psychology, (): –. doi: ./ Dunn, Judy, Brown, Jane, Slomkowski, Cheryl, Tesla, Caroline, and Youngblade, Lise (). Young children’s understanding of other people’s feelings and beliefs: individual differences and their antecedents. Child Development, (): –. doi: ./j.-..tb.x Durkheim, Emile (). The Elementary Forms of Religious Life. London: Allen and Unwin. Dwivedi, Veena D. (). Interpreting quantifier scope ambiguity: evidence of heuristic first, algorithmic second processing. PLoS ONE, (): e.
Dwivedi, Veena D., Phillips, Natalie A., Einagel, Stephanie, and Baum, Shari R. (). The neural underpinnings of semantic ambiguity and anaphora. Brain Research, (): –. doi: ./ j.brainres... Eady, Stephen and Cooper, William E. (). Speech intonation and focus location in matched statements and questions. Journal of the Acoustical Society of America, : –. Eady, Stephen J., Cooper, William E., Klouda, Gayle V., Mueller, Pamela R., and Lotts, Dan W. (). Acoustical characteristics of sentential focus: narrow vs. broad and single vs. dual focus environments. Language and Speech, : –. Ebeling, Karen S. and Gelman, Susan A. (). Coordination of size standards by young children. Child Development, : –. Ebert, Christian (). Formal investigations of underspecified representations. PhD thesis. London: King’s College. Eckert, Penelope (). Jocks and Burnouts: Social Categories and Identity in the High School. London: Teachers College Press. Eckert, Penelope (). Three waves of variation study : the emergence of meaning in the study of sociolinguistic variation. Annual Review of Anthropology, , –. doi: ./annurev-anthro- Eelen, Gino (). A Critique of Politeness Theories. Manchester: St Jerome. Egg, Markus (). Metonymie als Phänomen der Semantik-Pragmatik-Schnittstelle. Metaphorik, : –. Égré, Paul and Zehr, Jérémy (). Are gaps preferred to gluts? A closer look at borderline contradictions. In Elena Castroviejo, Galit Sassoon, and Louise McNally (eds.), The Semantics of Gradability, Vagueness, and Scale Structure: Experimental Perspectives. New York: Springer. Égré, Paul, de Gardelle, Vincent, and Ripley, David (). Vagueness and order effects in color categorization. Journal of Logic, Language and Information, : –. Ehrlich, Kate and Rayner, Keith (). Pronoun assignment and semantic integration during reading: eye-movements and immediacy of processing. Journal of Verbal Learning and Verbal Behavior, : –. doi: ./S-()- Eilers, Rebecca E., Oller, D. Kimbrough, and Ellington, Judy (). The acquisition of word meaning for dimensional adjectives: the long and short of it. Journal of Child Language, : –. Elkind, David (). Children’s conception of right and left: Piaget replication study IV. Journal of Genetic Psychology, (): –. doi: ./.. Elman, Jeffrey L., Hare, Mary, and McRae, Ken (). Cues, constraints, and competition in sentence processing. In Michael Tomasello and Dan I. Slobin (eds.), Beyond Nature-Nurture: Essays in Honor of Elizabeth Bates. Mahwah, NJ: Erlbaum. –. Engelhardt, Paul E. (). Generalized Implicatures: do uninformative environments eliminate default interpretations? PhD thesis. Michigan State University. Engelhardt, Paul E., Bailey, Karl G. D., and Ferreira, Fernanda (). Do speakers and listeners observe the Gricean Maxim of Quantity? Journal of Memory and Language, (): –. doi: ./j.jml... Engelhardt, Paul E., Demiral, Ş. Baris, and Ferreira, Fernanda (). Overspecified referring expressions impair comprehension: an ERP study. Brain and Cognition, : –. ./j.bandc... Erickson, Bonnie E., Lind, E. Allen, Johnson, Bruce C., and O’Barr, William M. (). Speech style and impression formation in a court setting: the effects of ‘powerful’ and ‘powerless’ speech. Journal of Experimental Social Psychology, (): –. Erickson, Thomas D. and Mattson, Mark E. (). From words to meaning: a semantic illusion. Journal of Verbal Learning and Verbal Behavior, : –. doi: ./S-()- Ernst, Marc O. and Banks, Martin S. (). Humans integrate visual and haptic information in a statistically optimal fashion. Nature, (): –. doi: ./a Eslinger, Paul J. (). Conceptualizing, describing, and measuring components of executive function. In G. Reid Lyon and Norman A. Krasnegor (eds.), Attention, Memory, and Executive Function. Baltimore: Paul H. Brookes. –.
Evang, Kilian and Bos, Johan (). Scope disambiguation as a tagging task. In Katrin Erk and Alexander Koller (eds.), Proceedings of the th International Conference on Computational Semantics. Potsdam: Association for Computational Linguistics. –. Evans, Jonathan St. B. T. (). Interpretation and matching bias in a reasoning task. Quarterly Journal of Experimental Psychology, : –. doi: ./ Evans, Jonathan St. B. T. (). Matching bias in conditional reasoning: do we understand it after years? Thinking and Reasoning, (): –. Evans, Jonathan St. B. T. and Over, David E. (). If. Oxford: Oxford University Press. Evans, Jonathan St. B. T., Clibbens, John, and Rood, Benjamin (). The role of implicit and explicit negation in conditional reasoning bias. Journal of Memory and Language, (): –. doi: ./jmla.. Evans, Jonathan St. B. T., Over, David E., and Handley, Simon J. (). Suppositions, extensionality, and conditionals: a critique of the mental model theory of Johnson-Laird and Byrne (). Psychological Review, (): –. doi: ./-X... Everett, Daniel (). Cultural constraints on grammar and cognition in Pirahã: another look at the design features of human language. Current Anthropology, (): –. Eviatar, Zohar and Just, Marcel Adam M. (). Brain correlates of discourse processing: an fMRI investigation of irony and conventional metaphor comprehension. Neuropsychologia, (), –. Faber, Myrthe and Gennari, Silvia P. (). Representing time in language and memory: the role of similarity structure. Acta Psychologica, : –. doi: ./j.actpsy... Falkum, Ingrid L. (). The how and why of polysemy: a pragmatic account. Lingua, , –. Falkum, Ingrid L., Recasens, Marta, and Clark, Eve V. (). ‘The moustache sits down first’: on the acquisition of metonymy. Journal of Child Language, (): –. doi: ./ S Fara, Delia Graff (). Shifting sands: an interest-relative theory of vagueness. Philosophical Topics, : –. Originally published under the name ‘Delia Graff ’. Farkas, Donka (). Quantifier scope and syntactic islands. In Roberta A. Hendrick, Carrie S. Masek, and Mary Frances Miller (eds.), Papers from the Seventeenth Regional Meeting of the Chicago Linguistic Society. Chicago: University of Chicago, Chicago Linguistic Society. –. Farkas, Donka and De Swart, Henriette (). The semantics and pragmatics of plurals. Semantics and Pragmatics, (): –. Farkas, Donka and Giannakidou, Anastasia (). How clause-bounded is the scope of universals? In Theresa Galloway and Justin Spence (eds.), Proceedings of Semantics and Linguistic Theory VI. Ithaca, NY: Cornell University, CLC Publications. –. Farkas, Donka and Sugioka, Yoko (). Restrictive if/when clauses. Linguistics and Philosophy, (): –. Fauconnier, Gilles (). Polarity and the scale principle. In Robin E. Grossman (ed.) Papers from the Eleventh Meeting of the Chicago Linguistic Society. Chicago Linguistic Society. –. Fauconnier, Gilles (). Mental Spaces: Aspects of Meaning Construction in Natural Language. Cambridge, MA: MIT Press. Fauconnier, Gilles (). Mental Spaces. Cambridge: Cambridge University Press. Fauconnier, Gilles (). Mappings in Thought and Language. Cambridge: Cambridge University Press. Fauconnier, Gilles and Turner, Mark (). Conceptual integration networks. Cognitive Science, : –. Fauconnier, Gilles and Turner, Mark (). The Way We Think: Conceptual Blending and the Mind’s Hidden Complexities. New York: Basic Books. Featherston, Sam (). Data in generative grammar: the stick and the carrot. Theoretical Linguistics, (): –. Federmeier, Kara D. and Kutas, Marta (). A rose by any other name: long-term memory structure and sentence processing. Journal of Memory and Language, (): –.
Feeney, Aidan and Bonnefon, Jean-François (). Politeness and honesty contribute additively to the interpretation of scalar expressions. Journal of Language and Social Psychology, : –. doi: ./X Feeney, Aidan, Scrafton, Susan, Duckworth, Amber, and Handley, Simon J. (). The story of some: everyday pragmatic inference by children and adults. Canadian Journal of Experimental Psychology, (): –. doi: ./h Fein, Ofer, Yeari, Menahem, and Giora, Rachel (). On the priority of salience-based interpretations: the case of sarcastic irony. Intercultural Pragmatics, (): –. doi: ./ip-– Feist, Michele I. (). Motion through syntactic frames. Cognition, (): –. Fenson, Larry, Dale, Philip S., Reznick, J. Steven, Bates, Elizabeth, Thal, Donna J., and Pethick, Stephen J. (). Variability in Early Communicative Development. Monograph of the Society for Research in Child Development, . Wiley. doi: ./ Ferenz, Krag and Prasada, Sandeep (). Singular or plural? Children’s knowledge of the factors that determine the appropriate form of count nouns. Journal of Child Language, : –. Ferguson, Heather J. (). Eye movements reveal rapid concurrent access to factual and counterfactual interpretations of the world. Quarterly Journal of Experimental Psychology, (): –. Ferguson, Heather J. and Breheny, Richard (). Listeners’ eyes reveal spontaneous sensitivity to others’ perspectives. Journal of Experimental Social Psychology, (): –. Ferguson, Heather J. and Cane, James E. (). Examining the cognitive costs of counterfactual language comprehension: evidence from ERPs. Brain Research, : –. Ferguson, Heather J. and Sanford, Anthony J. (). Anomalies in real and counterfactual worlds: an eye-movement investigation. Journal of Memory and Language, : –. Ferguson, Heather J., Sanford, Anthony J., and Leuthold, Hartmut (). Real-world interference in detecting violations of counterfactual and negated worlds. Conference paper presented at Cognitive Neuroscience Society Annual meeting, New York, NY. Ferguson, Heather J., Sanford, Anthony J., and Leuthold, Hartmut (). Eye-movements and ERPs reveal the time-course of processing negation and remitting counterfactual worlds. Brain Research, : –. Ferguson, Heather J., Breheny, Richard, Sanford, Anthony J., and Scheepers, Christoph (). Reading the minds of others: disentangling the gender-specific mechanisms. Conference paper presented at Annual Meeting of the Cognitive Neuroscience Society. San Francisco, CA. Ferguson, Heather J., Scheepers, Christoph, and Sanford, Anthony J. (). Expectations in counterfactual and theory of mind reasoning. Language and Cognitive Processes, : –. Ferguson, Heather J., Cane, James E., Douchkov, Michelle, and Wright, Daniel (). Empathy predicts false belief reasoning ability: evidence from the N. Social Cognitive and Affective Neuroscience, : –. doi: ./scan/nsu Fernando, Tim and Hans Kamp (). Expecting many. In Teresa Galloway and Justin Spence (eds.), Proceedings of Semantics and Linguistic Theory . Ithaca, NY: CLC Publications. Ferreira, Fernanda (). Psycholinguistics, formal grammars, and cognitive science. Linguistic Review, : –. Ferreira, Fernanda and Clifton, Charles, Jr. (). The independence of syntactic processing. Journal of Memory and Language, (): –. doi: ./-X()- Ferreira, Fernanda, Bailey, Karl G. D., and Ferraro, Vittorio (). Good-enough representations in language comprehension. Current Directions in Psychological Science, (): –. doi: ./. Ferreira, Victor S., Slevc, L. Robert, and Rogers, Erin S. (). How do speakers avoid ambiguous linguistic expressions? Cognition, (): –. doi: ./j.cognition... Ferstl, Evelyn C., Garnham, Alan, and Manouilidou, Christina (). Implicit causality bias in English: a corpus of verbs. Behavior Research Methods, : –. Féry, Caroline (). Information structural notions and the fallacy of invariant correlates. In Caroline Féry, Gisbert Fanselow, and Manfred Krifka (eds.), Interdisciplinary Studies on Information Structure . Potsdam. –.
Féry, Caroline (). Focus as prosodic alignment. Natural Language and Linguistic Theory, : –. Féry, Caroline and Krifka, Manfred (). Information structure: notional distinctions, ways of expression. In Piet van Sterkenburg (ed.), Unity and Diversity of Languages. Amsterdam: John Benjamins. –. Fiengo, Robert and May, Robert (). Indices and Identity. Cambridge, MA: MIT Press. Filik, Ruth and Leuthold, Hartmut (). The role of character-based knowledge in online narrative comprehension: evidence from eye movements and ERPs. Brain Research, : –. Filik, Ruth, Paterson, Kevin B., and Liversedge, Simon P. (). Processing doubly quantified sentences: evidence from eye movements. Psychonomic Bulletin and Review, (): –. Filik, Ruth, Paterson, Kevin B., and Liversedge, Simon P. (). Parsing with focus particles in context: eye movements during the processing of relative clause ambiguities. Journal of Memory and Language, : –. Filik, Ruth, Paterson, Kevin B., and Liversedge, Simon P. (). The influence of only and even on online semantic interpretation. Psychonomic Bulletin and Review, (): –. doi: ./ PBR... Filik, Ruth, Leuthold, Hartmut, Wallington, Katie, and Page, Jemma (). Testing theories of irony processing using eye-tracking and ERPs. Journal of Experimental Psychology: Learning, Memory, and Cognition, (): –. doi: ./a Fillenbaum, Samuel (). Or: Some uses. Journal of Experimental Psychology, (): –. Fillmore, Charles J. (). The case for case. In Emmon Bach and Robert T. Harms (eds.), Universals in Linguistic Theory. New York: Holt, Rinehart, and Winston. –. Fine, Kit (). Vagueness, truth and logic. Synthese, : –. Reprinted in Keefe and Smith (). Fischler, Ira, Bloom, Paul A., Childers, Donald G., Roucos, Salim E., and Perry, Nathan W. (). Brain potentials related to stages of sentence verification. Psychophysiology, (): –. doi: ./j.-..tb.x Fitzgerald, F. Scott (). The Great Gatsby. Cambridge: Cambridge University Press (first published ). Fodor, Janet Dean (). The mental representation of quantifiers. In Stanley Peters and Esa Saarinen (eds.), Processes, Beliefs and Questions. Dordrecht: Reidel. –. Fodor, Jerry A. (). Three reasons for not deriving ‘kill’ from ‘cause to die’. Linguistic Inquiry, (): –. Fodor, Jerry A. (). The Modularity of Mind: An Essay on Faculty Psychology. Bradford Books, MIT Press. Folli, Raffaella and Harley, Heidi (). Causation, obligation, and argument structure: on the nature of little v. Linguistic Inquiry, (): –. Foppolo, Francesca, Guasti, Maria Teresa, and Chierchia, Gennaro (). Scalar implicatures in child language: give children a chance. Language Learning and Development, : –. Foraker, Stephani and McElree, Brian (). The role of prominence in pronoun resolution: active versus passive representations. Journal of Memory and Language, : –. Foraker, Stephani and Murphy, Gregory L. (). Polysemy in sentence comprehension: effects of meaning dominance. Journal of Memory and Language, (): –. Ford, Cecilia E. and Thompson, Sandra A. (). Interactional units in conversation: syntactic, intonational, and pragmatic resources for the management of turns. In Elinor Ochs, Emanuel A. Schegloff and Sandra A. Thompson (eds.), Interaction and Grammar. Cambridge: Cambridge University Press. –. Forgács, Bálint (). Figures of language in cognitive science in the light of figurative language processing in the brain. PhD thesis. Budapest University of Technology and Economics (BME). Forgács, Bálint, Bohrn, Isabel, Baudewig, Jürgen, Hofmann, Markus J., Pléh, Csaba, and Jacobs, Arthur M. (). Neural correlates of combinatorial semantic processing of literal and figurative noun noun compound words. NeuroImage, (): –. doi: ./j.neuroimage...
Forgács, Bálint, Lukács, Ágnes, and Pléh, Csaba (). Lateralized processing of novel metaphors: disentangling figurativeness and novelty. Neuropsychologia, : –. Forgas, Joseph P. (a). Feeling and speaking: mood effects on verbal communication strategies. Personality and Social Psychology Bulletin, (): –. doi: ./ Forgas, Joseph P. (b). On feeling good and being rude: affective influences on language use and request formulations. Journal of Personality and Social Psychology, (): –. doi: ./ -... Forster, Kenneth I. (). Levels of processing and the structure of the language processor. In William E. Cooper and Edward C. T. Walker (eds.), Sentence Processing: Psycholinguistic Studies presented to Merrill Garrett. Hillsdale, NJ: Erlbaum. Forster, Kenneth I. and Chris Davis (). Repetition priming and frequency attenuation in lexical access. Journal of Experimental Psychology: Learning, Memory and Cognition, (): –. Fox, Danny (a). Economy and scope. Natural Language Semantics, : –. Fox, Danny (b). Condition C effects in ACD. In Rob Pensalfini and Hiroyuki Ura (eds.), Papers on Minimalist Syntax. MIT Working Papers in Linguistics . Cambridge, MA. –. Fox, Danny (). Economy and Semantic Interpretation. Cambridge, MA: MIT Press. Fox, Danny (). Antecedent-contained deletion and the copy theory of movement. Linguistic Inquiry, : –. Fox, Danny (). Free choice and the theory of scalar implicatures. In Uli Sauerland and Penka Stateva (eds.), Presupposition and Implicature in Compositional Semantics. Houndmills: Palgrave Macmillan. –. Fox, Danny (). Two short notes on Schlenker’s theory of presupposition projection. Theoretical Linguistics, (): –. Fox, Danny (a). Cancelling the maxim of quantity: another argument for a grammatical derivation of scalar implicatures. Ms. MIT/HUJI. Fox, Danny (b). Presupposition projection from quantificational sentences: trivalence, local accommodation, and presupposition strengthening. Ms. Hebrew University of Jerusalem. Fox, Danny and Katzir, Roni (). On the characterization of alternatives. Natural Language Semantics, (): –. Francik, Ellen Palmer (). Referential choice and focus of attention in narratives (discourse anaphora, topic continuity, language production). PhD thesis. Stanford University. Frank, Michael C. and Goodman, Noah D. (). Predicting pragmatic reasoning in language games. Science, : . doi: ./science. Frank, Michael C. and Goodman, Noah D. (). Inferring word meanings by assuming that speakers are informative. Cognitive Psychology, : –. Frank, Michael C., Goodman, Noah D., and Tenenbaum, Joshua B. (). Using speakers’ referential intentions to model early cross-situational word learning. Psychological Science, (): –. Franke, Michael (). Signal to act: game theory in pragmatics. PhD thesis. University of Amsterdam. Franke, Michael (). Quantity implicatures, exhaustive interpretation, and rational conversation. Semantics and Pragmatics, (): –. doi: ./sp.. Franke, Michael (). Typical use of quantifiers: a probabilistic model. In Paul Bello, Marcello Guarini, Marjorie McShane, and Brian Scassellati (eds.), Proceedings of the th Annual Meeting of the Cognitive Science Society. Austin, TX: Cognitive Science Society. –. Franke, Michael and Degen, Judith (). Reasoning in reference games: individual- vs. populationlevel probabilistic modeling. PLoS ONE, (): –. doi: ./journal.pone. Franke, Michael and Jäger, Gerhard (). Probabilistic pragmatics, or why Bayes’ rule is probably important for pragmatics. Zeitschrift für Sprachwissenschaft, : –. Fraser, Bruce and Nolen, William (). The association of deference with linguistic form. International Journal of the Sociology of Language, : –. doi: ./ijsl... Fraundorf, Scott H., Watson, Duane G., and Benjamin, Aaron S. (). Recognition memory reveals just how CONTRASTIVE contrastive accenting really is. Journal of Memory and Language, : –. doi: ./j.jml...
Frazier, Lyn (). On comprehending sentences: syntactic parsing strategies. PhD thesis. University of Connecticut, Storrs, CT. Distributed by the Indiana University Linguistics Club, Bloomington, IN. Frazier, Lyn (). Sentence processing: a tutorial review. In Max Coltheart (ed.), Attention and Performance XII: The Psychology of Reading. London: Erlbaum. –. Frazier, Lyn (). The big fish in a small pond: accommodation and the processing of novel definites. Ms. University of Massachusetts, Amherst. Frazier, Lyn and Clifton, Charles, Jr. (). Sentence reanalysis, and visibility. In Janet D. Fodor and Fernanda Ferreira (ed.), Reanalysis in Sentence Processing. Springer Netherlands. –. Frazier, Lyn and Rayner, Keith (). Taking on semantic commitments: processing multiple meanings vs. multiple senses. Journal of Memory and Language, (): –. Frazier, Lyn, Pacht, Jeremy M., and Rayner, Keith (). Taking on semantic commitments, II: collective versus distributive readings. Cognition, : –. Frazier, Lyn, Clifton, Charles, Jr., and Carlson, Katy (). Focus and VP ellipsis. Language and Speech, (): –. Frazier, Lyn, Clifton, Charles, Jr., and Stolterfoht, Britta (). Scale structure: processing minimum standard and maximum standard scalar adjectives. Cognition, : –. Frederiksen, John R. (). Understanding anaphora: rules used by readers in assigning pronominal referents. Discourse Processes, : –. doi: ./ Freeman, Norman H. (). Reasonable errors in basic reasoning. Educational Psychology, (–): –. Freeman, Norman H. and Stedmon, Jacqueline A. (). How children deal with natural language quantification. In Ida Kurcz, Grace W. Shugar, and Joseph H. Danks (eds.), Knowledge and Language. North-Holland: Elsevier. –. Freeman, Norman H., Sinha, Chris, and Stedmon, Jacqueline (). All the cars—which cars? From word meaning to discourse analysis. In Michael Beveridge (ed.), Children Thinking through Language. London: Edward Arnold. –. Frege, Gottlob (). Begriffsschrift, eine der arithmetischen nachgebildete formelsprache des reinen denkens. In Jean van Heijenoort (ed.), From Frege to Gödel: A Source Book in Mathematical Logic, –. Cambridge, MA: Harvard University Press, . Translated as Concept Script, a formal language of pure thought modelled upon that of arithmetic, by Stefan Bauer-Mengelberg. Frege, Gottlob (). On sense and reference. In Peter Geach and Max Black (eds.), Translations from the Philosophical Writings of Gottlob Frege. Oxford: Blackwell. –. Fricker, Miranda (). Epistemic Injustice: Power and the Ethics of Knowing. Oxford: Oxford University Press. Friederici, Angela D. and Weissenborn, Jürgen (). Mapping sentence form onto meaning: the syntax–semantic interface. Brain Research, : –. Frisson, Steven (). Semantic underspecification in language processing. Language and Linguistics Compass, (): –. Frisson, Steven and Frazier, Lyn (). Carving up word meaning: portioning and grinding. Journal of Memory and Language, (): –. Frisson, Steven and Pickering, Martin J. (). The processing of metonymy: evidence from eye movements. Journal of Experimental Psychology: Learning, Memory and Cognition, (): –. Frisson, Steven and Pickering, Martin J. (). The processing of familiar and novel senses of a word: why reading Dickens is easy but reading Needham can be hard. Language and Cognitive Processes, (): –. Frota, Sónia (). The prosody of focus: a case-study with cross-linguistic implications. In Bernard Bel and Isabelle Marlien (eds.), Proceedings of Speech Prosody . Aix-en-Provence: Speech Prosody. –. Fuellenbach, Kim and Husband, E. Matthew (). The role of determiners in generalizing principled and statistical connections. Talk given at the Workshop on the Semantic Contribution of Det and Num. (In)definiteness, Genericity and Referentiality, Barcelona.
Fukumura, Kumiko and Van Gompel, Roger P. G. (). Choosing anaphoric expressions: do people take into account likelihood of reference? Journal of Memory and Language, : –. Fukumura, Kumiko and Van Gompel, Roger P. G. (). Effects of order of mention and grammatical role on anaphor resolution. Journal of Experimental Psychology: Learning, Memory, and Cognition, (): –. doi: ./xlm Fukumura, Kumiko, Hyönä, Jukka, and Scholfield, Merete (). Gender affects semantic competition: the effect of gender in a non-gender-marking language. Journal of Experimental Psychology: Learning, Memory, and Cognition, (): –. doi: ./a Gajewski, Jon R. (). Neg-raising: polarity and presupposition. PhD thesis. MIT. Gajewski, Jon (). Another look at NPIs in definite descriptions: an experimental approach. In Pierre Larrivée and Chungmin Lee (eds.), Negation and Polarity: Experimental Perspectives. Language, Cognition, and Mind . New York: Springer International Publishing. –. Gajewski, Jon and Hsieh, I-ta Chris (). Comments on negative polarity items in definite description. In Luka Crnic and Uli Sauerland (eds.), The Art and Craft of Semantics: A Festschrift for Irene Heim, vol. . Cambridge, MA: MIT Working Papers in Linguistics. –. Gallistel, Charles R. (). The Organization of Learning. Cambridge, MA: Bradford Books/MIT Press. Gallistel, Charles R. and Cramer, Andrea E. (). Computations on metric maps in mammals: getting oriented and choosing a multi-destination route. Journal of Experimental Biology, : –. Gamut, L. T. F. (). Logic, Language, and Meaning. Chicago: University of Chicago. Gärdenfors, Peter (). Conceptual Spaces: The Geometry of Thought. Cambridge, MA: MIT Press. Garnham, Alan (). Mental models as representations of text. Memory and Cognition, : –. Garnham, Alan (). The other side of mental models: theories of language comprehension. In Jane Oakhill and Alan Garnham (eds.), Mental Models in Cognitive Science: Essays in Honour of Phil Johnson-Laird. Hove: Psychology Press. –. Garnham, Alan and Oakhill, Jane (). On-line resolution of anaphoric pronouns: effects of inference making and verb semantics. British Journal of Psychology, : –. Garnham, Alan, Traxler, Matthew, Oakhill, Jane, and Gernsbacher, Morton Ann (). The locus of implicit causality effects in comprehension. Journal of Memory and Language, (): –. doi: ./jmla.. Garrod, Simon, Freudenthal, Daniel, and Boyle, Elizabeth A. (). The role of different types of anaphor in the online resolution of sentences in a discourse. Journal of Memory and Language, : –. Garvey, Catherine and Caramazza, Alfonso (). Implicit causality in verbs. Linguistic Inquiry, : –. Garvey, Catherine, Caramazza, Alfonso, and Yates, Jack (). Factors influencing assignment of pronoun antecedents. Cognition, : –. Gatt, Albert, Van Gompel, Roger P. G., Krahmer, Emiel, and Van Deemter, Kees (). Does domain size impact speech onset time during reference production? In Naomi Miyake, David Peebles, and Richard C. Cooper (eds.), Proceedings of the th Annual Meeting of the Cognitive Science Society. Austin, TX: Cognitive Science Society. –. Gatt, Albert, Krahmer, Emiel, Van Deemter, Kees, and Van Gompel, Roger P. G. (). Models and empirical data for the production of referring expressions. Language, Cognition and Neuroscience, (): –. Gawron, Jean Mark (). Generalized paths. In Effi Georgala and Jonathan Howell (eds.), Proceedings of Semantics and Linguistic Theory . Ithaca, NY: CLC. Gazdar, Gerald (). Pragmatics: Implicature, Presupposition, and Logical Form. New York: Academic Press. Gelman, Susan A. (). The Essential Child: Origins of Essentialism in Everyday Thought. New York: Oxford University Press. Gelman, Susan A. (). Generics as a window onto young children’s concepts. In Francis J. Pelletier (ed.), Kinds, Things, and Stuff: The Cognitive Side of Generics and Mass Terms. New York: Oxford University Press. –. Gelman, Susan A. and Bloom, Paul (). Young children are sensitive to how an object was created when deciding what to name it. Cognition, (): –.
Gelman, Susan A., Goetz, Peggy J., Sarnecka, Barbara W., and Flukes, Jonathan (). Generic language in parent-child conversations. Language Learning and Development, : –. Gennari, Silvia and Poeppel, David (). Processing correlates of lexical semantic complexity. Cognition, (): B–B. Gennari, Silvia P., Sloman, Steven A., Malt, Barbara C., and Fitch, W. Tecumseh (). Motion events in language and cognition. Cognition, : –. doi: ./S-()- Gennari, Silvia, Meroni, Luisa, and Crain, Stephen (). Rapid relief of stress in dealing with ambiguity. In John C. Trueswell and Michael K. Tanenhaus (eds.), Approaches to Studying WorldSituated Language Use: Bridging the Language-as-Product and Language-as-Action Traditions. Cambridge, MA: MIT Press. –. Gentner, Dedre (). If a tree had a knee, where would it be? Children’s performance on simple spatial metaphors. Papers and Reports on Child Language Development, : –. Gentner, Dedre (). On relational meaning: the acquisition of verb meaning. Child Development, (): –. Gentner, Dedre (). Structure-mapping: a theoretical framework for analogy. Cognitive Science, (): –. Gentner, Dedre and Bowdle, Brian F. (). Metaphor as structure-mapping. In Raymond W. Gibbs Jr. (ed.), Cambridge Handbook of Metaphor and Thought. New York, NY: Cambridge University Press. –. Gentner, Dedre and Bowerman, Melissa (). Why some spatial semantic categories are harder to learn than others: the typological prevalence hypothesis. In Jiansheng Guo, Elena Lieven, Nancy Budwig, Susan Ervin-Tripp, Keiko Nakamura, and Seyda Ozkaliskan (eds.), Crosslinguistic Approaches to the Psychology of Language: Research in the Tradition of Dan Isaac Slobin. New York, NY: Lawrence Erlbaum. –. Gentner, Dedre and Clement, Catherine (). Evidence for relational selectivity in the interpretation of analogy and metaphor. The Psychology of Learning and Motivation: Advances in Research and Theory, : –. Gentner, Dedre and Wolff, Phillip (). Alignment in the processing of metaphor. Journal of Memory and Language, (): –. Gentner, Dedre, Bowdle, Brian F., Wolff, Phillip, and Boronat, Consuelo (). Metaphor is like analogy. In Dedre Gentner, Keith J. Holyoak, and Boicho N. Kokinov (eds.), The Analogical Mind: Perspectives from Cognitive Science. Cambridge, MA: The MIT Press. –. George, Benjamin R. (a). Predicting presupposition projection: some alternatives in the strong Kleene tradition. Ms. UCLA. Retrieved from http://semanticsarchive.net/Archive/DYYTgxN/ George, Benjamin R. (b). Presupposition repairs: a static, trivalent approach to predict projection. MA thesis. UCLA. George, Christian (). The endorsement of the premises: assumption-based or belief-based reasoning. British Journal of Psychology, : –. doi: ./j.-..tb.x George, Christian (). Reasoning from uncertain premises. Thinking and Reasoning, : –. Gergely, György and Csibra, Gergely (). Teleological reasoning in infancy: the naive theory of rational action. Trends in Cognitive Science, (): –. Gergely, György and Csibra, Gergely (). A few reasons why we don’t share Tomasello et al.’s intuitions about sharing. Behavioral and Brain Sciences, (): –. Gergely, György, Bekkering, Harold, and Király, Ildikó (). Developmental psychology: rational imitation in preverbal infants. Nature, : –. doi: ./a Gernsbacher, Morton Ann (). Mechanisms that improve referential access. Cognition, : –. Gernsbacher, Morton Ann (). Language Comprehension as Structure Building. Hillsdale, NJ: Lawrence Erlbaum. Gernsbacher, Morton Ann (). Activating knowledge of fictional characters’ emotional states. In Charles A. Weaver III, Suzanne Mannes, and Charles R. Fletcher (eds.), Discourse Comprehension: Essays in Honor of Walter Kintsch. Mahwah, NJ: Erlbaum. –.
Gernsbacher, Morton Ann and Hargreaves, David J. (). Accessing sentence participants: the advantage of first mention. Journal of Memory and Language, : –. doi: ./-X ()- Gernsbacher, Morton Ann, Keysar, Boaz, and Robertson, Rachel R. W. (). The role of suppression in metaphor interpretation. Paper presented at Meeting of the Psychonomic Society, Los Angeles, CA. Gernsbacher, Morton Ann, Keysar, Boaz, Robertson, Rachel R. W., and Werner, Necia K. (). The role of suppression and enhancement in understanding metaphors. Journal of Memory and Language, : –. Gernsbacher, Morton Ann and Pripas-Kapit, Sarah R. (). Who’s missing the point? A commentary on claims that autistic persons have a specific deficit in figurative language comprehension. Metaphor and Symbol, : –. doi: ./.. Gerstadt, Cherie L., Hong, Yoon Joo, and Diamond, Adele (). The relationship between cognition and action: performance of children ½– years old on a Stroop-like day/night test. Cognition, : –. doi: ./-()-X Geurts, Bart (). Generics. Journal of Semantics, (): –. Geurts, Bart (). Presuppositions and anaphors in attitude contexts. Linguistics and Philosophy, (): –. Geurts, Bart (). Quantifying kids. Language Acquisition, (): –. Geurts, Bart (). Entertaining alternatives: disjunctions as modals. Natural Language Semantics, (): –. doi: ./s--- Geurts, Bart (). Take ‘five’: the meaning and use of a number word. In Svetlana Vogeleer and Liliane Tasmowski (eds.), Non-Definiteness and Plurality. Amsterdam and Philadelphia, PA: John Benjamins. –. Geurts, Bart (). Quantity Implicatures. Cambridge: Cambridge University Press. Geurts, Bart and Nouwen, Rick (). At least et al.: the semantics of scalar modifiers. Language, (): –. Geurts, Bart and Pouscoulous, Nausicaa (). Embedded implicatures?!? Semantics and Pragmatics, (): –. Geurts, Bart and Rubio-Fernández, Paula (). Pragmatics and processing. Ratio, (): –. Geurts, Bart and Van der Sandt, Rob (). Interpreting focus. Theoretical Linguistics, : –. Geurts, Bart and Van Tiel, Bob (). Embedded scalars. Semantics and Pragmatics, (): –. Geurts, Bart and Van Tiel, Bob (). When ‘all the five circles’ are four: new exercises in domain restriction. Topoi, : –. Geurts, Bart, Katsos, Napoleon, Cummins, Chris, Moons, Jonas, and Noordman, Leo (). Scalar quantifiers: logic, acquisition and processing. Language and Cognitive Processes, : –. Giannakidou, Anastasia (). Polarity Sensitivity as (Non)veridical Dependency. Amsterdam, Philadelphia: John Benjamins. Giannakidou, Anastasia (). Affective dependencies. Linguistics and Philosophy, (): –. Giannakidou, Anastasia (). Only, emotive factive verbs, and the dual nature of polarity dependency. Language, , –. Giannakidou, Anastasia (). Negative and positive polarity items: variation, licensing, and compositionality. In Klaus von Heusinger, Claudia Maienborn, and Paul Portner (eds.), Semantics: An International Handbook of Natural Language Meaning, vol. . Berlin: De Gruyter. –. Giannakidou, Anastasia and Yoon, Suwon (). Scalar marking without scalar meaning: Nonscalar, nonexhaustive even-marked NPIs in Greek and Korean. Language, (): –. Gibbons, Pamela, Busch, Jon, and Bradac, James J. (). Powerful versus powerless language: consequences for persuasion, impression formation, and cognitive response. Journal of Language and Social Psychology, (): –. doi: ./X Gibbs, Raymond W, Jr. (). Contextual effects in understanding indirect requests. Discourse Processes, : –.
Gibbs, Raymond W., Jr. (). Do people always process the literal meanings of indirect requests? Journal of Experimental Psychology: Learning, Memory, and Cognition, : –. Gibbs, Raymond W., Jr. (). On the psycholinguistics of sarcasm. Journal of Experimental Psychology: General, (): –. Gibbs, Raymond W., Jr. (). Comprehending figurative referential descriptions. Journal of Experimental Psychology: Learning, Memory and Cognition, (): –. Gibbs, Raymond W., Jr. (a). Figurative thought and figurative language. In Morton Ann Gernsbacher (ed.), Handbook of Psycholinguistics. New York: Academic Press. –. Gibbs, Raymond W., Jr. (b). The Poetics of Mind: Figurative thought, Language and Understanding. Cambridge: Cambridge University Press. Gibbs, Raymond W., Jr. (). A new look at literal meaning in understanding what is said and implicated. Journal of Pragmatics, (): –. Gibbs, Raymond W., Jr. and Moise, Jessica F. (). Pragmatics in understanding what is said. Cognition, : –. Gibbs, Raymond W., Jr. and Tendahl, Markus (). Cognitive effort and effects in metaphor comprehension: relevance theory and psycholinguistics. Mind and Language, (): –. doi: ./j.-...x Gibbs, Raymond W., Jr., O’Brien, Jennifer E., and Doolittle, Shelly (). Inferring meanings that are not intended: speakers’ intentions and irony comprehension. Discourse Processes, (): –. doi: ./ Gibson, Edward and Pearlmutter, Neal J. (). Constraints on sentence comprehension. Trends in Cognitive Sciences, (): –. doi: ./S-()- Gibson, Edward, Jacobson, Pauline, Graff, Peter, Mahowald, Kyle, Fedorenko, Evelina, and Piantadosi, Steven T. (). A pragmatic account of complexity in definite antecedent-containeddeletion relative clauses. Journal of Semantics, : –. Gil, David (). Quantifier scope, linguistic variation, and natural language semantics. Linguistics and Philosophy, (): –. Gil, David (). Universal quantifiers and distributivity. In Emmon Bach, Eloise Jelinek, Angelika Kratzer, and Barbara H. Partee (eds.), Quantification in Natural Languages. Dordrecht: Kluwer Academic Publishers. –. Gildea, Patricia and Glucksberg, Sam (). On understanding metaphor: the role of context. Journal of Verbal Learning and Verbal Behavior, : –. Gillen, Kathryn (). The comprehension of doubly quantified sentences. PhD thesis. University of Durham. Gillis, Randall L. and Nilsen, Elizabeth S. (). Children’s use of information quality to establish speaker preferences. Developmental Psychology, (): –. doi: ./a Gillon, Brendan (). The readings of plural Noun Phrases in English. Linguistics and Philosophy, : –. Gilovich, Thomas and Medvec, Victoria Husted (). The temporal profile to the experience of regret. Journal of Personality and Social Psychology, : –. Ginsburg, Shiphra, Van der Vleuten, Cees, Eva, Kevin W., and Lingard, Lorelei (). Hedging to save face: a linguistic analysis of written comments on in-training evaluation reports. Advances in Health Sciences Education, : –. doi: ./s--- Ginzburg, Jonathan (). The Interactive Stance: Meaning for Conversation. Oxford: Oxford University Press. Giora, Rachel (). Understanding figurative and literal language: the graded salience hypothesis. Cognitive Linguistics, (): –. Giora, Rachel (). Literal vs. figurative language: different or equal? Journal of Pragmatics, (): –. Giora, Rachel (). On Our Mind: Salience, Context, and Figurative Language. New York: Oxford University Press.
Giora, Rachel (). Anything negatives can do affirmatives can do just as well, except for some metaphors. Journal of Pragmatics, (): –. doi: ./j.pragma... Giora, Rachel (). Is metaphor unique? In Raymond W. Gibbs Jr. (ed.), Cambridge Handbook of Metaphor and Thought. Cambridge: Cambridge University Press. –. Giora, Rachel (). Will anticipating irony facilitate it immediately? In Marta Dynel (ed.), The Pragmatics of Humour Across Discourse Domains. Amsterdam: John Benjamins. –. Giora, Rachel and Fein, Ofer (a). Irony: context and salience. Metaphor and Symbol, (): –. Giora, Rachel and Fein, Ofer (b). On understanding familiar and less-familiar figurative language. Journal of Pragmatics, : –. Giora, Rachel, Fein, Ofer, and Schwartz, Tamir (). Irony: graded salience and indirect negation. Metaphor and Symbol, (): –. Giora, Rachel, Fein, Ofer, Laadan, Dafna et al. (). Expecting irony: context versus salience-based effects. Metaphor and Symbol, (): –. Gisladottir, Rosa S., Chwilla, Dorothee J., Schriefers, Herbert, and Levinson, Stephen C. (). Speech act recognition in conversation: experimental evidence. In Naomi Miyake, David Peebles, and Richard P. Cooper (eds.), Proceedings of the th Annual Conference of the Cognitive Science Society. Austin, TX: Cognitive Science Society. –. Givón, Talmy (). Negation in language: pragmatics, function, ontology. In Peter Cole (ed.), Syntax and Semantics : Pragmatics. Academic Press, New York. –. Givón, Talmy (). Topic continuity in discourse: an introduction. In Talmy Givón (ed.), Topic Continuity in Discourse: A Quantitative Cross-Language Study. Amsterdam: John Benjamins, –. Givón, Talmy (). Functionalism and Grammar. Amsterdam: John Benjamins. Glanzberg, Michael (). Presuppositions, truth values and expressing propositions. In Gerhard Preyer and Georg Peter (eds.), Contextualism in Philosophy: Knowledge, Meaning, and Truth. Oxford: Oxford University Press. –. Gleitman, Lila R. and Newport, Elissa L. (). The invention of language by children: environmental and biological influences on the acquisition of language. In Lila R. Gleitman and Mark Liberman (eds.), Language, Vol. : An invitation to Cognitive Science, nd edn. Cambridge: MIT Press. Glenberg, Arthur M. (). What memory is for: creating meaning in the service of action. Behavioral and Brain Sciences, (): –. Glenberg, Arthur M. and Kaschak, Michael P. (). Grounding language in action. Psychonomic Bulletin and Review, : –. doi: ./BF Glenberg, Arthur M., Meyer, Marion, and Lindem, Karen (). Mental models contribute to foregrounding during text comprehension. Journal of Memory and Language, : –. doi: ./-X()- Glenberg, Arthur M., Robertson, David A., Jansen, Jennifer L., and Johnson-Glenberg, Mina C. (). Not propositions. Cognitive Systems Research, (): –. doi: ./S-()- Glucksberg, Sam (). Understanding Figurative Language: From Metaphor to Idioms. Oxford: OUP. doi: ./acprof:oso/.. Glucksberg, Sam (). The psycholinguistics of metaphor. Trends in Cognitive Sciences, (): –. doi: ./S-()- Glucksberg, Sam and Haught, Catrinel (). On the relation between metaphor and simile: when comparison fails. Mind and Language, (), –. doi: ./j.-...x Glucksberg, Sam and Keysar, Boaz (). Understanding metaphorical comparisons: beyond similarity. Psychological Review, : –. Glucksberg, Sam and Keysar, Boaz (). How metaphors work. In Andrew Ortony (ed.), Metaphor and Thought, nd edn. Cambridge: Cambridge University Press. –. doi: ./ CBO. Glucksberg, Sam, Gildea, Patricia, and Bookin, Howard B. (). On understanding nonliteral speech: can people ignore metaphors? Journal of Verbal Learning and Verbal Behavior, : –.
Glucksberg, Sam, McGlone, Matthew S., and Manfredi, Deanna (). Property attribution in metaphor comprehension. Journal of Memory and Language, : –. Godfrey, John J., Holliman, Edward C., and McDaniel, Jane (). SWITCHBOARD: Telephone speech corpus for research and development. Paper presented at the ICASSP-, IEEE International Conference on Acoustics, Speech, and Signal Processing. Goffman, Erving (). The Presentation of Self in Everyday Life. Garden City, NY: Anchor. Goffman, Erving (). Interaction Ritual: Essays on Face-to-Face Behavior. New York: Doubleday. Gold, Rinat, Faust, Miriam, and Goldstein, Abraham (). Semantic integration during metaphor comprehension in Asperger syndrome. Brain and Language, : –. Goldberg, Adele E. (). Constructions: a new theoretical approach to language. Trends in Cognitive Sciences, (): –. Goldin-Meadow, Susan, Gelman, Susan A., and Mylander, Carolyn (). Expressing generic concepts with and without a language model. Cognition, : –. Goldstein, Abraham, Arzouan, Yossi, and Faust, Miriam (). Killing a novel metaphor and reviving a dead one: ERP correlates of metaphor conventionalization. Brain and Language, (): –. Gómez-Veiga, Isabel, García-Madruga, Juan A., and Moreno-Ríos, Sergio (). Unless conditionals: new evidences from priming. Journal of Cognitive Psychology, (): –. Gonzales, Marti Hope, Pederson, Julie Haugen, Manning, Debra J., and Wetter, David W. (). Pardon my gaffe: effects of sex, status, and consequence severity on accounts. Journal of Personality and Social Psychology, (): –. doi: ./-... Goodglass, Harold, Gleason, Jean Berko, and Hyde, Mary R. (). Some dimensions of auditory language comprehension in aphasia. Journal of Speech, Language, and Hearing Research, : –. doi: ./jshr.. Goodman, Nelson (). The problem of counterfactual conditionals. Journal of Philosophy, : –. Goodman, Noah D. and Frank, Michael C. (). Pragmatic language interpretation as probabilistic inference. Trends in Cognitive Sciences, (): –. doi: ./j.tics... Goodman, Noah D. and Lassiter, Dan (). Probabilistic semantics and pragmatics: uncertainty in language and thought. In Shalom Lappin and Chris Fox (eds.), Handbook of Contemporary Semantic Theory, nd edn. New York: Wiley-Blackwell. Goodman, Noah D. and Stuhlmüller, Andreas (). Knowledge and implicature: modeling language understanding as social cognition. Topics in Cognitive Science, : –. doi: ./tops. Goodrich Smith, Whitney and Hudson Kam, Carla L. (). Pointing to ‘her’: the effect of co-speech gesture on pronoun resolution. Language and Cognition, : –. Gopnik, Alison and Meltzoff, Andrew N. (). Words, Thoughts, and Theories, Vol. . Cambridge, MA: MIT Press. Gordon, Peter (). The truth value judgment task. In Dana McDaniel, Cecile McKee, and Helen Smith Cairns (eds.), Methods for Assessing Children’s Syntax. Cambridge, MA: MIT Press. –. Gordon, Peter (). The origin of argument structure in infant event representations. In Alejna Brugos, Linnea Micciulla, and Christine E. Smith (eds.), Proceedings of the th Annual Boston University Conference on Language Development. Somerville, MA: Cascadilla Press. –. Gordon, Peter C. and Hendrick, Randall (). The representation and processing of coreference in discourse. Cognitive Science, (): –. doi: ./S-()- Gordon, Peter C. and Scearce, Kimberly A. (). Pronominalization and discourse coherence, discourse structure and pronoun interpretation. Memory and Cognition, (): –. doi: ./BF Gordon, Peter C., Grosz, Barbara J., and Gilliom, Laura A. (). Pronouns, names, and the centering of attention in discourse. Cognitive Science, : –. Gorman, Kristen S., Gegg-Harrison, Whitney, Marsh, Chelsea R., and Tanenhaus, Michael K. (). What’s learned together stays together: speakers’ choice of referring expression reflects shared experience. Journal of Experimental Psychology: Learning, Memory, and Cognition, (): –. doi: ./a
Goswami, Usha (). Analogical reasoning: what develops? A review of research and theory. Child Development, : –. Goswami, Usha (). Analogical reasoning in children. In Dedre Gentner, Keith J. Holyoak, and Boicho N. Kokinov (eds.), The Analogical Mind: Perspectives from Cognitive Science. Cambridge, MA: The MIT Press. –. Gotzner, Nicole and Spalek, Katharina (). Exhaustive inferences and additive presuppositions: interplay of focus operators and contrastive intonation. In Judith Degen, Michael Franke and Noah D. Goodman (eds.), Proceedings of the Formal and Experimental Pragmatics Workshop. Tübingen: ESSLLI. Gotzner, Nicole and Spalek, Katharina (). The role of contrastive and non-contrastive associates in the interpretation of focus particles. Discourse Processes, (): –. doi: ./ X... Gough, Philip B. (). Grammatical transformations and speed of understanding. Journal of Verbal Learning and Verbal Behavior, (): –. doi: ./S-()- Gough, Philip B. (). The verification of sentences: the effects of delay of evidence and sentence length. Journal of Verbal Learning and Verbal Behavior, (): –. doi: ./S-()- Gouvea, Ana C., Phillips, Colin, Kazanina, Nina, and Poeppel, David (). The linguistic processes underlying the P. Language and Cognitive Processes, (): –. Graesser, Arthur C., Millis, Keith K., and Zwaan, Rolf A. (). Discourse comprehension. Annual Review of Psychology, : –. doi: ./annurev.psych... Graf, Caroline, Degen, Judith, Hawkins, Robert X. D., and Goodman, Noah D. (). Animal, dog, or dalmatian? Level of abstraction in nominal referring expressions. In Anna Papafragou, Daniel Grodner, Daniel Mirman, and John C. Trueswell (eds.), Proceedings of the th Annual Conference of the Cognitive Science Society. Austin, TX: Cognitive Science Society. –. Graf, Eileen and Davies, Catherine (). The production and comprehension of referring expressions. In Danielle Matthews (ed.), Pragmatic Development in First Language Acquisition. Amsterdam: John Benjamins. –. Grant, Cathy, Riggs, Kevin J., and Boucher, Jill (). Counterfactual and mental state reasoning in children with autism. Journal of Autism and Developmental Disorders, : –. doi: ./B: JADD... Grassmann, Susanne, Stracke, Marén, and Tomasello, Michael (). Two-year-olds exclude novel objects as potential referents of novel words based on pragmatics. Cognition, : –. Greenberg, Yael (). Exceptions to generics: where vagueness, context dependence and modality interact. Journal of Semantics, (): –. Greene, Steven B. and McKoon, Gail (). Telling something we can’t know: experimental approaches to verbs exhibiting implicit causality. Psychological Science, (): –. doi: ./j.-..tb.x Greene, Steven B., McKoon, Gail, and Ratcliff, Roger (). Pronoun resolution and discourse models. Journal of Experimental Psychology: Learning, Memory and Cognition, : –. doi: ./-... Grice, H. Paul (). Meaning. Philosophical Review, (): –. Grice, H. Paul (). Logic and Conversation: The William James Lectures. Cambridge, MA: Harvard University. Published as Part of Grice (). Grice, H. Paul (). Utterer’s meaning and intentions. Philosophical Review, (), –. Grice, H. Paul (). Logic and conversation. In Peter Cole and Jerry L. Morgan (eds.), Syntax and Semantics, Vol. : Speech Acts. New York: Academic Press. –. Reprinted in Grice (), –. Grice, H. Paul (). Further notes on logic and conversation. In Peter Cole (ed.), Syntax and Semantics : Pragmatics. New York: Academic Press. –. Grice, H. Paul (). Studies in the Way of Words. Cambridge, MA: Harvard University Press. Grice, Martine and Baumann, Stefan (). An introduction to intonation—functions and models. In Jürgen Trouvain and Ulrike Gut (eds.), Non-Native Prosody: Phonetic Description and Teaching Practice. Berlin: De Gruyter. –.
Grigoroglou, Myrto, Johanson, Megan, and Papafragou, Anna (in press). Pragmatic factors in the acquisition of spatial language: The case of front and back. Developmental Psychology. doi: ./ dev. Grimm, Hannelore (). On the child’s acquisition of semantic structure underlying the wordfield of prepositions. Language and Speech, (): –. doi: ./ Grimm, Scott (). Plurality is distinct from number-neutrality. In Yelena Fainleib, Nicholas LaCara, and Yangsook Park (eds.), Proceedings of the st Annual Meeting of the North East Linguistic Society. Amherst, MA: University of Massachussetts at Amherst. –. Grodner, Daniel and Sedivy, Julie C. (). The effects of speaker-specific information on pragmatic inferences. In Edward Gibson and Neal J. Perlmutter (eds.), The Processing and Acquisition of Reference. Cambridge, MA: MIT Press. Grodner, Daniel J., Klein, Natalie M., Carbary, Kathleen M., and Tanenhaus, Michael K. (). ‘Some,’ and possibly all, scalar inferences are not delayed: evidence for immediate pragmatic enrichment. Cognition, (): –. doi: ./j.cognition... Groenendijk, Jeroen and Stokhof, Martin (). Studies in the semantics of questions and the pragmatics of answers. PhD Thesis. University of Amsterdam. Grosjean, François and Hirt, Cendrine (). Using prosody to predict the end of sentences in English and French: normal and brain-damaged subjects. Language and Cognitive Processes, (): –. Grosse, Gerlind and Tomasello, Michael (). Two-year-old children differentiate test questions from genuine questions. Journal of Child Language, (): –. Grosse, Gerlind, Moll, Henrike, and Tomasello, Michael (). -month-olds understand the cooperative logic of requests. Journal of Pragmatics, (): –. Grosz, Barbara J. (). The representation and use of focus in dialogue understanding. Technical Report No. , Artificial Intelligence Center, SRI International, Menlo Park, CA. Grosz, Barbara J. and Sidner, Candace L. (). Attention, intentions, and the structure of discourse. Computational Linguistics, (): –. Grosz, Barbara J., Joshi, Aravind K., and Weinstein, Scott (). Centering: a framework for modeling the local discourse. Computational Linguistics, (): –. Grüter, Theres, Rohde, Hannah, and Schafer, Amy (). Coreference and discourse coherence in L: the roles of grammatical aspect and referential form. Linguistic Approaches to Bilingualism, : –. Gualmini, Andrea (). Some knowledge children don’t lack. Linguistics, : –. Gualmini, Andrea and Schwarz, Bernhard (). Solving the learnability problems in the acquisition of semantics. Journal of Semantics, : –. Gualmini, Andrea, Crain, Stephen, Meroni, Luisa, Chierchia, Gennaro, and Guasti, Maria Teresa (). At the semantics/pragmatics interface in child language. In Rachel Hastings, Brendan Jackson, and Zsofia Zvolenszky (eds.), Proceedings of Semantics and Linguistic Theory . Cornell University, Ithaca, NY: CLC Publications. –. Gualmini, Andrea, Hulsey, Sarah, Hacquard, Valentine, and Fox, Danny (). The QuestionAnswer Requirement for scope assignment. Natural Language Semantics, : –. Guasti, Maria Teresa, Chierchia, Gennaro, Crain, Stephen, Foppolo F., Gualmini, Andrea, and Meroni, Luisa (). Why children and adults sometimes (but not always) compute implicatures. Language and Cognitive Processes, : –. Gundel, Jeanette K., Hedberg, Nancy, and Zacharski, Ron (). Cognitive status and the form of referring expressions in discourse. Language, (): –. Gunlogson, Christine (). Declarative questions. In Brendan Jackson (ed.), Proceedings of Semantics and Linguistic Theory . Ithaca, NY: CLC Publications. –. Gussenhoven, Carlos (). Testing the reality of focus domains. Language and Speech, : –. Gussenhoven, Carlos (). On the limits of focus projection in English. In Peter Bosch and Rob van der Sandt (eds.), Focus: Linguistic, Cognitive, and Computational Perspectives. Cambridge: Cambridge University Press. –.
Gussenhoven, Carlos (). Phonology of intonation. Glot International, : –. Gweon, Hyowon, Pelton, Hannah, Konopka, Jaclyn A., and Schulz, Laura E. (). Sins of omission: children selectively explore when teachers are under-informative. Cognition, (): –. doi: ./j.cognition... Hackl, Martin (). On the grammar and processing of proportional quantifiers: most versus more than half. Natural Language Semantics, : –. Hackl, Martin, Koster-Hale, Jorie, and Varvoutis, Jason (). Quantification and ACD: evidence from real-time sentence processing. Journal of Semantics, : –. Hacohen, Aviya, Kozlowski, Dana, and Cohen, Ariel (). The truth shall make you slow: superlative quantifiers as illocutionary operators. Talk at XPRAG , Barcelona. Hafri, Alon, Papafragou, Anna, and Trueswell, John C. (). Getting the gist of events: recognition of two-participant actions from brief displays. Journal of Experimental Psychology: General, (): –. Hagoort, Peter and Brown, Colin M. (). Brain responses to lexical ambiguity resolution and parsing. In Charles Clifton Jr., Lyn Frazier, and Keith Rayner (eds.), Perspectives on Sentence Processing. Hillsdale, NJ: Laurence Erlbaum. –. Hagoort, Peter, Brown, Colin M., and Groothusen, Jolanda (). The syntactic positive shift (SPS) as an ERP measure of syntactic processing. Language and Cognitive Processes, (): –. Hagoort, Peter, Hald, Lea, Bastiaansen, Marcel C. M., and Petersson, Karl Magnus (). Integration of word meaning and world knowledge in language comprehension. Science, (): –. Haigh, Matthew and Stewart, Andrew J. (). The influence of clause order, congruency, and probability on the processing of conditionals. Thinking and Reasoning, (): –. doi: ./.. Hájek, Petr (). Metamathematics of Fuzzy Logic. Dordrecht: Kluwer. Hale, John (). A probabilistic early parser as a psycholinguistic model. Proceedings of the Second Meeting of the North American Chapter of the Association for Computational Linguistics. Stroudsburg, PA: Association for Computational Linguistics. Halliday, Michael A. K. (). Notes on transitivity and theme in English (Part ). Journal of Linguistics, : –. Halliday, Michael A. K. (). An Introduction to Functional Grammar. London and Baltimore, MD: Edward Arnold. Halvorsen, Per-Kristian (). The syntax and semantics of cleft constructions. PhD thesis. University of Texas at Austin. Hamm, Robert M. (). Selection of verbal probabilities: a solution for some problems of verbal probability expression. Organizational Behavior and Human Decision Processes, : –. doi: ./-()-I Hampton, James A. (). Testing the prototype theory of concepts. Journal of Memory and Language, (): –. Hampton, James A. (). Typicality, graded membership, and vagueness. Cognitive Science, : –. Hankamer, Jorge and Sag, Ivan (). Deep and surface anaphora. Linguistic Inquiry, : –. Hanna, Joy E. and Brennan, Susan E. (). Speakers’ eye gaze disambiguates referring expressions early during face-to-face conversation. Journal of Memory and Language, : –. Hanna, Joy E., Tanenhaus, Michael K., and Trueswell, John C. (). The effects of common ground and perspective on domains of referential interpretation. Journal of Memory and Language, (): –. doi: ./S-X()- Hansen, May-Britt Mosegaard and Waltereit, Richard (). On the role of generalized conversational implicature in semantic change. Proceedings from the Annual Meeting of the Chicago Linguistic Society, (): –. Hansen, Nat and Chemla, Emmanuel (). Color adjectives, standards and thresholds: an experimental investigation. Linguistics and Philosophy, : –. Hanssen, Judith, Peters, Jörg, and Gussenhoven, Carlos (). Prosodic effects of focus in Dutch declaratives. Proceedings of Speech Prosody . Campiñas: Speech Prosody. –.
Happé, Francesca G. (). Communicative competence and theory of mind in autism: a test of relevance theory. Cognition, (): –. Harris, Adam J. L. and Corner, Adam (). Communicating environmental risks: clarifying the severity effect in interpretations of verbal probability expressions. Journal of Experimental Psychology: Learning, Memory, and Cognition, : –. doi: ./a Harris, Adam J. L., Corner, Adam J., and Hahn, Ulrike (). Estimating the probability of negative events. Cognition, : –. doi: ./j.cognition... Harris, Jess, Theobald, Maryanne, Danby, Susan J., Reynolds, Edward, and Rintel, Sean (). ‘What’s going on here?’ The pedagogy of a data analysis session. In Alison Lee and Susan J. Danby (eds.), Reshaping Doctoral Education: International Approaches and Pedagogies. London and New York: Routledge. –. Harris, Lauren Julius (). Discrimination of left and right, and development of the logic of relations. Merrill-Palmer Quarterly of Behavior and Development, (): –. Harris, Lauren Julius and Strommen, Ellen A. (). The development of understanding of the spatial terms front and back. Advances in Child Development and Behavior, : –. doi: ./S-()- Harris, Paul L. (). Desires, beliefs and language. In Peter Carruthers and Peter K. Smith (eds.), Theories of Theories of Mind. Cambridge: Cambridge University Press. –. Harris, Paul L. (). Acquiring the art of conversation: children’s developing conception of their conversation partner. In Mark Bennett (ed.), Developmental Psychology: Achievements and Prospects. Philadelphia, PA: Psychology Press. –. Harris, Paul L., Ronfard, Samuel, and Bartz, Deborah (). Young children’s developing conception of knowledge and ignorance: work in progress. European Journal of Developmental Psychology, (): –. doi: ./.. Harris, Paul L., Yang, Bei, and Cui, Yixin (). ‘I don’t know’: children’s early talk about knowledge. Mind and Language, (): –. Hartshorne, Joshua K. and Snedeker, Jesse (). Verb argument structure predicts implicit causality: the advantages of finer-grained semantics. Language and Cognitive Processes, : –. Hartshorne, Joshua K., Nappa, Rebecca, and Snedeker, Jesse (). Development of the first-mention bias. Journal of Child Language, (): –. doi: ./S Hartshorne, Joshua K., Snedeker, Jesse, Azar, Stephanie Yen-Mun Liem, and Kim, Albert E. (). The neural computation of scalar implicature. Language, Cognition and Neuroscience, (): –. Haslanger, Sally (). Ideology, generics, and common ground. In Charlotte Witt (eds.), Feminist Metaphysics: Explorations in the Ontology of Sex, Gender, and the Self. Dordrecht: Springer. –. Hasson, Uri and Glucksberg, Sam (). Does understanding negation entail affirmation? An examination of negated metaphors. Journal of Pragmatics, (): –. doi: ./j.pragma... Haugh, Michael (). The discursive challenge to politeness research: an interactional alternative. Journal of Politeness Research, (): –. Haun, Daniel B. M., Rapold, Christian J., Call, Josep, Janzen, Gabriele, and Levinson, Stephen C. (). Cognitive cladistics and cultural override in Hominid spatial cognition. Proceedings of the National Academy of Sciences of the United States of America, (): –. doi: ./ pnas. Haun, Daniel B. M., Rapold, Christian J., Janzen, Gabriele, and Levinson, Stephen C. (). Plasticity of human spatial memory: spatial language and cognition covary across cultures. Cognition, : –. doi: ./j.cognition... Haviland, Susan E. and Clark, Herbert H. (). What’s new? Acquiring new information as a process in comprehension. Journal of Verbal Learning and Verbal Behavior, (): –. doi: ./S-()- Hawkins, J., Pea, Roy D., Glick, J., and Scribner, S. (). ‘Merds that laugh don’t like mushrooms’: evidence for deductive reasoning by preschoolers. Developmental Psychology, (): –. doi: ./-...
Hawkins, Robert X. D. and Goodman, Noah D. (). Conversational expectations account for apparent limits on theory of mind use. In Anna Papafragou, Daniel Grodner, Daniel Mirman, and John C. Trueswell (eds.), Proceedings of the th Annual Conference of the Cognitive Science Society. Austin, TX: Cognitive Science Society. Healey, Patrick G. T., Purver, Matthew, and Howes, Christine (). Divergence in dialogue. PLoS ONE, (): e. Hedberg, Nancy (). Discourse pragmatics and cleft sentences in English. PhD thesis. University of Minnesota, Minnesota. Hedberg, Nancy (). On the referential status of clefts. Language, : –. Hedberg, Nancy (). Multiple focus and cleft sentences. In Katharina Hartmann and Tonjes Veenstra (eds.), Cleft Structures. Amsterdam: John Benjamins. –. Heim, Irene (a). The Semantics of Definite and Indefinite Noun Phrases. New York: Garland. Heim, Irene (b). The semantics of definite and indefinite noun phrases. PhD thesis. University of Massachusetts, Amherst. Heim, Irene (). On the projection problem for presuppositions. In Michael Barlow, Daniel Flickinger, and Michael T. Wescoat (eds.), Proceedings of the Second West Coast Conference on Formal Linguistics. Standford: Stanford University. –. Heim, Irene (). Presupposition projection. In Rob van der Sandt (ed.), Reader for the Nijmegen Workshop on Presupposition, Lexical Meaning, and Discourse Processes. University of Nijmegen. Heim, Irene (). Artikel und Definitheit. In Arnim von Stechow and Dieter Wunderlich (eds.), Semantik: Ein internationales Handbuch des zeitgenossischen Forschung. Berlin: De Gruyter. –. Heim, Irene (). Degree operators and scope. In Brendan Jackson and Tanya Matthews (eds.), Proceedings of the th Semantics and Linguistic Theory Conference. Ithaca, NY: CLC Publications. –. Heim, Irene and Kratzer, Angelika (). Semantics in Generative Grammar. Malden and Oxford: Blackwell. Heim, Irene, Lasnik, Howard, and May, Robert (). Reciprocity and plurality. Linguistic Inquiry, : –. Heldner, Mattias and Edlund, Jens (). Pauses, gaps and overlaps in conversations. Journal of Phonetics, (): –. Heller, Daphna, Grodner, Daniel, and Tanenhaus, Michael K. (). The role of perspective in identifying domains of reference. Cognition, (): –. doi: ./j.cognition... Heller, Daphna, Gorman, Kristen S., and Tanenhaus, Michael K. (). To name or to describe: shared knowledge affects referential form. Topics in Cognitive Science, : –. doi: ./ j.–...x Heller, Daphna, Parisien, Christopher, and Stevenson, Suzanne (). Perspective-taking behavior as the probabilistic weighing of multiple domains. Cognition, : –. doi: ./j.cognition... Helming, Katharina A., Strickland, Brent, and Jacob, Pierre (). Making sense of early false-belief understanding. Trends in Cognitive Sciences, (): –. Hendriks, Herman (). Compositionality and model-theoretic interpretation. PhD thesis. University of Amsterdam. Hendriks, Petra (). Cognitive modeling of individual variation in reference production and comprehension. Frontiers in Psychology, : . doi: ./fpsyg.. Hendriks, Petra, Koster, Charlotte, and Hoeks, John C. J. (). Referential choice across the lifespan: why children and elderly adults produce ambiguous pronouns. Language, Cognition and Neuroscience, (): –. doi: ./.. Herbert, Cornelia and Kübler, Andrea (). Dogs cannot bark: event-related brain responses to true and false negated statements as indicators of higher-order conscious processing. PLoS ONE, (): e. doi: ./journal.pone. Hermer-Vazquez, Linda, Spelke, Elizabeth S., and Katsnelson, Alla S. (). Sources of flexibility in human cognition: dual-task studies of space and language. Cognitive Psychology, : –. doi: ./cogp..
Herrmann, Esther, Call, Josep, Hernàndez-Lloreda, María Victoria, Hare, Brian, and Tomasello, Michael (). Humans have evolved specialized skills of social cognition: the cultural intelligence hypothesis. Science, (), –. doi: ./science. Herskovits, Annette (). Semantics and pragmatics of locative expressions. Cognitive Science, (): –. doi: ./scog_ Hespos, Susan J. and Baillargeon, Renée (). Reasoning about containment events in very young infants. Cognition, (): –. doi: ./S-()- Hespos, Susan J. and Spelke, Elizabeth S. (). Conceptual precursors to language. Nature, : –. doi: ./nature Heyes, Cecilia (). Submentalizing: I am not really reading your mind. Perspectives on Psychological Science, (): –. doi: ./ Heyes, Cecilia (). When does social learning become cultural learning? Developmental Science, (). doi: ./desc. Hickmann, Maya (). The relativity of motion in first language. In Maya Hickmann and Stéphane Robert (eds.), Space in Languages: Linguistic Systems and Cognitive Categories. Amsterdam: John Benjamins. –. Higgins, Derrick and Sadock, Jerrold M. (). A machine learning approach to modelling scope preferences. Computational Linguistics, (): –. doi: ./ Hill, Beverly, Ide, Sachiko, Ikuta, Shoko, Kawasaki, Akika, and Ogino, Tsunao (). Universals of linguistic politeness: quantitative evidence from Japanese and American English. Journal of Pragmatics, (): –. Hill, Clifford A. (). Up/down, front/back, left/right: a contrastive study of Hausa and English. In Jürgen Weissenborn and Wolfgang Klein (eds.), Here and There: Cross-Linguistic Studies on Deixis and Demonstration. Amsterdam: John Benjamins. –. Hirsch, Aron and Hackl, Martin (). Incremental presupposition evaluation in disjunction. In Jyoti Iyer and Leland Kusmer (eds.), Proceedings of the th Meeting of the North East Linguistic Society, vol. . Amherst, MA: GLSA. –. Hirsch, Aron, Schwarz, Florian, and Zehr, Jérémy (). Presupposition projection from disjunction in online processing. Poster presented at Sinn und Bedeutung , Edinburgh. Hirschberg, Julia (). A theory of scalar implicature. PhD thesis. University of Pennsylvania. Hirschberg, Julia Bell (). A Theory of Scalar Implicature. New York: Garland. Hirst, Daniel and Di Cristo, Albert (). Intonation Systems: A Survey of Twenty Languages. Cambridge: Cambridge University Press. Hobbs, Jerry R. (). Coherence and coreference. Cognitive Science, : –. Hobbs, Jerry R., Stickel, Mark E., Appelt, Douglas E., and Martin, Paul (). Interpretation as abduction. Artificial Intelligence, : –. Hochstein, Lara, Bale, Alan, Fox, Danny, and Barner, David (). Ignorance and inference: do problems with Gricean epistemic reasoning explain children’s difficulty with scalar implicature? Journal of Semantics, (): –. doi: ./jos/ffu Hochstein, Lara, Bale, Alan, and Barner, David (). Scalar implicature in absence of epistemic reasoning? The case of autism spectrum disorder. Language Learning and Development. doi: ./.. Hoeks John C. J. and Brouwer, Harm (). Electrophysiological research on conversation and discourse processing. In Thomas M. Holtgraves (ed.), Oxford Handbook of Language and Social Psychology. New York, NY: Oxford University Press. –. Hoeks, John C. J., Schoot, Lotte, Taylor, Ryan C., and Brouwer, Harm (). Did you just say ‘No’ to me? An ERP study on politeness in dialogue. Poster presented at XPRAG , Utrecht. Hoeksema, Jack (). Plurality and conjunction. In Alice G. B. ter Meulen (ed.), Studies in Modeltheoretic Semantics. Dordrecht: Foris. –. Hohenstein, Jill M., Naigles, Letitia R., and Eisenberg, Ann R. (). Keeping verb acquisition in motion: a comparison of English and Spanish. In D. Geoffrey Hall and Sandra R. Waxman (eds.), Weaving a Lexicon. Cambridge, MA: MIT Press. –.
Holcomb, Phillip J. and Grainger, Jonathan (). On the time course of visual word recognition: an event-related potential investigation using masked repetition priming. Journal of Cognitive Neuroscience, (): –. Holcomb, Phillip J., Grainger, Jonathan, and O’Rourke, Timothy B. (). An electrophysiological study of the effects of orthographic neighborhood size on printed word perception. Journal of Cognitive Neuroscience, (): –. Holler, Anke and Irmen, Lisa (). Empirically assessing effects of the right frontier constraint. In António Branco (ed.), Anaphora: Analysis, Algorithms and Applications: th Discourse Anaphora and Anaphor Resolution Colloquium, Selected Papers. LNCS/LNAI . Springer. Holtgraves, Thomas M. (). Language structure in social interaction: perceptions of direct and indirect speech acts and interactants who use them. Journal of Personality and Social Psychology, (): –. Holtgraves, Thomas M. (). Politeness and memory for the wording of remarks. Memory and Cognition, : –. Holtgraves, Thomas M. (). Preference organization and reply comprehension. Discourse Processes, : –. Holtgraves, Thomas M. and Kraus, Brian (). Processing scalar implicatures in conversational contexts: an ERP study. Journal of Neurolinguistics, : –. Holtgraves, Thomas M. and Lasky, Benjamin (). Linguistic power and persuasion. Journal of Language and Social Psychology, (): –. Holtgraves, Thomas M. and McNamara, Patrick (). Parkinson’s Disease and politeness. Journal of Language and Social Psychology, (): –. Holtgraves, Thomas M. and Perdew, Audrey (). Politeness and the communication of uncertainty, Cognition, : –. Holtgraves, Thomas M. and Yang, Joong-nam (). Politeness as universal: cross-cultural perceptions of request strategies and inferences based on their use. Journal of Personality and Social Psychology, : –. doi: ./-... Holtgraves, Thomas M. and Yang, Joong-nam (). Interpersonal underpinnings of request strategies: general principles and differences due to culture and gender. Journal of Personality and Social Psychology, (): –. Holtgraves, Thomas M., Fogle, Kelly, and Marsh, Lindsay (). Pragmatic language production deficits in Parkinson’s Disease. Advances in Parkinson’s Disease, : –. doi: ./ apd.. Honda, Hidehito and Yamagishi, Kimihiko (). Directional verbal probabilities: inconsistencies between preferential judgments and numerical meanings. Experimental Psychology, : –. doi: ./-... Honda, Hidehito and Yamagishi, Kimihiko (). Perceived certainty based on verbal probability phrases: effect of directionality and its dependence on method. Japanese Psychological Research, : –. doi: ./j.–...x Horn, Laurence R. (). On the semantic properties of logical operators in English. PhD thesis. University of California, Los Angeles. Distributed by Indiana University Linguistics Club. Horn, Laurence R. (). Lexical incorporation, implicature, and the least effort hypothesis. In Donka Farkas, Wesley M. Jacobsen, and Karol W. Todrys (eds.), Proceedings of the Parasession on the Lexicon. Chicago: Chicago Linguistic Society. –. Horn, Laurence R. (). Exhaustiveness and the semantics of clefts. In Victoria A. Burke and James Pustejovsky (eds.), Proceedings of the th Annual Meeting of the North East Linguistic Society. Amhurst, MA: University of Massachussetts. –. Horn, Laurence R. (). Toward a new taxonomy for pragmatic inference: Q-based and R-based implicature. In Deborah Schiffrin (ed.), Meaning, Form, and Use in Context: Linguistic Applications. Washington, DC: Georgetown University Press. –. Horn, Laurence R. (). Metalinguistic negation and pragmatic ambiguity. Language, : –. Horn, Laurence R. (). A Natural History of Negation. Chicago: University of Chicago Press.
Horn, Laurence R. (). The said and the unsaid. In Chris Barker and David Dowty (eds.), Proceedings from the Second Conference on Semantics and Linguistic Theory. Ohio: Ohio State University Department of Linguistics. –. Horn, Laurence R. with Kleinedler, Steven R. (). Parasitic reference vs. R-based narrowing: Lexical pragmatics meets He-Man. Paper presented at annual meeting of the Linguistic Society of America, Chicago. Horn, Laurence R. (). Implicature. In Laurence R. Horn and Gregory Ward (eds.), Handbook of Pragmatics. Oxford: Blackwell. –. Horn, Laurence R. (). The border wars: a neo-Gricean perspective. In Klaus von Heusinger and Ken Turner (eds.), Where Semantics Meets Pragmatics. Leiden: Brill. –. Horn, Laurence R. (). Exhaustivity and focus revisited. Talk given at Annual Meeting of the German Linguistics Society (DGfS), Berlin. Hornby, Peter A. (). Surface structure and presupposition. Journal of Verbal Learning and Verbal Behavior, : –. doi: ./S-()- Hornstein, Norbert (). An argument for minimalism: the case of antecedent-contained deletion. Linguistic Inquiry, : –. Horowitz, Alexandra C., Schneider, Rose M., and Frank, Michael C. (in press). The trouble with quantifiers: explaining children’s deficits in scalar implicature. Child Development. Horton, William S. and Gerrig, Richard J. (). Conversational common ground and memory processes in language production. Discourse Processes, : –. Horton, William S. and Keysar, Boaz (). When do speakers take into account common ground? Cognition, (): –. doi: ./-()- Horton, William S. and Slaten, Daniel G. (). Anticipating who will say what: the influence of speaker-specific memory associations on reference resolution. Memory and Cognition, : –. Horváth, Julia (). Separating focus movement from focus. In Simin Karimi, Vida Samiian, and Wendy K. Wilkins (eds.), Phrasal and Clausal Architecture: Syntactic Derivation and Interpretation. Amsterdam: Benjamins. –. Hosman, Lawrence A. and Wright, John W. II (). The effects of hedges and hesitations on impression formation in a simulated courtroom context. Western Journal of Communication (includes Communication Reports), (): –. doi: ./ Hovland, Carl I. and Weiss, Walter (). Transmission of information concerning concepts through positive and negative instances. Journal of Experimental Psychology, (): –. doi: ./h Howard, James (). The attentional demands of negation in a memory-scanning task. Memory and Cognition, (): –. Hruska, Claudia, Steinhauer, Karsten, Alter, Kai, and Steube, Anita (). Can wrong prosodic information be mistaken by the brain? Journal of Cognitive Neuroscience, Supplement S, January, . http://hdl.handle.net//-M---AB-C Huang, Yi Ting and Snedeker, Jesse (a). On-line interpretation of scalar quantifiers: insight into the semantics-pragmatics interface. Cognitive Psychology, (): –. doi: ./j.cogpsych.... Huang, Yi Ting and Snedeker, Jesse (b). Semantic meaning and pragmatic interpretation in five-year olds: evidence from real time spoken language comprehension. Developmental Psychology, : –. Huang, Yi Ting and Snedeker, Jesse (). Logic and conversation revisited: evidence for a division between semantic and pragmatic content in real-time language comprehension. Language and Cognitive Processes, (): –. Huang, Yi Ting, Spelke, Elizabeth, and Snedeker, Jesse (). What exactly do numbers mean? Language Learning and Development, (): –. doi: ./.. Hudson-D’Zmura, Susan and Tanenhaus, Michael K. (). Assigning antecedents to ambiguous pronouns: the role of the center of attention as the default assignment. In Marilyn A. Walker, Aravind K. Joshi, and Ellen F. Prince (eds.), Centering Theory in Discourse. Oxford: Oxford University Press. –.
Hulsey, Sarah, Hacquard, Valentine, Fox, Danny, and Gualmini, Andrea (). The question answer requirement and scope assignment. In Aniko Csirmaz, Andrea Gualmini, and Andrew Nevins (eds.)., Plato’s Problems: Papers on Language Acquisition. MIT Working Papers in Linguistics . Cambridge, MA: MIT. Hunt, Lamar, Politzer-Ahles, Stephen, Gibson, Linzi, Minai, Utako, and Fiorentino, Robert (). Pragmatic inferences modulate N during sentence comprehension: evidence from picture– sentence verification. Neuroscience Letters, : –. Husband, E. Matthew (). Rescuing manner/result complementary from certain death. Proceedings of the Annual Meeting of the Chicago Linguistics Society, (): –. Husband, E. Matthew and Ferreira, Fernanda (). The role of selection in the comprehension of focus alternatives. Language, Cognition and Neuroscience, (), –. Husband, E. Matthew and Stockall, Linnaea (). Building aspectual interpretations online. In Roberto G. de Almeida and Christina Manouilidou (eds.), Cognitive Science Perspectives on Verb Representation and Processing. New York: Springer International Publishing. –. Hyde, Dominic (). From heaps and gaps to heaps of gluts. Mind, : –. Inhelder, Bärbel and Piaget, Jean (). The Growth of Logical Thinking. New York: Basic Books. Inhelder, Bärbel and Piaget, Jean (a). The Growth of Logical Thinking from Childhood to Adolescence. New York: Basic Books. Inhelder, Bärbel and Piaget, Jean (b). The Early Growth of Logic in the Child: Classification and Seriation. Oxford: Routledge. Inhoff, Albrecht Werner and Rayner, Keith (). Parafoveal word processing during eye fixations in reading: effects of word frequency. Perception and Psychophysics, : –. Ionin, Tania, Montrul, Silvina, Kim, Ji-Hye, and Philippov, Vadim (). Genericity distinctions and the interpretation of determiners in L acquisition. Language Acquisition, : –. Ionin, Tania, Montrul, Silvina, and Santos, Hélade (). An experimental investigation of the expression of genericity in English, Spanish and Brazilian Portuguese. Lingua, : –. Ionin, Tania, Montrul, Silvina, and Crivos, Mónica (). A bidirectional study on the acquisition of plural NP interpretation in English and Spanish. Applied Psycholinguistics, : –. Ioup, Georgette (). Some universals for quantifier scope. In John P. Kimball (ed.), Syntax and Semantics . New York: Academic Press. –. Ippolito, Michela (). On the meaning of some focus-sensitive particles. Natural Language Semantics, (): –. Irwin, R. John and Newland, Jacqueline K. (). Children’s knowledge of left and right: Research note. Journal of Child Psychology and Psychiatry, (): –. doi: ./j.-.. tb.x Isaacs, Ellen A. and Clark, Herbert H. (). References in conversations between experts and novices. Journal of Experimental Psychology: General, : –. Ito, Kiwako and Speer, Shari R. (). Using interactive tasks to elicit natural dialogue. In Stefan Sudhoff, Denisa Lenertová, Roland Meyer, Sandra Pappert, Petra Augurzky, Ina Mleinek, Nicole Richter, and Johannes Schließer (eds.), Methods in Empirical Prosody Research. Berlin, New York: De Gruyter. –. Ito, Kiwako and Speer, Shari R. (). Anticipatory effects of intonation: eye movements during instructed visual search. Journal of Memory and Language, : –. Ito, Kiwako, Speer, Shari R., and Beckman, Mary E. (). Informational status and pitch accent distribution in spontaneous dialogues in English. In Bernard Bel and Isabelle Marlien (eds.), Proceedings of Speech Prosody . Nara: Speech Prosody. –. Itzhak, Inbal and Baum, Shari R. (). Misleading bias-driven expectations in referential processing and the facilitative role of contrastive accent. Journal of Psycholinguistic Research, : –. Ivanko, Stacey L. and Pexman, Penny M. (). Context incongruity and irony processing. Discourse Processes, (): –. doi: ./SDP_ Ivlieva, Natalia (). Scalar implicatures and the grammar of plurality and disjunction. PhD thesis. Massachusetts Institute of Technology.
Jackendoff, Ray (). Semantic Interpretation in Generative Grammar. Cambridge, MA: MIT Press. Jackendoff, Ray (). Toward an explanatory semantic representation. Linguistic Inquiry, (): –. Jackendoff, Ray (). Semantic Structures. Cambridge/London: MIT Press. Jackendoff, Ray (). The Architecture of the Language Faculty. Cambridge, MA: MIT Press. Jacobs, Joachim (). Fokus und Skalen. Tübingen: Niemeyer. Jaffé, Joseph and Feldstein, Stanley (). Rhythms of Dialogue. New York: Academic Press. Jäger, Gerhard (). Game-theoretical pragmatics. In Johan F. van Benthem and Alice ter Meulen (eds.), Handbook of Logic and Language, nd edn. London: Elsevier. –. Jansen, Carel J. M. and Pollmann, Mathijs M. W. (). On round numbers: pragmatic aspects of numerical expressions. Journal of Quantitative Linguistics, : –. Jansen, Stefanie, Wesselmeier, Hendrik, de Ruiter, Jan P., and Müller, Horst M. (). Using the readiness potential of button-press and verbal response within spoken language processing. Journal of Neuroscience Methods, : –. Janus, Raizi A. and Bever, Thomas G. (). Processing of metaphoric language: an investigation of the three-stage model of metaphor comprehension. Journal of Psycholinguistic Research, (): –. doi: ./BF Järvikivi, Juhani, Van Gompel, Roger P. G., Hyönä, Jukka, and Bertram, Raymond (). Ambiguous pronoun resolution: contrasting the first-mention and subject-preference accounts. Psychological Science, (): –. doi: ./j.-...x Jaśkowski, Stanislaw (). Propositional calculus for contradictory deductive systems. Studia Logica, : –. Originally published in Polish in Studia Scientarium Torunensis, Sec. A II (): –. Jayez, Jacques (). Presupposition triggers and orthogonality. Ms. Lyon. Jayez, Jacques, Mongelli, Valeria, Reboul, Anne, and Van der Henst, Jean-Baptiste (). Weak and strong triggers. In Florian Schwarz (ed.), Experimental Perspectives on Presuppositions. Cham: Springer International Publishing. –. Jefferson, Gail (). Two Explorations of the Organization of Overlapping Talk in Conversation. Tilburg Papers in Language and Literature . Tilburg University, Department of Language and Literature. Jefferson, Gail (). Notes on some orderlinesses of overlap onset. In Valentina D’Urso and Paolo Leonardi (eds.), Discourse Analysis and Natural Rhetoric. Padua: Cleup. –. Jefferson, Gail (). A sketch of some orderly aspects of overlap in natural conversation. In Gene H Lerner (ed.), Conversation Analysis: Studies from the First Generation. Amsterdam: John Benjamins. –. Jenkins, Adrianna C. and Mitchell, Jason P. (). Mentalizing under uncertainty: dissociated neural responses to ambiguous and unambiguous mental state inferences. Cerebral Cortex, (): –. doi: ./cercor/bhp Jenkins, Sarah, Harris, Adam J. L., and Lark, Richard Murray (). ‘Unlikely’ outcomes might never occur, but what about ‘unlikely (% chance)’ outcomes? Paper presented at the Annual Meeting of the Cognitive Science Society, Philadelphia, PA. Jeong, Sunwoo and Potts, Christopher (). Intonational sentence-type conventions for perlocutionary effects: an experimental investigation. In Mary Moroney, Carol-Rose Little, Jacob Collard, and Dan Burgdorf (eds.), Proceedings of the th Semantics and Linguistic Theory Conference. Ithaca, NY: CLC Publications. –. Jiang, Xiaoming, Li, Yi, and Zhou, Xiaolin (). Is it over-respectful or disrespectful? Differential patterns of brain activity in perceiving pragmatic violation of social status information during utterance comprehension. Neuropsychologia, (): –. doi: ./j.neuropsychologia... Jiang, Xiaoming, Li, Yi, and Zhou, Xiaolin (). ERPs reveal the real time processing of disrespectful reference in Mandarin utterance comprehension. In Marina Terkourafi (ed.), Interdisciplinary Perspectives on (Im)politeness. Amsterdam: John Benjamins. –.
Johanson, Megan and Papafragou, Anna (). Universality and language-specificity in the acquisition of path vocabulary. In Katie Franich, Kate M. Iserman, and Lauren L. Keil (eds.), Proceedings from the th Annual Boston University Conference on Language Development. Somerville, MA: Cascadilla Press. –. Johnson, Carl Nils and Maratsos, Michael P. (). Early comprehension of mental verbs: think and know. Child Development, (): –. doi: ./ Johnson, Susan, Slaughter, Virginia, and Carey, Susan (). Whose gaze will infants follow? The elicitation of gaze-following in -month-olds. Developmental Science, (): –. Johnson-Laird, Philip N. (). On understanding logically complex sentences. Quarterly Journal of Experimental Psychology, : –. Johnson-Laird, Philip N. (). Mental Models: Toward a Cognitive Science of Language, Inference and Consciousness. Cambridge, MA: Harvard University Press. Johnson-Laird, Philip N. and Byrne, Ruth M. J. (). Deduction. London: Erlbaum. Johnson-Laird, Philip N. and Byrne, Ruth M. J. (). Conditionals: a theory of meaning, pragmatics, and inference. Psychological Review, : –. Johnson-Laird, Philip N., Oakhill, Jane, and Bull, Deborah (). Children’s syllogistic reasoning. Quarterly Journal of Experimental Psychology, (): –. Johnson-Laird, Philip N., Byrne, Ruth M. J., and Tabossi, Patrizia (). Reasoning by model: the case of multiple quantification. Psychological Review, : –. Johnston, Judith R. (). Acquisition of locative meanings: behind and in front of. Journal of Child Language, : –. doi: ./S Johnston, Judith R. and Slobin, Dan I. (). The development of locative expressions in English, Italian, Serbo-Croatian and Turkish. Journal of Child Language, (): –. doi: ./ SX Jönsson, Martin L. and Hampton, James A. (). The inverse conjunction fallacy. Journal of Memory and Language, : –. Jorgensen, Julia, Miller, George A., and Sperber, Dan (). Test of the mention theory of irony. Journal of Experimental Psychology: General, (): –. Juanchich, Marie and Sirota, Miroslav (). Do people really say it is ‘likely’ when they believe it is only ‘possible’? Effect of politeness on risk communication. Quarterly Journal of Experimental Psychology, : –. Juanchich, Marie and Sirota, Miroslav (). How to improve people’s interpretation of probabilities of precipitation. Journal of Risk Research, : –. doi: ./.. Juanchich, Marie and Sirota, Miroslav (). Improving climate change risk communication: the effect of format on prediction calibration, probability perception and decision making. Journal of Experimental Psychology: Applied. doi:./xap. Juanchich, Marie and Villejoubert, Gaëlle (). How do speakers choose uncertainty phrases to express guilt probabilities? Paper presented at the st Annual Conference of the Cognitive Science Society. Juanchich, Marie, Teigen, Karl H., and Villejoubert, Gaëlle (). Is guilt ‘likely’ or ‘not certain’? Contrast with previous probabilities determines choice of verbal terms. Acta Psychologica, : –. doi: ./j.actpsy... Juanchich, Marie, Sirota, Miroslav, and Butler, Christina Lea (). The perceived functions of linguistic risk quantifiers and their effect on risk, negativity perception and decision making. Organizational Behavior and Human Decision Processes, : –. doi: ./j.obhdp... Juanchich, Marie, Sirota, Miroslav, Karelitz, Tzur M., and Villejoubert, Gaëlle (). Can membership-functions capture the directionality of verbal probabilities? Thinking and Reasoning, : –. doi: ./.. Juanchich, Marie, Teigen, Karl H., and Gourdon, Amélie (). Top scores are possible, bottom scores are certain (and middle scores are not worth mentioning): a pragmatic view of verbal probabilities. Judgment and Decision Making, : –.
Jucks, Regina, Päuler, Lena, and Brummernhenrich, Benjamin (). ‘I need to be explicit: you’re wrong’: impact of face threats on social evaluations in online instructional communication. Interacting with Computers, (): –. doi: ./iwc/iwu Jun, Sun-Ah (). The phonetics and phonology of Korean prosody. PhD thesis. Ohio State University. Jun, Sun-Ah (). K-ToBI (Korean ToBI) labelling conventions (ver. .). UCLA Working Papers in Phonetics, : –. Jun, Sun-Ah (ed.) (). Prosodic Typology: The Phonology of Intonation and Phrasing. Oxford: Oxford University Press. Jun, Sun-Ah (). Prosodic typology: by prominence type, word prosody, and macro-rhythm. In Sun-Ah Jun (ed.), Prosodic Typology II. Oxford: Oxford University Press. Jun, Sun-Ah and Kim, Hee-Sun (). VP focus and narrow focus in Korean. In William J. Barry and Jürgen Trouvain (eds.), Proceedings of the XVIth International Congress of Phonetic Sciences. Saarbrücken: Universität des Saarlandes. –. Jun, Sun-Ah and Lee, Hyuck-Joon (). Phonetic and phonological markers of contrastive focus in Korean. In Robert H. Mannell and Jordi Robert-Ribes (eds.), Proceedings of the th International Conference on Spoken Language Processing. Sydney: ICSLP. –. Jurafsky, Daniel (). A probabilistic model of lexical and syntactic access and disambiguation. Cognitive Science, (): –. doi: ./scog_ Kaan, Edith, Harris, Anthony, Gibson, Edward, and Holcomb, Phillip (). The P as an index of syntactic integration difficulty. Language and Cognitive Processes, (): –. Kadmon, Nirit and Landman, Fred (). Any. Linguistics and Philosophy, (): –. Kahn, Jason M. and Arnold, Jennifer E. (). A processing-centered look at the contribution of givenness to durational reduction. Journal of Memory and Language, (): –. doi: ./j.jml... Kahneman, Daniel (). Varieties of counterfactual thinking. In Neal J. Roese and James M. Olson (eds.), What Might Have Been: The Social Psychology of Counterfactual Thinking. Hillsdale, NJ: Erlbaum. –. Kahneman, Daniel (). Thinking, Fast and Slow. New York: Farrar, Straus and Giroux. Kahneman, Daniel and Frederick, Shane (). Representativeness revisited: attribute substitution in intuitive judgement. In Thomas Gilovich, Dale Griffin, and Daniel Kahneman (eds.), Heuristics and Biases: The Psychology of Intuitive Judgement. New York: Cambridge University Press. –. Kahneman, Daniel and Miller, Dale T. (). Norm theory: comparing reality to its alternatives. Psychological Review, (): –. Kahneman, Daniel and Tversky, Amos (). The simulation heuristic. In Daniel Kahneman, Paul Slovic, and Amos Tversky (eds.), Judgement Under Uncertainty: Heuristics and Biases. New York: Cambridge University Press. –. Kaiser, Elsi (). Investigating the consequences of focus on the production and comprehension of referring expressions. International Review of Pragmatics, : –. Kaiser, Elsi, Runner, Jeffrey T., Sussman, Rachel S., and Tanenhaus, Michael K. (). Structural and semantic constraints on the resolution of pronouns and reflexives. Cognition, : –. doi: ./j.cognition.... Kalbe, Elke, Schlegel, Marius, Sack, Alexander T. et al. (). Dissociating cognitive from affective theory of mind: a TMS study. Cortex, (): –. Kalmus, Hans (). Dependence of colour naming and monochromator setting on the direction of preceding changes in wavelength. British Journal of Physiological Optics, : –. Kameyama, Megumi (). Indefeasible semantics and defeasible pragmatics. In Makoto Kanazawa, Christopher Piñon, and Henriëtte De Swart (eds.), Quantifiers, Deduction, and Context. Stanford, CA: CSLI. Kamide, Yuki, Altmann, Gerry T. M., and Haywood, Sarah L. (). The time-course of prediction in incremental sentence processing: evidence from anticipatory eye movements. Journal of Memory and Language, (): –.
Kamp, Hans (). Two theories about adjectives. In Edward L. Keenan (ed.), Formal Semantics of Natural Language. Cambridge: Cambridge University Press. –. Kamp, Hans (a). A theory of truth and semantic representation. In Jeroen A. G. Groenendijk, Theo M. V. Janssen, and Martin J. B. Stokhof (eds.), Formal Methods in the Study of Language. Amsterdam: Mathematical Centre Tracts . –. Kamp, Hans (b). The paradox of the heap. In Uwe Mönnich (ed.), Aspects of Philosophical Logic. Dordrecht: Riedel. –. Kamp, Hans (). Presupposition computation and presupposition justification. In Myriam Bras and Laure Vieu (eds.), Semantic and Pragmatic Issues in Discourse and Dialogue. Amsterdam: Elsevier. –. Kamp, Hans and Partee, Barbara (). Prototype theory and compositionality. Cognition, : –. Kamp, Hans and Reyle, Uwe (). From Discourse to Logic: Introduction to Model Theoretic Semantics of Natural Language, Formal Logic and Discourse Representation Theory. Dordrecht: Kluwer. Kampe, Knut K. W., Frith, Chris D., and Frith, Uta (). ‘Hey John’: signals conveying communicative intention toward the self activate brain regions associated with ‘mentalizing,’ regardless of modality. Journal of Neuroscience, (): –. Kang, Hye-Kyung (). Quantifier spreading: linguistic and pragmatic considerations. Lingua, (): –. doi: ./S-()- Kant, Immanuel (). The Critique of Pure Reason, nd edn. First edn published . Kao, Justine T., Wu, Jean Y., Bergen, Leon, and Goodman, Noah D. (). Nonliteral understanding of number words. Proceedings of the National Academy of Sciences of the United States of America, : –. Karelitz, Tzur M. and Budescu, David V. (). You say ‘probable’ and I say ‘likely’: improving interpersonal communication with verbal probability phrases. Journal of Experimental Psychology: Applied, : –. doi: ./-X... Karmiloff-Smith, Annette (). A Functional Approach to Child Language. Cambridge: Cambridge University Press. Karmiloff-Smith, Annette (). Language and cognitive processes from a developmental perspective. Language and Cognitive Processes, : –. Karttunen, Lauri (). Presuppositions of compound sentences. Linguistic Inquiry, (): –. Karttunen, Lauri (). Presupposition and linguistic context. Theoretical Linguistics, : –. Katsos, Napoleon and Bishop, Dorothy V. M. (). Pragmatic tolerance: implications for the acquisition of informativeness and implicature. Cognition, (): –. Katsos, Napoleon, Andrés Roqueta, Clara, Estevan, Rosa Ana C., and Cummins, Chris (). Are children with Specific Language Impairment competent with the pragmatics and logic of quantification? Cognition, : –. Katsos, Napoleon, Cummins, Chris, Ezeizabarrena, Maria et al. (). Cross-linguistic patterns in the acquisition of quantifiers. Proceedings of the National Academy of Sciences of the United States of America, (): –, doi: ./pnas. Katz, Albert N. (). Discourse and sociocultural factors in understanding nonliteral language. In Herbert Colston and Albert N. Katz (eds.), Figurative Language Comprehension: Social and Cultural Influences. Mahwah: Erlbaum. –. Katz, Albert N., Paivio, Allan, Marschark, Mark, and Clark, James M. (). Norms for literary and nonliterary metaphors on psychological dimensions. Metaphor and Symbol, (): –. doi: ./sms_ Katz, Jonah and Selkirk, Elisabeth (). Contrastive focus vs. discourse-new: evidence from phonetic prominence in English. Language, : –. Katzir, Roni (). Structurally-defined alternatives. Linguistics and Philosophy, (): –. doi: ./s---y Kaup, Barbara and Zwaan, Rolf A. (). Effects of negation and situational presence on the accessibility of text information. Journal of Experimental Psychology: Learning, Memory and Cognition, : –.
Kaup, Barbara, Zwaan, Rolf A., and Lüdtke, Jana (). Effects of negation, truth value, and delay on picture recognition after reading affirmative and negative sentences. In Bruno G. Bara, Lawrence Barsalou, and Monica Bucciarelli (eds.), Proceedings of the th Annual Conference of the Cognitive Science Society. Mahwah, NJ: Lawrence Erlbaum Associates. –. Kaup, Barbara, Zwaan, Rolf A., and Lüdtke, Jana (). Processing negated sentences with contradictory predicates: is a door that is not open mentally closed? Journal of Pragmatics, : –. Kaup, Barbara, Yaxley, Richard H., Madden, Carol J., Zwaan, Rolf A., and Lüdtke, Jana (). Experiential simulations of negated text information. Quarterly Journal of Experimental Psychology, : –. Kaup, Barbara, Zwaan, Rolf A., and Lüdtke, Jana (). The experiential view of language comprehension: how is negated text information represented? In Franz Schmalhofer and Charles A. Perfetti (eds.), Higher Level Language Processes in the Brain: Inference and Comprehension Processes. Mahwah, NJ: Lawrence Erlbaum. –. Kawachi, Kazuhiro (). Korean putting verbs do not contrast space contrastively in terms of ‘tightness of fit’. Lingua, : –. doi: ./j.lingua... Kay, Paul and Regier, Terry (). Language, thought and color: recent developments. Trends in Cognitive Sciences, (): –. doi: ./j.tics... Keefe, Rosanna (). Theories of Vagueness. Cambridge: Cambridge University Press. Keefe, Rosanna and Smith, Peter (eds.) (a). Vagueness: A Reader. Cambridge: MIT Press. Keefe, Rosanna and Smith, Peter (b). Introduction: theories of vagueness. In Rosanna Keefe and Peter Smith (eds.), Vagueness: A Reader. Cambridge, MA: MIT Press. –. Keenan, Janice M., MacWhinney, Brian, and Mayhew, Deborah (). Pragmatics in memory: a study of natural conversation. Journal of Verbal Learning and Verbal Behavior, (): –. Kehler, Andrew (). Coherence and the resolution of ellipsis. Linguistics and Philosophy, (): –. Kehler, Andrew (). Coherence, Reference, and the Theory of Grammar. Stanford, CA: CSLI Publications. Kehler, Andrew and Rohde, Hannah (). A probabilistic reconciliation of coherence-driven and centering-driven theories of pronoun interpretation. Theoretical Linguistics, : –. Kehler, Andrew and Rohde, Hannah (). Evaluating an expectation-driven Question-UnderDiscussion model of discourse interpretation. Discourse Processes, (): –. doi: ./ X.. Kehler, Andrew, Kertz, Laura, Rohde, Hannah, and Elman, Jeffrey L. (). Coherence and coreference revisited. Journal of Semantics, : –. doi: ./jos/ffm Keil, Frank C. (). Conceptual domains and the acquisition of metaphor. Cognitive Development, : –. Keil, Frank C. (). Explanation, association, and the acquisition of word meaning. Lingua, : –. Keitel, Anne, Prinz, Wolfgang, Friederici, Angela D., Von Hofsten, Claes, and Daum, Moritz M. (). Perception of conversations: the importance of semantics and intonation in children’s development. Journal of Experimental Child Psychology, (): –. doi: ./j.jecp... Keller, Frank and Alexopoulou, Theodora (). Phonology competes with syntax: experimental evidence for the interaction of word order and accent placement in the realization of information structure. Cognition, : –. doi: ./S-()- Keltner, Dacher, Young, Randall C., Heerey, Erin A., Oemig, Carmen, and Monarch, Natalie D. (). Teasing in hierarchical and intimate relations. Journal of Personality and Social Psychology, (): –. doi: ./-... Kemtes, Karen A. and Kemper, Susan (). Aging and resolution of quantifier scope effects. Journal of Gerontology, : –. Kennedy, Christopher (). Antecedent-contained deletion and the syntax of quantification. Linguistic Inquiry, : –.
Kennedy, Christopher (). Vagueness and grammar: the semantics of relative and absolute gradable adjectives. Linguistics and Philosophy, : –. Kennedy, Christopher (). Two sources of subjectivity: qualitative assessment and dimensional uncertainty. Inquiry: An Interdisciplinary Journal of Philosophy, : –. Kennedy, Christopher (). A ‘de-Fregean’ semantics (and neo-Gricean pragmatics) for modified and unmodified numerals. Semantics and Pragmatics (): –. doi: ./sp.. Kennedy, Christopher and Levin, Beth (). Measure of change: the adjectival core of degree achievements. In Louise McNally and Christopher Kennedy (eds.), Adjectives and Adverbs: Syntax, Semantics and Discourse. Oxford: Oxford University Press. –. Kennedy, Christopher and McNally, Louise (). Scale structure, degree modification and the semantics of gradable predicates. Language, : –. Kennedy, Linda, Bill, Cory, Schwarz, Florian, Crain, Stephen, Folli, Raffaella, and Romoli, Jacopo (). Scalar implicatures vs presuppositions: the view from Broca’s aphasia. In Seda Kan, Claire Moore-Cantwell, and Robert Staubs (eds.), Proceedings of the th Annual Meeting of the North East Linguistic Society. Amherst, MA: GLSA. Kertz, Laura (). Verb phrase ellipsis: the view from information structure. Language, (): –. Keysar, Boaz (). On the functional equivalence of literal and metaphorical interpretations in discourse. Journal of Memory and Language, : –. Keysar, Boaz, Barr, Dale J., Balin, Jennifer A., and Brauner, Jason S. (). Taking perspective in conversation: the role of mutual knowledge in comprehension. Psychological Science, : –. doi: ./-. Khemlani, Sangeet, Leslie, Sarah-Jane, Glucksberg, Sam, and Rubio-Fernández, Paula (). Do ducks lay eggs? How humans interpret generic assertions. In Danielle McNamara and Greg Trafton (eds.), Proceedings of the th annual conference of the Cognitive Science Society. Austin, TX: Cognitive Science Society. Khemlani, Sangeet, Leslie, Sarah-Jane, and Glucksberg, Sam (). Generics, prevalence, and default inferences. In Niels A. Taatgen and Hedderik van Rijn (eds.), Proceedings of the st Annual Conference of the Cognitive Science Society. Austin, TX: Cognitive Science Society. Khemlani, Sangeet, Leslie, Sarah-Jane, and Glucksberg, Sam (). Inferences about members of kinds: the generics hypothesis. Language and Cognitive Processes, : –. Khemlani, Sangeet, Orenes, Isabel, and Johnson-Laird, Philip N. (). Negation: a theory of its meaning, representation, and use. Journal of Cognitive Psychology, (): –. doi: ./ .. Khetarpal, Naveen, Majid, Asifa, and Regier, Terry (). Spatial terms reflect near-optimal spatial categories. In Niels A. Taatgen and Hedderik van Rijn (eds.), Proceedings of the st Annual Conference of the Cognitive Science Society. Austin, TX: Cognitive Science Society. –. Khetarpal, Naveen, Neveu, Grace, Majid, Asifa, Michael, Lev, and Regier, Terry (). Spatial terms across languages support near-optimal communication: evidence from Peruvian Amazonia, and computational analyses. In Markus Knauff, Michael Pauen, Natalie Sebanz, and Ipke Wachsmuth (eds.), Proceedings of the th Annual Meeting of the Cognitive Science Society. Austin, TX: Cognitive Science Society. –. Kiguchi, Hirohisa and Thornton, Rosalind (). Binding principles and ACD constructions in child grammars. Syntax, : –. Kim, Albert and Osterhout, Lee (). The independence of combinatory semantic processing: evidence from event-related potentials. Journal of Memory and Language, (): –. Kim, Christina S. (). Processing presupposition: verifying sentences with ‘only’. Proceedings of the th Penn Linguistics Colloquium. University of Pennsylvania Working Papers in Linguistics, (). Kim, Christina S. (). Presupposition satisfaction, locality and discourse constituency. In Florian Schwarz (ed.), Experimental Perspectives on Presuppositions. Cham: Springer International Publishing. –.
Kim, Christina S., Gunlogson, Christine, Tanenhaus, Michael K., and Runner Jeffrey T. (). Context-driven expectations about focus alternatives. Cognition, : –. doi: ./j. cognition... Kim, Hee-Sun, Jun, Sun-Ah, Lee, Hyuck-Joon, and Kim, Jong-Bok (). Argument structure and focus projection in Korean. Proceedings of Speech Prosody . Dresden: Speech Prosody. Kim, Sahyang (). The role of prosodic phrasing in Korean word segmentation. PhD thesis. UCLA. Kim, So Hyun, Paul, Rhea, Tager-Flusberg, Helen, and Lord, Catherine (). Language and communication in autism. In Fred Volkmar, Rhea Paul, Sally J. Rogers, and Kevin A. Pelphrey (eds.), Handbook of Autism and Pervasive Developmental Disorders . Hoboken, NJ: Wiley. –. Kintsch, Walter (). The role of knowledge in discourse comprehension: a constructiveintegration model. Psychological Review, : –. Kiparsky, Paul (). Remarks on denominal verbs. In Alex Alsina, Joan Bresnan, and Peter Sells (eds.), Complex Predicates. Stanford, CA: CSLI Publications. –. Kirby, Simon (). Function, Selection, and Innateness: The Emergence of Language Universals. Oxford: Oxford University Press. Kiss, Katalin É. (). Identificational focus versus information focus. Language, : –. Kiss, Katalin É. and Zétényi, Tamás (). Quantifier spreading: children misled by ostensive cues. Glossa: A Journal of General Linguistics, (): . doi: ./gjgl. Kleene, Stephen (). Introduction to Metamathematics. Amsterdam: North-Holland. Klein, Devorah E. and Murphy, Gregory L. (). The representation of polysemous words. Journal of Memory and Language, (): –. Klein, Devorah E. and Murphy, Gregory L. (). Paper has been my ruin: conceptual relations of polysemous senses. Journal of Memory and Language, (): –. Klein, Ewan (). A semantics for positive and comparative adjectives. Linguistics and Philosophy, : –. Kleinschmidt, Dave F. and Jaeger, T. Florian (). Robust speech perception: recognize the familiar, generalize to the similar, and adapt to the novel. Psychological Review, (): –. doi: ./ a Klepousniotou, Ekaterini (). The processing of lexical ambiguity: homonymy and polysemy in the mental lexicon. Brain and Language, : –. Klepousniotou, Ekatarini and Baum, Shari R. (). Processing homonymy and polysemy: effects of sentential context and time-course following unilateral brain damage. Brain and Language, : –. Klepousniotou, Ekaterini, Titone, Debra, and Romero, Carolina (). Making sense of word senses: the comprehension of polysemy depends on sense overlap. Journal of Experimental Psychology: Learning, Memory, and Cognition, (): –. Klepousniotou, Ekaterini, Pike, G. Bruce, Steinhauer, Karsten, and Gracco, Vincent L. (). Not all ambiguous words are created equal: an EEG investigation of homonymy and polysemy. Brain and Language, (): –. Klima, Edward S. (). Negation in English. In Jerry A. Fodor and Jerrold J. Katz (eds.), The Structure of Language. Englewood Cliffs, NJ: Prentice-Hall. Klinedinst, Nathan (). Plurals, possibilities, and conjunctive disjunction. UCL Working Papers in Linguistics, : –. Klinedinst, Nathan (). THCSP. Ms. UCL. Knežević, Natasa (). Numerals and distributivity in Serbian: at the syntax-semantics acquisition interface. PhD thesis. University of Nantes. Knill, David C. (). Robust cue integration: a Bayesian model and evidence from cue-conflict studies with stereoscopic and figure cues to slant. Journal of Vision, (): . doi: ./.. Köbel, Max (). Faultless disagreement. Proceedings of the Aristotelean Society, : –. Koch, Karsten (). Intonation and focus in NìePkepmxcin. PhD thesis. University of British Columbia.
Koch, Peter (). Metonymy: unity in diversity. Journal of Historical Pragmatics, : –. Koenig, Jean-Pierre, Mauner, Gail, and Bienvenue, Breton (). Arguments for adjuncts. Cognition, (): –. Koenig, Melissa A. and Harris, Paul L. (). Preschoolers mistrust ignorant and inaccurate speakers. Child Development, (): –. doi: ./j.-...x Kong, Augustine, Barnett, G. Octo, Mosteller, Frederick, and Youtz, Cleo (). How medical professionals evaluate expressions of probability. New England Journal of Medicine, : –. Koolen, Ruud, Gatt, Albert, Goudbeek, Martijn and Krahmer, Emiel (). Factors causing overspecification in definite descriptions. Journal of Pragmatics, : –. Koolen, Ruud, Goudbeek, Martijn, and Krahmer, Emiel (). The effect of scene variation on the redundant use of colour in definite reference. Cognitive Science, (): –. Koolen, Ruud, Krahmer, Emiel, and Swerts, Marc (). How distractor objects trigger referential overspecification: testing the effects of visual clutter and distractor distance. Cognitive Science, : –. doi: ./cogs. Koornneef, Arnout W. and Sanders, Ted J. M. (). Establishing coherence relations in discourse: the influence of implicit causality and connectives on pronoun resolution. Language and Cognitive Processes, : –. Koornneef, Arnout W. and Van Berkum, Jos J. A. (). On the use of verb-based implicit causality in sentence comprehension: evidence from self-paced reading and eye tracking. Journal of Memory and Language, : –. Kosinski, Robert J. (). A literature review on reaction time. https://archive.li/jwW-selection.-. Kovács, Ágnes M., Téglás, Ernő, Gergely, György, and Csibra, Gergely (in press). Seeing behind the surface: communicative demonstration boosts category disambiguation in -month-olds. Developmental Science. Kövecses, Zoltán (). Metaphor: A Practical Introduction. Oxford: Oxford University Press. Kowatch, Kristy, Whalen, Juanita M., and Pexman, Penny M. (). Irony comprehension in action: a new test of processing for verbal irony. Discourse Processes, (): –. Krahmer, Emiel and Swerts, Marc (). On the alleged existence of contrastive accents. Speech Communication, : –. Kratzer, Angelika (). Modality. In Arnim von Stechow and Dieter Wunderlich (eds.), Semantik. Ein internationales Handbuch der zeitgnössischen Forschung. Berlin: De Gruyter. –. Kratzer, Angelika (). Building resultatives. In Claudia Maienborn and Angelika Wöllstein (eds.), Event Arguments: Foundations and Applications. Tübingen: Niemeyer. –. Krifka, Manfred (). An outline of genericity. Partly in collaboration with Claudia Gerstner. Technical Report SNS-Bericht –, Seminar für natürliche Systeme, Tübingen University. Krifka, Manfred (). Nominal reference, temporal constitution and quantification in event semantics. In Renate Bartsch, Johan van Benthem, and Peter van Emde Boas (eds.), Semantics and Contextual Expressions. Dordrecht: Foris. –. Krifka, Manfred (a). Thematic relations as links between nominal reference and temporal constitution. In Ivan A. Sag and Anna Szabolcsi (eds.), Lexical Matters. Stanford: CSLI Publications. –. Krifka, Manfred (b). A compositional semantics for multiple focus constructions. In Joachim Jacobs (ed.), Informationsstruktur und Grammatik. Opladen: Westdeutscher Verlag. Krifka, Manfred (a). The semantics and pragmatics of polarity items. Linguistic Analysis, : –. Krifka, Manfred (b). Common nouns: a contrastive analysis of Chinese and English. In Greg Carlson and Francis J. Pelletier (eds.), The Generic Book. Chicago: University of Chicago Press. –. Krifka, Manfred (). The origins of telicity. In Susan Rothstein (ed.), Events and Grammar. Dordrecht: Springer. –. Krifka, Manfred (a). At least some determiners aren’t determiners. In Ken Turner (ed.), The Semantics/Pragmatics Interface from Different Points of View, Vol. . Elsevier. –.
Krifka, Manfred (b). Additive particles under stress. In Devon Strolovitch and Aaron Lawson (eds.), Proceedings of Semantics and Linguistic Theory . CLC Publications. –. doi: ./ salt.vi. Krifka, Manfred (). Basic notions of information structure. In Caroline Féry, Gisbert Fanselow, and Manfred Krifka (eds.), Interdisciplinary Studies of Information Structure . Potsdam: Universitätsverlag Potsdam. –. Krifka, Manfred (). Approximate interpretations of number words: a case for strategic communication. In Erhard Hinrichs and Jon Nerbonne (eds.), Theory and Evidence in Semantics. Stanford, CA: CSLI Publications. –. Krifka, Manfred, Pelletier, Francis J., Carlson, Greg, ter Meulen, Alice, Chierchia, Gennaro, and Link, Godehard (). Genericity: an introduction. In Greg Carlson and Francis J. Pelletier (eds.), The Generic Book. Chicago: University of Chicago Press. –. Kripke, Saul (). Presupposition and anaphora: remarks on the formulation of the projection problem. Ms. Princeton. Kripke, Saul A. (). Presupposition and anaphora: remarks on the formulation of the projection problem. Linguistic Inquiry, (): –. doi: ./ling.... Križ, Manuel (). Referentiality, exhaustivity, and trivalence in it-clefts. Ms. IJN. Kronmüller, Edmundo and Barr, Dale J. (). Perspective-free pragmatics: broken precedents and the recovery-from-preemption hypothesis. Journal of Memory and Language, (): –. doi: ./j.jml... Kronmüller, Edmundo and Barr, Dale J. (). Referential precedents in spoken language comprehension: a review and meta-analysis. Journal of Memory and Language, : –. Kronmüller, Edmundo, Morisseau, Tiffany, and Noveck, Ira A. (). Show me the pragmatic contribution: a developmental investigation of contrastive inference. Journal of Child Language, : –. doi: ./S Kuczaj, Stan A. II and Maratsos, Michael P. (). On the acquisition of front, back and side. Child Development, : –. doi: ./ Kügler, Frank and Skopeteas, Stavros (). Interaction of lexical tone and information structure of Yucatec Maya. In Agnès Belotel-Grenié and Michel Grenié (eds.), Proceedings of the Second International Symposium on Tonal Aspects of Languages. La Rochelle: Université de la Rochelle. –. Kügler, Frank and Skopeteas, Stavros (). On the universality of prosodic reflexes of contrast: the case of Yucatec Maya. In William J. Barry and Jürgen Trouvain (eds.), Proceedings of the XVIth International Congress of Phonetic Sciences. Saarbrücken: Universität des Saarlandes. –. Kulakova, Eugenia and Nieuwland, Mante S. (). Understanding counterfactuality: a review of experimental evidence for the dual meaning of counterfactuals. Language and Linguistics Compass, (): –. doi: ./lnc. Kulakova, Eugenia and Nieuwland, Mante S. (in press). Pragmatic skills predict online counterfactual comprehension: evidence from the N. Cognitive, Affective, and Behavioral Neuroscience. Kulakova, Eugenia, Freunberger, Dominik, and Roehm, Dietmar (). Marking the counterfactual: ERP evidence for pragmatic processing of German subjunctives. Frontiers in Human Neuroscience, : . Kuperberg, Gina R. (). Neural mechanisms of language comprehension: challenges to syntax. Brain Research, : –. Kuperberg, Gina R. (). The pro-active comprehender: what event-related potentials tell us about the dynamics of reading comprehension. In Brett Miller, Laurie E. Cutting, and Peggy McCardle (eds.), Unraveling the Behavioral, Neurobiological, and Genetic Components of Reading Comprehension. Baltimore: Paul Brookes Publishing. –. Kuperberg, Gina R., Choi, Arim, Cohn, Neil, Paczynski, Martin, and Jackendoff, Ray (). Electrophysiological correlates of complement coercion. Journal of Cognitive Neuroscience, (): –. doi: ./jocn.. Kupisch, Tanja (). The Acquisition of Determiners in Bilingual German-Italian and GermanFrench Children. München: Lincom Europa.
Kupisch, Tanja (). Specific and generic subjects in the Italian of German-Italian simultaneous bilinguals and L learners. Bilingualism: Language and Cognition, (): –. Kupisch, Tanja and Barton, Dagmar (). How bilinguals deal with variation: generic reference in adult German bilinguals. Studia Linguistica, (): –. Kurtzman, Howard S. and MacDonald, Maryellen C. (). Resolution of quantifier scope ambiguities. Cognition, : –. Kurumada, Chigusa, Brown, Meredith, and Tanenhaus, Michael K. (). Pragmatic interpretation of contrastive prosody: it looks like speech adaptation. In Naomi Miyake, David Peebles, and Richard P. Cooper (eds.), Proceedings of the th Annual Conference of the Cognitive Science Society. Austin, TX: Cognitive Science Society. –. Kurumada, Chigusa, Brown, Meredith, Bibyk, Sarah, Pontillo, Daniel F., and Tanenhaus, Michael K. (). Is it or isn’t it: listeners make rapid use of prosody to infer speaker meanings. Cognition, : –. doi: ./j.cognition... Kurumada, Chigusa, Brown, Meredith, Bibyk, Sarah, and Tanenhaus, Michael K. (n.d.). Adaptation and inference in pragmatic interpretation of English contrastive prosody. Cognitive Psychology. Kutas, Marta and Federmeier, Kara D. (). Thirty years and counting: finding meaning in the N components of the event related brain potential (ERP). Annual Review of Psychology, : –. Kutas, Marta and Hillyard, Steven A. (a). Reading between the lines: event-related brain potentials during natural sentence processing. Brain and Language, (): –. Kutas, Marta and Hillyard, Steven A. (b). Reading senseless sentences: brain potentials reflect semantic incongruity. Science, : –. Kutas, Marta and Hillyard, Steven A. (). Brain potentials during reading reflect word expectancy and semantic association. Nature, (): –. doi: ./a Kyburg, Alice and Morreau, Michael (). Fitting words: vague language in context. Linguistics and Philosophy, : –. Ladd, D. Robert (). The Structure of Intonational Meaning: Evidence from English. Bloomington, IN: Indiana University Press. Ladd, D. Robert (). Intonational Phonology, nd edn. Cambridge: Cambridge University Press. Ladd, D. Robert and Morton, Rachel (). The perception of intonational emphasis: continuous or categorical? Journal of Phonetics, : –. Ladusaw, William A. (). Polarity sensitivity as inherent scope relations. PhD thesis. University of Texas at Austin. Ladusaw, William A. (). On the notion ‘affective’ in the analysis of Negative Polarity Items. Journal of Linguistic Research, : –. Lahiri, Utpal (). Focus and negative polarity in Hindi. Natural Language Semantics, (): –. doi: ./A: Lai, Vicky Tzuyin and Curran, Tim (). ERP evidence for conceptual mappings and comparison processes during the comprehension of conventional and novel metaphors. Brain and Language, (): –. Laka, Itziar (). On the Syntax of Negation. Outstanding Dissertations in Linguistics Series. New York/London: Garland Publishing Co. Lakoff, George (). Irregularity in Syntax. New York: Holt, Rinehart and Winston. Lakoff, George (). Presupposition and relative well-formedness. In D. D. Steinberg and L. A. Jakobovits (eds.), Semantics. Cambridge: Cambridge University Press. Lakoff, George (). Hedges: a study in meaning criteria and the logic of fuzzy concepts. Journal of Philosophical Logic, : –. Lakoff, George (). Women, Fire and Dangerous Things. Chicago, IL: University of Chicago Press. doi: ./chicago/.. Lakoff, George (). The contemporary theory of metaphor. In Andrew Ortony (ed.), Metaphor and Thought, nd edn. Cambridge: Cambridge University Press. –. doi: ./ CBO.
Lakoff, George and Johnson, Mark (a). The metaphorical structure of the human conceptual system. Cognitive Science, (): –. Lakoff, George and Johnson, Mark (b). Metaphors We Live By. Chicago, IL: University of Chicago Press. Lakoff, Robin Tolmach (). The logic of politeness: minding your p’s and q’s. In Claudia W. Corum, Thomas Cedric Smith-Stark, and Ann Weiser (eds.), Papers from the Ninth Regional Meeting of the Chicago Linguistic Society. Chicago: Chicago Linguistics Society. –. Lakoff, Robin Tolmach (). Stylistic strategies within a grammar of style. Annals of the New York Academy of Sciences, : –. Lakusta, Laura and Carey, Susan (). Twelve-month-old infants’ encoding of goal and source paths in agentive and non-agentive motion events. Language Learning and Development, (): –. doi: ./.. Lakusta, Laura and Landau, Barbara (). Starting at the end: the importance of goals in spatial language. Cognition, (): –. Lakusta, Laura and Landau, Barbara (). Language and memory for motion events: origins of the asymmetry between source and goal paths. Cognitive Science, (): –. Lakusta, Laura, Wagner, Laura, O’Hearn, Kirsten, and Landau, Barbara (). Conceptual foundations of spatial language: evidence for a goal bias in infants. Language Learning and Development, (): –. Lambert, Bruce L. (). Face and politeness in pharmacist-physician interaction. Social Science and Medicine, (): –. Lambrecht, Knut (). Information Structure and Sentence Form. Cambridge: Cambridge University Press. Lammertink, Imme, Casillas, Marisa, Benders, Titia, Post, Brechtje, and Fikkert, Paula (). Dutch and English toddlers’ use of linguistic cues in predicting upcoming turn transitions. Frontiers in Psychology, : . doi: ./fpsyg.. Landau, Barbara and Jackendoff, Ray (). ‘What’ and ‘where’ in spatial language and spatial cognition. Behavioral and Brain Sciences, : –. doi: ./SX Landau, Barbara and Zukowski, Andrea (). Objects, motions, and paths: spatial language in children with Williams syndrome. Developmental Neuropsychology, (): –. doi: ./ SDNand_ Landau, Barbara, Johannes, Kristen, Skordos, Dimitrios, and Papafragou, Anna (). Containment and support: core and complexity in spatial language learning. Cognitive Science, (S): –. doi: ./cogs. Landman, Fred (a). Groups I. Linguistics and Philosophy, : –. Landman, Fred (b). Groups II. Linguistics and Philosophy, : –. Landman, Fred (). Plurality. In Shalom Lappin (ed.), Handbook of Contemporary Semantics, –. Oxford: Blackwell. Landman, Fred (). Events and Plurality: The Jerusalem Lectures. Dordrecht: Kluwer. Lange, K. W., Paul, G., Robbins, T., and Marsden, C. (). L-dopa and frontal cognitive function in Parkinson’s disease. Advances in Neurology, : –. Langendoen, D. Terence and Savin, Harris (). The projection problem for presuppositions. In Charles J. Fillmore and D. Terence Langendoen (eds.), Studies in Linguistic Semantics. New York: Holt, Reinhardt and Winston. –. Larson, Meredith, Doran, Ryan, McNabb, Yaron et al. (). Distinguishing the said from the implicated using a novel experimental paradigm. In Uli Sauerland and Kazuko Yatsushiro (eds.), Semantics and Pragmatics: From Experiment to Theory. Berlin: Palgrave Macmillan. –. Larson, Richard K. (). On the double object construction. Linguistic Inquiry, (): –. Larson, Richard and May, Robert (). Antecedent containment or vacuous movement: Reply to Baltin. Linguistic Inquiry, : –. Lasersohn, Peter (). Group action and spatio-temporal proximity. Linguistics and Philosophy, : –.
Lasersohn, Peter (). Existence presuppositions and background knowledge. Journal of Semantics, (): –. Lasersohn, Peter (). Plurality, Conjunction, and Events. Dordrecht: Kluwer. Lasersohn, Peter (). Generalized distributivity operators. Linguistics and Philosophy, : –. Lasersohn, Peter (). Pragmatic halos. Language, : –. Lasersohn, Peter (). Context dependence, disagreement, and predicates of personal taste. Linguistics and Philosophy, : –. Lassiter, Dan and Goodman, Noah D. (). Context, scale structure, and statistics in the interpretation of positive-form adjectives. In Todd Snider (ed.), Proceedings of Semantics and Linguistic Theory . Santa Cruz, CA: SLT. –. Lau, Ellen F., Phillips, Colin, and Poeppel, David (). A cortical network for semantics: (de) constructing the N. Nature Reviews Neuroscience, (): –. Lau, Ellen F., Wagers, Matt, Stroud, Clare, and Phillips, Colin (). Agreement and the subject of confusion. Talk given at CUNY Sentence Processing Conference , Chapel Hill, NC. Lawler, John (). Generic to a fault. In Paul M. Peranteau, Judith N. Levi, and Gloria C. Phares (eds.), Papers from the th Regional Meeting of the Chicago Linguistic Society. –. Chicago, IL: Chicago Linguistic Society. Lawler, John (). Studies in English Generics. University of Michigan Papers in Linguistics :. Ann Arbor, MI: University of Michigan Press. Lazaridou-Chatzigoga, Dimitra and Stockall, Linnaea (). Genericity, exceptions and domain restriction: experimental evidence from comparison with universals. In Emmanuel Chemla, Vincent Homer, and Grégoire Winterstein (eds.), Proceedings of Sinn und Bedeutung . Paris: École Normale Supérieure. –. Lazaridou-Chatzigoga, Dimitra, Katsos, Napoleon, and Stockall, Linnaea (). Genericity is easy? Formal and experimental perspectives. Ratio: An International Journal of Analytic Philosophy, (): –. Lazaridou-Chatzigoga, Dimitra, Katsos, Napoleon, and Stockall, Linnaea (a). Generic and universal generalisations: contextualising the ‘Generic Overgeneralisation’ effect. In Glenn Gunzelmann, Andrew Howes, Thora Tenbrink, and Eddy J. Davelaar (eds.), Proceedings of the th Annual Meeting of the Cognitive Science Society. Austin, TX: Cognitive Science Society. –. Lazaridou-Chatzigoga, Dimitra, Katsos, Napoleon, and Stockall, Linnaea (b). Exceptionality and strikingness in the acquisition of generics. Poster at IASCL , Lyon, France. Lazaridou-Chatzigoga, Dimitra, Stockall, Linnaea, and Katsos, Napoleon (). A new look at the ‘Generic Overgeneralisation’ effect. Inquiry. Advance online publication. doi: ./X.. Lebeaux, David (). Language acquisition and the form of the grammar. PhD thesis. University of Massachusetts, Amherst. Lee, James J. and Pinker, Steven (). Rationales for indirect speech: the theory of the strategic speaker. Psychological Review, : –. doi: ./a Lee, Thomas (). Studies on quantification in Chinese. PhD thesis. University of California, Los Angeles. Leech, Geoffrey N. (). Pragmatics and conversational rhetoric. In Herman Parret, Marina Sbisà, and Jef Verschueren (eds.), Possibilities and Limitations of Pragmatics. Amsterdam: Benjamins. Leech, Geoffrey N. (). Principles of Pragmatics. London: Longman. Leevers, Hilary J. and Harris, Paul L. (). Counterfactual syllogistic reasoning in normal -yearolds, children with learning disabilities, and children with autism. Journal of Experimental Child Psychology, : –. doi: ./jecp.. Legate, Julie and Yang, Charles (). Empirical re-assessment of the stimulus poverty arguments. Linguistic Review, : –. Léger, Catherine (). The acquisition of two types of factive complements. In Anna Gavarró and M. João Freitas (eds.), Language Acquisition and Development: Proceedings of GALA. Newcastle: Cambridge Scholars. –.
Lehrer, Adrienne (). Polysemy, conventionality, and the structure of the lexicon. Cognitive Linguistics, (): –. Leichty, Greg and Applegate, James L. (). Social-cognitive and situational influences on the use of face-saving persuasive strategies. Human Communication Research, (): –. doi: ./ j.-..tb.x Leslie, Alan M. (). Pretence and representation: the origins of a ‘theory of mind’. Psychological Review, : –. Leslie, Sarah-Jane (). Generics and the structure of the mind. Philosophical Perspectives, : –. Leslie, Sarah-Jane (). Generics: cognition and acquisition. Philosophical Review, (): –. Leslie, Sarah-Jane (forthcoming). The original sin of cognition: fear, prejudice, and generalisation. Journal of Philosophy. Leslie, Sarah-Jane and Gelman, Susan A. (). Quantified statements are recalled as generics: evidence from preschool children and adults. Cognitive Psychology, (): –. Leslie, Sarah-Jane, Khemlani, Sangeet, and Glucksberg, Sam (). All ducks lay eggs: the Generic Overgeneralization effect. Journal of Memory and Language, (): –. Leuthold, Hartmut, Filik, Ruth, Murphy, Kirsty, and Mackenzie, Ian G. (). The on-line processing of socio-emotional information in prototypical scenarios: inferences from brain potentials. Social Cognitive and Affective Neuroscience, : –. doi: ./scan/nsr Levin, Beth (). A constraint on verb meanings: manner/result complementarity. Cognitive Science Department Colloquium Series, Brown University, Providence, RI, March . Levin, Beth and Rappaport Hovav, Malka (). The lexical semantics of verbs of motion: the perspective from unaccusativity. In Iggy M. Roca (ed.), Thematic Structure: Its Role in Grammar. Berlin/New York: Foris Publications. –. Levin, Beth and Rappaport Hovav, Malka (). Unaccusativity: At the Syntax-Lexical Semantics Interface. Cambridge, MA: MIT Press. Levin, Beth and Rappaport Hovav, Malka (). Constraints on the complexity of verb meaning and VP structure. Between, : –. Levin, Beth and Rappaport Hovav, Malka (). Lexicalized manner and result are in complementary distribution. In Yehuda N. Falk (ed.), Proceedings of the th Annual Conference of the Israel Association for Theoretical Linguistics. Jerusalem: Israel Association of Theoretical Linguistics. Levin, Beth and Rappaport Hovav, Malka (). Lexicalized meaning and manner/result complementarity. In Boban Arsenijevic, Berit Gehrke, and Rafael Marín (eds.), Studies in the Composition and Decomposition of Event Predicates. New York: Springer. –. Levine, Susan Cohen and Carey, Susan (). Up front: the acquisition of a concept and a word. Journal of Child Language, : –. doi: ./s Levinson, Stephen C. (). Pragmatics. Cambridge: Cambridge University Press. Levinson, Stephen C. (). Vision, shape and linguistic description: Tzeltal body-part terminology and object description. Linguistics, : –. doi: ./ling...-. Levinson, Stephen C. (). Three levels of meaning. In Frank R. Palmer (ed.), Grammar and Meaning: Essays in Honour of Sir John Lyons. Cambridge: Cambridge University Press. –. Levinson, Stephen C. (). Frames of reference and Molyneux’s question: cross-linguistic evidence. In Paul Bloom, Mary A. Peterson, Lynn Nadel, and Merrill F. Garrett (eds.), Language and Space. Cambridge, MA: MIT Press. –. Levinson, Stephen C. (a). Presumptive Meanings: The Theory of Generalized Conversational Implicature. Cambridge, MA: MIT Press. Levinson, Stephen C. (b). H. P. Grice on Location on Rossel Island. In Steve S. Chang, Lily Liaw, and Josef Ruppenhofer (eds.), Proceedings of the th Annual Meeting of the Berkeley Linguistic Society. Berkeley, CA: Berkeley Linguistics Society. –. Levinson, Stephen C. (). Space in Language and Cognition: Explorations in Linguistic Diversity. Cambridge: Cambridge University Press. Levinson, Stephen C. (). Turn-taking in human communication: origins and implications for language processing. Trends in Cognitive Sciences, (): –.
Levinson, Stephen C. and Torreira, Francisco (). Timing in turn-taking and its implications for processing models of language. Frontiers in Psychology, : . Levinson, Stephen C. and Wilkins, David P. (eds.) (). Grammars of Space: Explorations in Cognitive Diversity. Cambridge: Cambridge University Press. Levinson, Stephen C., Kita, Sotaro, Haun, Daniel B. M., and Rasch, Björn H. (). Returning the tables: language affects spatial reasoning. Cognition, : –. doi: ./S-()- Lev-Rana, S., Shamay-Tsoory, Simone G., Zangenc, A., and Levkovitz, Yechiel (). Transcranial magnetic stimulation of the ventromedial prefrontal cortex impairs theory of mind learning. European Psychiatry, (): –. doi: ./j.eurpsy... Levy, Roger (). Expectation-based syntactic comprehension. Cognition, (): –. doi: ./j.cognition... Levy, Roger and Jaeger, T. Florian, (). Speakers optimize information density through syntactic reduction. In Bernhard Schölkopf, John Platt, and Thomas Hoffman (eds.), Advances in Neural Information Processing Systems. Cambridge, MA: MIT Press. –. Lewis, David (). Convention: A Philosophical Study. Cambridge, MA: Harvard University Press. Lewis, David (). Counterfactuals. Oxford: Blackwell. Lewis, David (). Adverbs of quantification. In Edward Keenan (ed.), Formal Semantics in Natural Languages. Cambridge: Cambridge University Press. –. Lewis, David (). Scorekeeping in a language game. In Rainer Bäuerle, Urs Egli, and Arnim von Stechow (eds.), Semantics from Different Points of View. Berlin: Springer. –. Lewis, David (). Logic for equivocators. Noûs, : –. Lewis, Fiona M., LaPointe, Leonard L., Murdoch, Bruce E., and Chenery, Helen J. (). Language impairment in Parkinson’s disease. Aphasiology, (): –. doi: ./ Lewis, Richard L. and Vasishth, Shravan (). An activation-based model of sentence processing as skilled memory retrieval. Cognitive Science, (): –. Lewis, Richard L., Vasishth, Shravan, and Van Dyke, Julie A. (). Computational principles of working memory in sentence comprehension. Trends in Cognitive Sciences, (): –. doi: ./j.tics... Lewis, Shevaun (). Pragmatic enrichment in language processing and development. PhD thesis. University of Maryland. Li, Peggy and Gleitman, Lila (). Turning the tables: language and spatial reasoning. Cognition, (): –. doi: ./S-()- Li, Peggy, Abarbanell, Linda, Gleitman, Lila, and Papafragou, Anna (). Spatial reasoning in Tenejapan Mayans. Cognition, (): –. doi: ./j.cognition... Liao, Shen-yi and Meskin, Aaron (). Aesthetic adjectives: experimental semantics and contextsensitivity. Philosophy and Phenomenological Research, : –. Liao, Shen-yi, McNally, Louise, and Meskin, Aaron (). Aesthetic adjectives lack uniform behavior. Inquiry: An Interdisciplinary Journal of Philosophy, : –. Liberman, Nira and Trope, Yaacov (). The psychology of transcending the here and now. Science, (): –. Lichtenstein, Sarah and Newman, J. Robert (). Empirical scaling of common verbal phrases associated with numerical probabilities. Psychonomic Science, : –. doi: ./BF Lidz, Jeffrey and Musolino, Julien (). Children’s command of quantification. Cognition, : –. Liebal, Kristin, Behne, Tanya, Carpenter, Malinda, and Tomasello, Michael (). Infants use shared experience to interpret pointing gestures. Developmental Science, : –. doi: ./j....x Liebesman, David (). Simple generics. Noûs, : –. Lieven, Elena (). Usage-based approaches to language development: where do we go from here? Language and Cognition, (): –. Light, Leah L. and Capps, Janet L. (). Comprehension of pronouns in young and older adults. Developmental Psychology, (): –. doi: ./-...
Lim, Tae-Seop and Bowers, John Waite (). Facework solidarity, approbation, and tact. Human Communication Research, (): –. doi: ./j.-..tb.x Lin, Jing (). Acquiring negative polarity items. PhD thesis. University of Amsterdam. Lin, Jing (). Distributionally constrained items in child language: the acquisition of superweak NPI shenme ‘a/some’ in Mandarin Chinese. Glossa: A Journal of General Linguistics, (): . doi: ./gjgl. Lin, Jing, Weerman, Fred, and Zeijlstra, Hedde (). Mandarin SHENME as a superweak NPI. In Jacob Hoeksema and Dirk Gilbers (eds.), Black Book: A Festschrift in Honor of Frans Zwarts. University of Groningen. –. Lin, Jo-Wang (). Polarity licensing and wh-phrase quantification in Chinese. PhD thesis. University of Massachussetts. Lin, Jo-Wang (). Distributivity in Chinese and its implications. Natural Language Semantics, : –. Linebarger, Marcia C. (). The grammar of negative polarity. PhD thesis. MIT, Cambridge, MA. Linebarger, Marcia C. (). Negative polarity and grammatical representation. Linguistics and Philosophy, (): –. Link, Godehard (). The logical analysis of plural and mass terms: a lattice theoretic approach. In Rainer Bäuerle, Christoph Schwarze, and Arnim von Stechow (eds.), Meaning, Use and Interpretation of Language. Berlin: De Gruyter. –. Link, Godehard (). Generalized quantifiers and plurals. In Peter Gärdenfors (ed.), Generalized Quantifiers. Dordrecht: D. Reidel. –. Link, Godehard (). Plurals. In Arnim von Stechow and Dieter Wunderlich (eds.), Semantik. Ein internationales Handbuch der zeitgenössischen Forschung. –. Berlin: De Gruyter. Liszkowski, Ulf, Carpenter, Malinda, and Tomasello, Michael (). Twelve-month-olds communicate helpfully and appropriately for knowledgeable and ignorant partners. Cognition, : –. Liu, In-mao, Lo, King-chung, and Wu, Jei-tun (). A probabilistic interpretation of ‘if-then’. Quarterly Journal of Experimental Psychology A: Human Experimental Psychology, A: –. doi: ./ Liu, Mingming (). Varieties of alternatives. PhD thesis. Rutgers, The State University of New Jersey. Liu, Ying and Yang, Yu’An (). Exhaustivity in Mandarin shi . . . (de) sentences: experimental evidence. In Marisa Köllner and Ramon Ziai (eds.), Proceedings of European Summer School in Logic Language and Information Student Session, Bolzano: Free University of Bozen. –. Online publication. Livesey, David J. and Morgan, Gillian A. (). The development of response inhibition in - and -year-old children. Australian Journal of Psychology, : –. doi: ./ Local, John, Wells, Bill, and Sebba, Mark (). Phonetic aspects of turn delimination in London Jamaican. Journal of Pragmatics, : –. Local, John, Kelly, John, and Wells, Bill (). Towards a phonology of conversation: turn-taking in Tyneside English. Journal of Linguistics, : –. Locher, Miriam A. (). Power and Politeness in Action: Disagreements in Oral Communication. Berlin: De Gruyter. Loewenstein, Jeffrey and Gentner, Dedre (). Relational language and the development of relational mapping. Cognitive Psychology, : –. http://groups.psych.northwestern.edu/gentner/papers/LoewensteinGentner.pdf Lombard, Lawrence B. (). How not to flip the prowler: transitive verbs of action and the identity of actions. In Ernest Lepore and Brian P. McLaughlin (eds.), Actions and Events: Perspectives on the Philosophy of Donald Davidson. New York: Routledge. –. Longobardi, Giuseppe (). Reference and proper names: a theory of N-movement in syntax and logical form. Linguistic Inquiry, (): –. Loukusa, Soile and Moilanen, Irma (). Pragmatic inference abilities in individuals with Asperger syndrome or high-functioning autism: a review. Research in Autism Spectrum Disorders, (): –.
Low, Jason, Apperly, Ian A., Butterfill, Stephen A., and Rakoczy, Hannes (). Cognitive architecture of belief reasoning in children and adults: a primer on the two-systems account. Child Development Perspectives, (): –. Lowder, Matthew W. and Gordon, Peter C. (). Focus takes time: structural effects on reading. Psychonomic Bulletin and Review, : –. Lucas, Amanda J. and Lewis, Charlie (). Should we trust experiments on trust? Human Development, (): –. doi: ./ Lüdtke, Jana and Kaup, Barbara (). Context effects when reading negative and affirmative sentences. In Ron Sun and Naomi Miyake (eds.), Proceedings of the th Annual Conference of the Cognitive Science Society. Mahwah, NJ: Lawrence Erlbaum Associates, Inc. –. Lüdtke, Jana, Friedrich, Claudia K., De Filippis, Mónica, and Kaup, Barbara (). Event-related potential correlates of negation in a sentence-picture verification paradigm. Journal of Cognitive Neuroscience, (): –. doi: ./jocn.. Łukasiewicz, Jan (). On three-valued logic. In Selected Works, ed. Ludwik Borkowski. Amsterdam: North-Holland. Originally published in Polish in Ruch Filozoficzny, (), –. MacDonald, Maryellen C., Pearlmutter, Neal J., and Seidenberg, Mark S. (). The lexical nature of syntactic ambiguity resolution. Psychological Review, : –. MacGregor, Lucy J., Bouwsema, Jennifer, and Klepousniotou, Ekaterini (). Sustained meaning activation for polysemous but not homonymous words: evidence from EEG. Neuropsychologia, : –. doi: ./j.neuropsychologia... Machina, Kenton F. (). Truth, belief, and vagueness. Journal of Philosophical Logic, : –. Reprinted in Keefe and Smith (a). MacKay, Gilbert and Shaw, Adrienne (). A comparative study of figurative language in children with autistic spectrum disorders. Child Language Teaching and Therapy, : –. Macnamara, John, Baker, Erica, and Olson, Chester L. (). Four-year-olds’ understanding of pretend, forget, and know: evidence for propositional operations. Child Development, (): –. MacWhinney, Brian, Keenan, Janice M., and Reinke, Peter (). The role of arousal in memory for conversation. Memory and Cognition, (): –. doi: ./BF Maes, Alfons, Arts, Anja, and Noordman, Leo (). Reference management in instructive discourse. Discourse Processes, (): –. doi: ./sdp_ Magri, Giorgio (). A theory of individual-level predicates based on blind mandatory scalar implicatures. Natural Language Semantics, (): –. doi: ./s---x Magri, Giorgio (). An account for the homogeneity effects triggered by plural definites and conjunction based on double strengthening. In Salvatore Pistoia Reda (ed.), Pragmatics, Semantics and the Case of Scalar Implicatures. London: Palgrave Macmillan. –. Maguire, Mandy J., Hirsh-Pasek, Kathy, Golinkoff, Roberta Michnick et al. (). A developmental shift from similar to language-specific strategies in verb acquisition: A comparison of English, Spanish and Japanese. Cognition, : –. doi: ./j.cognition... Magyari, Lilla and De Ruiter, Jan P. (). Prediction of turn-ends based on anticipation of upcoming words. Frontiers in Psychology, : . doi: ./fpsyg.. Magyari, Lilla, Bastiaansen, Marcel C. M., De Ruiter, Jan P., and Levinson, Stephen C. (). Early anticipation lies behind the speed of response in conversation. Journal of Cognitive Neuroscience, (): –. Mahowald, Kyle, Fedorenko, Evelina, Piantadosi, Steven T., and Gibson, Edward (). Info/information theory: speakers choose shorter words in predictive contexts. Cognition, (): –. doi: ./j.cognition... Majid, Asifa (). Words for parts of the body. In Barbara C. Malt and Phillip Wolff (eds.), Words and the Mind: How Words Capture Human Experience. Oxford: Oxford University Press. –. Majid, Asifa, Bowerman, Melissa, Kita, Sotaro, Haun, Daniel B. M., and Levinson, Stephen C. (). Can language restructure cognition? The case for space. Trends in Cognitive Science, : –. doi: ./j.tics...
Malt, Barbara C. (). The role of discourse structure in understanding anaphora. Journal of Memory and Language, : –. Malt, Barbara C. and Majid, Asifa (). How thought is mapped into words. Wiley Interdisciplinary Reviews: Cognitive Science, (): –. Mandelkern, Matthew (). Dissatisfaction theory. Talk presented at Semantics and Linguistic Theory . Mandler, Jean M. (). How to build a baby: II. Conceptual primitives. Psychological Review, (): –. Mangold, Roland and Pobel, Rupert (). Informativeness and instrumentality in referential communication. Journal of Language and Social Psychology, : –. doi: ./X Mann, William C. and Thompson, Sandra A. (). Rhetorical structure theory: toward a functional theory of text organization. Text—Interdisciplinary Journal for the Study of Discourse, : –. Manshadi, Mehdi Hafezi (). Dealing with quantifier scope ambiguity in natural language understanding. PhD thesis. University of Rochester. Marantz, Alec (). Generative linguistics within the cognitive neuroscience of language. Linguistic Review, : –. Marchman, Virginia A., Plunkett, Kim, and Goodman, Judith (). Overregularization in English plural and past tense inflectional morphology: a response to Marcus (). Journal of Child Language, (): –. Marcus, Gary F. (). The Algebraic Mind: Integrating Connectionism and Cognitive Science. Cambridge, MA: MIT Press. Marcus, Gary F., Pinker, Steven, Ullman, Michael et al. (). Overregularization in Language Acquisition. Monographs of the Society for Research in Child Development (, Serial no. ). Chicago: University of Chicago Press. Mari, Alda, Beyssade, Claire, and del Prete, Fabio (). Genericity. Oxford: Oxford University Press. Markman, Ellen M. and Wachtel, Gwyn F. (). Children’s use of mutual exclusivity to constrain the meanings of words. Cognitive Psychology, : –. doi: ./-()- Markman, Keith D. and Tetlock, Philip E. (). ‘I couldn’t have known’: accountability, foreseeability and counterfactual denials of responsibility. British Journal of Social Psychology, (): –. doi: ./ Marr, David (). Vision: A Computational Investigation into the Human Representation and Processing of Visual Information. New York: W.H. Freeman. Marslen-Wilson, William, Levy, Elena, and Tyler, Lorraine K. (). Producing interpretable discourse: the establishment and maintenance of reference. In Robert J. Jarvella and Wolfgang Klein (eds.), Speech, Place, and Action. Chichester: Wiley. –. Martin, Alex and Chao, Linda L. (). Semantic memory and the brain: structure and processes. Current Opinion in Neurobiology, : –. doi: ./S-()- Marty, Paul and Chemla, Emmanuel (). Scalar implicatures: working memory and a comparison with ‘only’. Frontiers in Psychology, : . doi: ./fpsyg.. Marty, Paul, Chemla, Emmanuel, and Spector, Benjamin (). Interpreting numerals and scalar items under memory load. Lingua, : –. doi: ./j.Lingua.... Marty, Paul, Chemla, Emmanuel, and Spector, Benjamin (). Phantom readings: the case of modified numerals. Language, Cognition and Neuroscience, (): –. Mateu, Jaume (). Argument structure: relational construal at the syntax-semantics interface. PhD thesis. Universitat Autònoma de Barcelona. Mateu, Jaume and Acedo-Matellán, Victor (). The Manner/Result complementarity revisited: a syntactic approach. In María Christina Cuervo and Yves Roberge (eds.), The End of Argument Structure. Bingley: Emerald Group. –. Matsumoto, Yo (). The conversational condition on Horn scales. Linguistics and Philosophy, : –. Matthews, Danielle, Lieven, Elena, Theakston, Anna, and Tomasello, Michael (). The effect of perceptual availability and prior discourse on young children’s use of referring expressions. Applied Psycholinguistics, (): –.
Matthews, Danielle, Lieven, Elena, and Tomasello, Michael (). What’s in a manner of speaking? Children’s sensitivity to partner-specific referential precedents. Developmental Psychology, : –. Mauner, Gail and Koenig, Jean-Pierre (). Linguistic vs. conceptual sources of implicit agents in sentence comprehension. Journal of Memory and Language, (): –. doi: ./jmla.. Mauner, Gail, Tanenhaus, Michael K., and Carlson, Gregory N. (). Implicit arguments in sentence processing. Journal of Memory and Language, (): –. May, Richard (). Logical Form: Its Structure and Derivation. Cambridge, MA: MIT Press. Mayo, Ruth, Schul, Yaacov, and Burnstein, Eugene (). ‘I am not guilty’ vs ‘I am innocent’: successful negation may depend on the schema used for its encoding. Journal of Experimental Social Psychology, (): –. doi: ./j.jesp... Mayol, Laia and Castroviejo, Elena (). Projective meaning and implicature cancellation. In Gregory Kierstead (ed.), Proceedings of the ESSLLI Workshop on Projective Content. Columbus, OH: Ohio State University. Mayr, Clemens (). Implicatures of modified numerals. In Ivano Caponigro and Carlo Cecchetto (eds.), From Grammar to Meaning: The Spontaneous Logicality of Language. Cambridge: Cambridge University Press. –. Mayr, Clemens (). Plural definite NPs presuppose multiplicity via embedded exhaustification. In Sarah D’Antonio, Mary Moroney, and Carol Rose Little (eds.), Proceedings of the th Semantics and Linguistic Theory Conference. Washington, DC: Linguistic Society of America. –. doi: ./salt.vi. Mayr, Clemens and Sauerland, Uli (). Accommodation and the strongest meaning hypothesis. In Pre-proceedings of the Amsterdam Colloquium . Amsterdam: Cisco System. Mazzarella, Diana (). Associative and inferential approaches to pragmatics: the state of the art of experimental investigation. Methode—Analytic Perspectives, (): –. Mazzarella, Diana (). Is inference necessary to pragmatics? Belgian Journal of Linguistics, : –. doi: ./bjl..maz Mazzone, Marco (). Schemata and associative processes in pragmatics. Journal of Pragmatics, : –. doi: ./j.pragma... Mazzone, Marco (a). The continuum problem: modified Occam’s Razor and conventionalisation of meaning. International Review of Pragmatics, (): –. Mazzone, Marco (b). Crossing the associative/inferential divide: ad hoc concepts and the inferential power of schemata. Review of Philosophy and Psychology, (): –. Mazzone, Marco (). What kind of associative and inferential processes? A response to RubioFernández (). International Review of Pragmatics, (): –. McCauliff, Catherine M. A. (). Burdens of proof: degrees of belief, quanta of evidence, or constitutional guarantees. Vanderbilt Law Review, , –. McCawley, James D. (). Pre-lexical syntax. In Richard J. O’Brien (ed.), Report on the nd Annual Round Table Meeting on Linguistics and Language Studies. Washington, DC: Georgetown University Press. –. McClelland, James L. and Patterson, Karalyn (). Rules or connections in past-tense inflections: what does the evidence rule out? Trends in Cognitive Sciences, (): –. McClelland, James L. and Rumelhart, David E. (). An interactive activation model of context effects in letter perception, part : an account of basic findings. Psychological Review, (): –. McConnell-Ginet, S. (). Adverbs and logical form: a linguistically realistic theory. Language, : –. McCourt, Michael, Green, Jeffrey J., Lau, Ellen, and Williams, Alexander (). Processing implicit control: evidence from reading times. Frontiers in Psychology, : . doi: ./fpsyg.. McDonald, Janet L. and MacWhinney, Brian (). The time course of anaphor resolution: effects of implicit verb causality and gender. Journal of Memory and Language, (): –. doi: ./ jmla..
McDonough, Laraine, Choi, Soonja, and Mandler, Jean M. (). Understanding spatial relations: flexible infants, lexical adults. Cognitive Psychology, : –. doi: ./S-()– McElree, Brian and Nordlie, Johanna (). Literal and figurative interpretations are computed in parallel. Psychonomic Bulletin and Review, : –. doi: ./BF McElree, Brian, Frisson, Steven, and Pickering, Martin J. (). Deferred interpretations: why starting Dickens is taxing but reading Dickens isn’t. Cognitive Science, (): –. McGlone, Matthew S. (). Concepts as metaphors. In Sam Glucksberg (ed.), Understanding Figurative Language: From Metaphor to Idioms. Oxford: Oxford University Press. –. McGlone, Matthew S. and Manfredi, Deanna A. (). Topic-vehicle interaction in metaphor comprehension. Memory and Cognition, : –. doi: ./BF McKenna, Peter E., Glass, Alexandra, Rajendran, Gnanathusharan, and Corley, Martin (). Strange words: autistic traits and the processing of non-literal language. Journal of Autism and Developmental Disorders, (): –. McKinstry, Chris, Dale, Rick, and Spivey, Michael J. (). Action dynamics reveal parallel competition in decision making. Psychological Science, (): –. doi: ./j.-...x McKoon, Gail and Love, Jessica (). Verbs in the lexicon: why is hitting easier than breaking? Language and Cognition, (): –. McKoon, Gail and Macfarland, Talke (). Event templates in the lexical representations of verbs. Cognitive Psychology, (): –. doi: ./S-()-X McKoon, Gail and Ratcliff, Roger (). Meanings, propositions, and verbs. Psychonomic Bulletin and Review, (): –. McLaughun, M. L., Cody, Michael J., and O’Hair, H. Dan (). The management of failure events: some contextual determinants of accounting behavior. Human Communication Research, (): –. doi: ./j.-..tb.x McNabb, Yaron (). Syntax and semantics of degree modification. PhD thesis. University of Chicago. McNabb, Yaron and Penka, Doris (). An experimental investigation of ignorance inferences and authoritative interpretations of superlative modifiers. Under review. McNally, Louise (). The relative role of property type and scale structure in explaining the behavior of gradable adjectives. In Rick Nouwen, Robert van Rooij, Uli Sauerland, and HansChristian Schmitz (eds.), Vagueness in Communication. Heidelberg: Springer. –. McNally, Louise (). Scalar alternatives and scalar inference involving adjectives: a comment on Van Tiel, et al. (). In Jason Ostrove, Ruth Kramer, and Joseph Sabbagh (eds.), Asking the Right Questions: Essays in Honor of Sandra Chung. –. Santa Cruz, CA: Linguistics Research Center, University of California. McNally, Louise and Stojanovic, Isidora (). Aesthetic adjectives. In James O. Young (ed.), The Semantics of Aesthetic Judgment. Oxford: Oxford University Press. –. McNamara, Patrick and Durso, Raymon (). Pragmatic communication skills in patients with Parkinson’s disease. Brain and Language, (): –. McNamara, Patrick, Durso, Raymon, and Harris, Erica (). Alterations of the sense of self and personality in Parkinson’s disease. International Journal of Geriatric Psychiatry, (): –. doi: ./gps. McRae, Ken, Spivey-Knowlton, Michael J., and Tanenhaus, Michael K. (). Modeling the influence of thematic fit (and other constraints) in on-line sentence comprehension. Journal of Memory and Language, (): –. doi: ./jmla.. Medin, Douglas L. and Ortony, Andrew (). Psychological essentialism. In Stella Vosniadou and Andrew Ortony (eds.), Similarity and Analogical Reasoning. New York: Cambridge University Press. –. Medin, Douglas L. and Schaffer, Marguerite M. (). Context theory of classification learning. Psychological Review, (): –. doi: ./-X... Meehan, Julie E. and Byrne, Ruth M. J. (). The temporal order effect in children’s counterfactual thinking. In Bruno G. Bara, Lawrence Barsalou, and Monica Bucciarelli (eds.), Proceedings of the
th Annual Conference of the Cognitive Science Society. Mahwah, NJ: Lawrence Erlbaum Associates. –. Mehlberg, Henry (). The Reach of Science. Toronto: University of Toronto Press. Mehrabian, Albert (). Nonverbal Communication. Transaction Publishers. Meibauer, Jörg and Steinbach, Markus (eds.) (). Experimental Pragmatics/Semantics. Amsterdam: John Benjamins. Meltzoff, Andrew N. (). Understanding the intentions of others: re-enactment of intended acts by -month-old children. Developmental Psychology, (): –. doi: ./-... Merchant, Jason (a). Antecedent-contained deletion in negative polarity items. Syntax, : –. Merchant, Jason (b). Economy, the copy theory, and antecedent-contained deletion. Linguistic Inquiry, : –. Merin, Arthur (). Information, relevance, and social decisionmaking: some principles and results of decision-theoretic semantics. In Lawrence S. Moss, Jonathan Ginzburg, and Maarten de Rijke (eds.), Logic, Language, and Computation, Vol. . Stanford, CA: CSLI Publications. –. Mervis, Carolyn B. and Johnson, Kathy E. (). Acquisition of the plural morpheme: a case study. Developmental Psychology, : –. Metzing, Charles and Brennan, Susan E. (). When conceptual pacts are broken: partner-specific effects on the comprehension of referring expressions. Journal of Memory and Language, : –. Meyer, Christian (). Predicaments of polyphony: sequence, self and style in Wolof conversations. Africana Studia, : –. Meyer, Marie-Christine (). Ignorance and grammar. PhD thesis. MIT. Meyer, Meredith, Gelman, Susan A., and Stilwell, Sarah M. (). Generics are a cognitive default: evidence from sentence processing. In Laura Carlson, Christoph Hölscher, and Thomas F. Shipley (eds.), Proceedings of the rd Annual Conference of the Cognitive Science Society. Austin, TX: Cognitive Science Society. Micham, Dennis L., Catlin, Jack, VanDerveer, Nancy J., and Loveland, Katherine A. (). Lexical and structural cues to quantifier scope relations. Journal of Psycholinguistic Research, (): –. doi: ./BF Miller, George A. and Johnson-Laird, Philip N. (). Language and Perception. Cambridge, MA: Harvard University Press. Miller, Karen and Schmitt, Cristina (). Wide-scope indefinites in English child language. In Jacqueline van Kampen and Sergio Bauuw (eds.), Proceedings of Generative Approaches to Language Acquisition . Utrecht: LOT Occasional series. –. Miller, Karen, Schmitt, Cristina, Chang, Hsiang-Hua, and Munn, Alan (). Young children understand some implicatures. In Alejna Brugos, Manuella R. Clark-Cotton, and Seungwan Ha (eds.), Proceedings of the th Annual Boston University Conference on Language Development. Somerville, MA: Cascadilla Press. –. Milligan, Karen, Astington, Janet W., and Dack, Lisa Ain (). Language and Theory of Mind: meta-analysis of the relation between language ability and false-belief understanding. Child Development, (): –. Mills, Candice M. (). Knowing when to doubt: developing a critical stance when learning from others. Developmental Psychology, (): –. doi: ./a Minai, Utako, Jincho, Nobuyuki, Yamane, Naoto, and Mazuka, Reiko (). What hinders child semantic computation: children’s universal quantification and the development of cognitive control. Journal of Child Language, (): –. Mitchell, Don C. (). Lexical guidance in human parsing: locus and processing characteristics. In Max Coltheart (ed.), Attention and Performance XII: The Psychology of Reading. Hillsdale, NJ: Erlbaum. –. Mittwoch, Anita (). In defence of Vendler’s achievements. Belgian Journal of Linguistics, (): –.
Mo, Yoonsook, Cole, Jennifer, and Lee, Eun-Kyung (). Naïve listeners’ prominence and boundary perception. Proceedings of Speech Prosody . Campiñas: Speech Prosody. –. Moens, Marc and Steedman, Mark (). Temporal ontology and temporal reference. Computational Linguistics, (): –. Moll, Henrike and Meltzoff, Andy N. (). Perspective-taking and its foundation in joint attention. In Naomi Eilan, Hemdat Lerman, and Johannes Roessler (eds.), Perception, Causation, and Objectivity. Oxford: Oxford University Press. –. Moll, Henrike and Tomasello, Michael (). How - and -month-olds know what others have experienced. Developmental Psychology, (): –. Moll, Henrike, Richter, Nadja, Carpenter, Malinda, and Tomasello, Michael (). Fourteen-montholds know what ‘we’ have shared in a special way. Infancy, (): –. Moll, Henrike, Carpenter, Malinda, and Tomasello, Michael (). Social engagement leads -yearolds to overestimate others’ knowledge. Infancy, (): –. Monetta, Laura and Pell, Marc D. (). Effects of verbal working memory deficits on metaphor comprehension in patients with Parkinson’s disease. Brain and Language, : –. Montague, Richard (). The proper treatment of quantification in ordinary English. In Patrick Suppes, Jaakko Hintikka, and Julius Moravcsik (eds.), Approaches to Natural Language. Dordrecht: Reidel. –. Montale, Eugenio (). Satura: –. Milan: Mondadori. Montrul, Silvina and Ionin, Tania (). Transfer effects in the interpretation of definite articles by Spanish heritage speakers. Bilingualism: Language and Cognition, (): –. Montrul, Silvina and Ionin, Tania (). Dominant language transfer in Spanish heritage speakers and second language learners in the interpretation of definite articles. Modern Language Journal, : –. Moore, Chris and Davidge, Jane (). The development of mental terms: pragmatics or semantics? Journal of Child Language, (): –. doi: ./SX Moore, Chris, Bryant, Dana, and Furrow, David (). Mental terms and the development of certainty. Child Development, (): –. doi: ./ Morison, Patricia and Gardner, Howard (). Dragons and dinosaurs: the child’s capacity to differentiate fantasy from reality. Child Development, : –. Morisseau, Tiffany, Davies, Catherine, and Matthews, Danielle (). How do - and -year-olds respond to over- and under-informative utterances? Journal of Pragmatics, (A): –. Morris, Robin K. and Folk, Jocelyn R. (). Focus as a contextual priming mechanism in reading. Memory and Cognition, (): –. doi: ./BF Morrow, Daniel G. and Clark, Herbert H. (). Interpreting words in spatial descriptions. Language and Cognitive Processes, : –. doi: ./ Morzycki, Marcin (). Modification. Cambridge University Press. Most, Robert B. and Saltz, Eli (). Information structure in sentences: new information. Language and Speech, : –. Moxey, Linda M. and Sanford, Anthony J. (). Communicating Quantities: A Psychological Perspective. Hillsdale, NJ: Erlbaum. Moxey, Linda M. and Sanford, Anthony J. (). Communicating quantities: a review of psycholinguistic evidence of how expressions determine perspectives. Applied Cognitive Psychology, : –. doi: ./(SICI)- Mulders, Iris and Szendrői, Kriszta (). Early association of prosodic focus with alleen ‘only’: evidence from eye movements in the visual-world paradigm. Frontiers in Psychology, : . doi: ./fpsyg.. Mullet, Etienne and Rivet, Isabelle (). Comprehension of verbal probability expressions in children and adolescents. Language and Communication, : –. doi: ./-()-I Munnich, Edward, Landau, Barbara, and Dosher, Barbara Anne (). Spatial language and spatial representation: a cross-linguistic comparison. Cognition, (): –. doi: ./ S-()-
Munsell, Albert Henry (). Atlas of the Munsell Color System. Malden, MA: Wadsworth, Howland and Company. Murphy, Gregory, L. (). Parsimony and the psychological representation of polysemous words. In Marina Rakova, Gergely Pethő, and Csilla Rákosi (eds.), The Cognitive Basis of Polysemy. Frankfurt: Peter Lang. Murphy, Gregory, L. and Medin, Douglas L. (). The role of theories in conceptual coherence. Psychological Review, (): –. doi: ./-X... Musolino, Julien (). The semantics and acquisition of number words: integrating linguistic and developmental perspectives. Cognition, (): –. Musolino, Julien (). The logical syntax of number words: theory, acquisition and processing. Cognition, : –. Musolino, Julien and Lidz, Jeffrey (). Why children aren’t universally successful with quantification, Linguistics, : –. Musolino, Julien, Crain, Stephen, and Thornton, Rosalind (). Navigating negative quantificational space. Linguistics, : –. Nadig, Aparna S. and Sedivy, Julie C. (). Evidence of perspective-taking constraints in children’s online reference resolution. Psychological Science, (): –. doi: ./j.-...x Naigles, Letitia R. and Terrazas, Paula (). Motion-verb generalizations in English and Spanish: influences of language and syntax. Psychological Science, : –. doi: ./-. Naigles, Letitia R., Eisenberg, Ann R., Kako, Edward T., Highter, Melissa, and McGraw, Nancy (). Speaking of motion: verb use in English and Spanish. Language and Cognitive Processes, : –. doi: ./ Nappa, Rebecca and Arnold, Jennifer E. (). The road to understanding is paved with the speaker’s intentions: cues to the speaker’s attention and intentions affect pronoun comprehension. Cognitive Psychology, : –. doi: ./j.cogpsych... Neale, Stephen (). Paul Grice and the philosophy of language. Linguistics and Philosophy, (): –. Neely, James H. (). Semantic priming and retrieval from lexical memory: roles of inhibitionless spreading activation and limited-capacity attention. Journal of Experimental Psychology: General, (): –. doi: ./-... Neldner, Karri, Collier-Baker, Emma, and Nielsen, Mark (). Chimpanzees (Pan troglodytes) and human children (Homo sapiens) know when they are ignorant about the location of food. Animal Cognition, : –. doi: ./s--- Nerlich, Brigitte, Clarke, David D., and Todd, Zazie (). ‘Mummy, I like being a sandwich’: metonymy in language acquisition. In Klaus-Uwe Panther and Günter Radden (eds.), Metonymy in Language and Thought. Amsterdam: John Benjamins. –. Newcombe, Nora S. and Huttenlocher, Janellen (). Making Space: The Development of Spatial Representation and Reasoning. Cambridge, MA: MIT Press. Newstead, Stephen E., Keeble, Stephanie, and Manktelow, Kenneth I. (). Children’s performance on set-inclusion and linear-ordering relationships. Bulletin of the Psychonomic Society, (): –. doi: ./BF Ni, Weijia, Crain, Stephen, and Shankweiler, Donald (). Sidestepping garden paths: assessing the contributions of syntax, semantics and plausibility in resolving ambiguities. Language and Cognitive Processes, (): –. Ni, Weijia, Fodor, Janet Dean, Crain, Stephen, and Shankweiler, Donald P. (). Anomaly detection: eye movement patterns. Journal of Psycholinguistic Research, (): –. doi: ./ A: Nickel, Bernhard (). Between Logic and the World. Oxford: Oxford University Press. Nickel, Bernhard (forthcoming). Generics. In Bob Hale, Alexander Miller, and Crispin Wright (eds.), Blackwell Companion to the Philosophy of Language, nd Edition. Oxford: Blackwell. Niedeggen, Michael, Rösler, Frank, and Jost, Kerstin (). Processing of incongruous mental calculation problems: evidence for an arithmetic N effect. Psychophysiology, (): –.
Nieuwland, Mante S. (). Establishing propositional truth-value in counterfactual and real-world contexts during sentence comprehension: differential sensitivity of the left and right inferior frontal gyri. NeuroImage, : –. Nieuwland, Mante S. (). ‘If a lion could speak . . . ’: online sensitivity to propositional truth-value of unrealistic counterfactual sentences. Journal of Memory and Language, : –. Nieuwland, Mante S. and Kuperberg, Gina R. (). When the truth is not too hard to handle: an event-related potential study on the pragmatics of negation. Psychological Science, (): –. doi: ./j.–...x Nieuwland, Mante S. and Martin, Andrea E. (). If the real world were irrelevant, so to speak: the role of propositional truth-value in counterfactual sentence comprehension. Cognition, : –. doi: ./j.cognition... Nieuwland, Mante S. and Van Berkum, Jos J. A. (). When peanuts fall in love: N evidence for the power of discourse. Journal of Cognitive Neuroscience, : –. Nieuwland, Mante S., Ditman, Tali, and Kuperberg, Gina R. (). On the incrementality of pragmatic processing: an ERP investigation of informativeness and pragmatic abilities. Journal of Memory and Language, : –. Nilsen, ystein (). The Syntax of Circumstantial Adverbials. Oslo: Novus Press. Nippold, Marilyn A. (/). Later Language Development: The School-Age and Adolescent Years, nd edn. Austin, TX: Pro-Ed. Norbury, Courtenay F. (). The relationship between theory of mind and metaphor: evidence from children with language impairment and autistic spectrum disorder. British Journal of Developmental Psychology, : –. Norbury, Heather M., Waxman, Sandra R., and Song, Hyun-Joo (). Tight and loose are not created equal: an asymmetry underlying the representation of fit in English and Korean speakers. Cognition, (): –. doi: ./j.cognition... Norrick, Neal (). Semiotic Principles in Semantic Theory. Amsterdam/Philadelphia: John Benjamins. Nouwen, Rick (). Two kinds of modified numerals. Semantics and Pragmatics, (): –. Nouwen, Rick (). Modified numerals: the epistemic effect. In Luis Alonso-Ovalle and Paula Ménendez-Benito (eds.), Epistemic Indefinites. Oxford: Oxford University Press. Noveck, Ira A. (). When children are more logical than adults: experimental investigations of scalar implicatures. Cognition, : –. Noveck, Ira A. (). Experimental Pragmatics: The Making of a Cognitive Science. Cambridge: Cambridge University Press. Noveck, Ira A. and Posada, Andres (). Characterizing the time course of an implicature: an evoked potentials study. Brain and Language, (): –. doi: ./s–x()–. Noveck, Ira A. and Reboul, Anne (). Experimental pragmatics: a Gricean turn in the study of language. Trends in Cognitive Sciences, (): –. Noveck, Ira A. and Sperber, Dan (eds.) (). Experimental Pragmatics. Houndmills: Palgrave Macmillan. Noveck, Ira A. and Sperber, Dan (). The why and how of experimental pragmatics: the case of ‘scalar inferences’. In Noel Burton-Roberts (ed.), Advances in Pragmatics. Basingstoke: Palgrave. Noveck, Ira A. and Spotorno, Nicola (). Narrowing. In Laurence Goldstein (ed.), Brevity. Oxford: Oxford University Press. Noveck, Ira A., Bianco, Maryse, and Castry, Alain (). The costs and benefits of metaphor. Metaphor and Symbol, (–): –. doi: ./.. Noveck, Ira A., Guelminger, Raphaële, Georgieff, Nicolas, and Labruyere, Nelly (). What autism can reveal about every . . . not sentences. Journal of Semantics, (): –. doi: ./jos/ffl Nunberg, Geoffrey (). The non-uniqueness of semantic solutions: polysemy. Linguistics and Philosophy, : –. Nunberg, Geoffrey (). Transfers of meaning. Journal of Semantics, : –. Nunberg, Geoffrey (). The pragmatics of deferred interpretation. In Laurence R. Horn and Gregory Ward (eds.), Handbook of Pragmatics. Oxford: Blackwell. –.
Nunberg, Geoffrey and Pan, Chiahua (). Inferring quantification in generic sentences. In Robin E. Grossman, L. James San, and Timothy J. Vance (eds.), Papers from the XIth Regional Meeting of the Chicago Linguistic Society. Chicago, IL: Chicago Linguistic Society. –. Nurmsoo, Erika and Bloom, Paul (). Preschoolers’ perspective-taking in word learning: do they blindly follow eye gaze? Psychological Science, : –. doi: ./j.-...x Oakhill, Jane (). Mental models of children’s text comprehension. In Jane Oakhill and Alan Garnham (eds.), Mental Models in Cognitive Science: Essays in Honour of Phil Johnson-Laird. Hove: Psychology Press. –. Oaksford, Mike and Moussakowski, Marek (). Negations and natural sampling in data selection: ecological versus heuristic explanations of matching bias. Memory and Cognition, (): –. doi: ./BF Oaksford, Mike and Stenning, Keith (). Reasoning with conditionals containing negated constituents. Journal of Experimental Psychology: Learning, Memory and Cognition, (): –. doi: ./-... O’Brien, Bernie J. (). Words or numbers? The evaluation of probability expressions in general practice. Journal of the Royal College of General Practitioners, : –. O’Brien, Edward J., Shank, Dolores M., Myers, Jerome L., and Rayner, Keith (). Elaborative inferences during reading: do they occur on-line? Journal of Experimental Psychology: Learning, Memory and Cognition, (): –. doi: ./-.... O’Donnell, Timothy J. (). Productivity and Reuse in Language: A Theory of Linguistic Computation and Storage. Cambridge, MA: MIT Press. O’Donoghue, Josie (). Is a metaphor (like) a simile? Differences in meaning, effects and processing. UCL Working Papers in Linguistics, : –. O’Grady, William, Suzuki, Takaaki, and Yoshinaga, Naoko (). Quantifier spreading: new evidence from Japanese. Language Learning and Development, (): –. Oiler, D. Kimbrough (). The effect of position in utterance on speech segment duration in English. Journal of the Acoustical Society of America, : –. Okamoto, Shinichiro and Robinson, W. Peter (). Determinants of gratitude expressions in England. Journal of Language and Social Psychology, (): –. doi: ./X Olson, Michael J. and Budescu, David V. (). Patterns of preference for numerical and verbal probabilities. Journal of Behavioral Decision Making, : –. Onea, Edgar and Beaver, David (). Hungarian focus is not exhausted. In Ed Cormany, Satoshi Ito, and David Lutz (eds.), Proceedings of the th Semantics and Linguistic Theory Conference. Ithaca, NY: CLC Publications. eLanguage. –. O’Neill, Daniela K. (). Two-year-old children’s sensitivity to a parent’s knowledge state when making requests. Child Development, (): –. doi: ./j.-..tb.x Onishi, Kristine H. and Baillargeon, Renée (). Do -month-old infants understand false beliefs? Science, (): –. Oostdijk, Nelleke (). Het Corpus Gesproken Nederlands. Amsterdam: Dutch Language Union. Orenes, Isabel, Beltrán, David, and Santamaría, Carlos (). How negation is understood: evidence from the visual world paradigm. Journal of Memory and Language, : –. doi: ./j. jml... Orenes, Isabel, Moxey, Linda M., Scheepers, Christoph, and Santamaría, Carlos (). Negation in context: evidence from the visual world paradigm. Quarterly Journal of Experimental Psychology, (): –. doi: ./.. Ortony, Andrew (). Beyond literal similarity. Psychological Review, (): –. Ortony, Andrew, Schallert, Diane L., Reynolds, Ralph E., and Antos, Stephen J. (). Interpreting metaphors and idioms: some effects of context on comprehension. Journal of Verbal Learning and Verbal Behavior, : –. Osherson, Daniel N. and Edward E. Smith (). Gradedness and conceptual combination. Cognition, : –.
Osterhout, Lee and Holcomb, Phillip J. (). Event-related brain potentials elicited by syntactic anomaly. Journal of Memory and Language, (): –. Osterhout, Lee, Holcomb, Phillip J., and Swinney, David A. (). Brain potentials elicited by garden-path sentences: evidence of the application of verb information during parsing. Journal of Experimental Psychology: Learning, Memory, and Cognition, (): –. Osterhout, Lee, Allen, Mark, and McLaughlin, Judith (). Words in the brain: lexical determinants of word-induced brain activity. Journal of Neurolinguistics, (–): –. Ouyang, Iris Chuoying and Kaiser, Elsi (). Prosody and information structure in a tone language: an investigation of mandarin chinese. Language, Cognition and Neuroscience, (–): –. Owen, Adrian M., James, M. F., Leigh, Peter Nigel et al. (). Fronto-striatal cognitive deficits at different stages of Parkinson’s disease. Brain, (): –. doi: ./brain/.. Özçaliskan, Şeyda (). On learning to draw the distinction between physical and metaphorical motion: is metaphor an early emerging cognitive and linguistic capacity? Journal of Child Language, : –. Özçaliskan, Şeyda (). Metaphors we move by: children’s developing understanding of metaphorical motion in typologically distinct languages. Metaphor and Symbol, (): –. Ozturk, Ozge and Papafragou, Anna (). The acquisition of epistemic modality: from semantic meaning to pragmatic interpretation. Language Learning and Development, (): –. doi: ./.. Pablos, Leticia, Doetjes, Jenny, Ruijgrok, Bobby, and Cheng, Lisa L. (). Backward NPI dependencies in Dutch: an ERP investigation. Poster presented at the th Annual CUNY Sentence Processing Conference, CUNY Graduate Center, New York. Paczynski, Martin, Jackendoff, Ray, and Kuperberg, Gina (). When events change their nature: the neurocognitive mechanisms underlying aspectual coercion. Journal of Cognitive Neuroscience, (): –. Pagliarini, Elena, Fiorin, Gaetano, and Dotlačil, Jakub (). The acquisition of distributivity in pluralities. In Alia K. Biller, Esther Y. Chung, and Amelia E. Kimball (eds.), Proceedings of the th Annual Boston University Conference on Language Development. Somerville, MA: Cascadilla Press. –. Panizza, Daniele, Chierchia, Gennaro, and Clifton, Charles, Jr. (). On the role of entailment patterns and scalar implicatures in the processing of numerals. Journal of Memory and Language, (): –. Panther, Klaus-Uwe and Thornburg, Linda L. (). Metonymy. In Dirk Geeraerts and Hubert Cuyckens (eds.), Oxford Handbook of Cognitive Linguistics. Oxford: Oxford University Press. –. Papafragou, Anna (a). On metonymy. Lingua, (): –. Papafragou, Anna (b). Figurative language and the semantics-pragmatics distinction. Language and Literature, (): –. Papafragou, Anna (). Source-goal asymmetries in motion representation: implications for language production and comprehension. Cognitive Science, (): –. doi: ./ j.–...x Papafragou, Anna (). The representation of events in language and cognition. In Eric Margolis and Stephen Laurence (eds.), The Conceptual Mind: New Directions in the Study of Concepts. Cambridge, MA: MIT Press. –. Papafragou, Anna and Musolino, Julien (). Scalar implicatures: experiments at the semanticspragmatics interface. Cognition, : –. Papafragou, Anna and Selimis, Stathis (). Lexical and structural biases in the acquisition of motion verbs. Language Learning and Development, (): –. doi: ./ Papafragou, Anna and Skordos, Dimitrios (). Scalar implicature. In Jeffrey Lidz, William Snyder, and Joe Pater (eds.), Oxford Handbook of Developmental Linguistics. Oxford: Oxford University Press. Papafragou, Anna, Massey, Christine, and Gleitman, Lila (). Shake, rattle, ’n’ roll: the representation of motion in language and cognition. Cognition, : –. doi: ./S()-X
Papafragou, Anna, Massey, Christine, and Gleitman, Lila (). When English proposes what Greek presupposes: the cross-linguistic encoding of motion events. Cognition, (): B–B. doi: ./j.cognition... Papafragou, Anna, Hulbert, Justin, and Trueswell, John C. (). Does language guide event perception? Evidence from eye movements. Cognition, (): –. Papafragou, Anna, Viau, Joshua B., and Landau, Barbara (). The ins and outs of spatial language: paths, places, and negative spatial prepositions. Paper presented at the th Annual Meeting of the Boston University Conference on Language Development, Boston, MA. Paraboni, Ivandré, Van Deemter, Kees, and Masthoff, Judith (). Generating referring expressions: making referents easy to identify. Computational Linguistics, : –. ./coli.... Paris, Scott G. (). Comprehension of language connectives and propositional logical relationships. Journal of Experimental Child Psychology, : –. doi: ./-()- Parisi, Domenico and Antinucci, Francesco (). Lexical competence. In Giovanni B. Flores d’Arcais and Willem J. M. Levelt (eds.), Advances in Psycholinguistics. Amsterdam: North-Holland. Parker, Dan and Phillips, Colin (). Negative polarity illusions and the format of hierarchical encodings in memory. Cognition, : –. Parsons, Terence (). Modifiers and quantifiers in natural language. Canadian Journal of Philosophy, sup. vol. VI: –. Parsons, Terence (). Events in the Semantics of English: A Study in Subatomic Semantics. Cambridge, MA: MIT Press Passanini, Alessia and Hampton, James A. (). Gender and generics: what makes a gender-specific generic sentence true? In Gabriella Airenti, Bruno G. Bara, Giulio Sandini, and Marco Cruciani (eds.), Proceedings of the EuroAsianPacific Joint Conference on Cognitive Science, th European Conference on Cognitive Science and th International Conference on Cognitive Science. –. Turin: CEUR-WS.org. Sun SITE Central Europe. Paterson, Kevin B., Liversedge, Simon P., and Underwood, Geoffrey (). The influence of focus operators on syntactic processing of short relative clause sentences. Quarterly Journal of Experimental Psychology Section A, (): –. Paterson, Kevin B., Liversedge, Simon P., Filik, Ruth, Juhasz, Barbara J., White, Sarah J., and Rayner, Keith (). Focus identification during sentence comprehension: evidence from eye movements. Quarterly Journal of Experimental Psychology, (): –. Paterson, Kevin B., Filik, Ruth, and Liversedge, Simon P. (). Competition during the processing of quantifier scope ambiguities: evidence from eye movements during reading. Quarterly Journal of Experimental Psychology, : –. Patson, Nicole and Warren, Tessa (). Evidence for distributivity effects in comprehension. Journal of Experimental Psychology: Learning, Memory, and Cognition, : –. Patson, Nikole D. (). Evidence in support of a scalar implicature account of plurality. Journal of Experimental Psychology: Learning, Memory, and Cognition, (): –. Patson, Nikole D., George, Gerret, and Warren, Tessa (). The conceptual representation of number. Quarterly Journal of Experimental Psychology, : –. Paul, Hermann (). Prinzipien der Sprachgeschichte. Tübingen: Niemeyer. Paul, Hermann (). Deutsche Grammatik, vol. . Tübingen: Niemeyer. Pearlmutter, Neal J., Garnsey, Susan M., and Bock, Kathryn (). Agreement processes in sentence comprehension. Journal of Memory and Language, (): –. doi: ./jmla.. Pearson, Barbara Zurer (). The comprehension of metaphor by preschool children. Journal of Child Language, (): –. doi: ./S Pearson, Hazel, Khan, Manizeh, and Snedeker, Jesse (). Even more evidence for the emptiness of plurality: an experimental investigation of plural interpretation as a species of scalar implicature. In Nan Li and David Lutz (eds.), Proceedings of the th Conference on Semantics and Linguistic Theory. Ithaca, NY: CLC Publications. Pechmann, Thomas (). Incremental speech production and referential overspecification. Linguistics, (): –.
Pederson, Eric, Danziger, Eve, Wilkins, David, Levinson, Stephen C., Kita, Sotaro, and Senft, Gunter (). Semantic typology and spatial conceptualization. Language, : –. doi: ./ Peirce, Charles S. (). Abduction and induction. In Justus Buchler (ed.), Philosophical Writings of Peirce. Dover Publications. Peleg, Orna, Giora, Rachel, and Fein, Ofer (). Salience and context effects: two are better than one. Metaphor and Symbol, (–): –. Pell, Marc D. (). Reduced sensitivity to prosodic attitudes in adults with focal right hemisphere brain damage. Brain and Language, (): –. doi: ./j.bandl... Pelletier, Francis J. (). Generics: a philosophical introduction. In Francis J. Pelletier (ed.), Kinds, Things, and Stuff: The Cognitive Side of Generics and Mass Terms. New York: Oxford University Press. –. Percus, Orin (). Prying open the cleft. In Kiyomi Kusumoto (ed.), Proceedings of the North East Linguistic Society . Amherst, MA: Graduate Linguistic Student Association, University of Massachusetts. –. Percus, Orin (). Antipresuppositions. In Ayumi Ueyama (ed.), Theoretical and Empirical Studies of Reference and Anaphora: Toward the Establishment of Generative Grammar as an Empirical Science. Report of the Grant-in-Aid for Scientific Research (B), Project No. , Supported by Japan Society for the Promotion of Science, Kyushu University. –. Pérez-Leroux, Ana T. (). The expression of genericity. In Jeffrey Lidz, William Snyder, and Joe Pater (eds.), Handbook of Developmental Linguistics. Oxford: Oxford University Press. –. Pérez-Leroux, Ana T., Munn, Alan, Schmitt, Cristina, and DeIrish, Michelle (). Learning definite determiners: genericity and definiteness in English and Spanish. In Alejna Brugos, Linnea Micciulla, and Christine E. Smith (eds.), Supplement to the Proceedings of the th Boston University Conference on Language Development. Boston, MA: Boston University Press. http:// www.bu.edu/bucld/proceedings/supplement/vol/ Perner, Josef and Ruffman, Ted (). Infants’ insight into the mind: how deep? Science, (): –. doi: ./science. Perner, Josef, Stummer, Sandra, Sprung, Manuel, and Doherty, Martin (). Theory of mind finds its Piagetian perspective: why alternative naming comes with understanding belief. Cognitive Development, : –. Perner, Josef, Brandl, Johannes L., and Garnham, Alan (). What is a perspective problem? Developmental issues in understanding belief and dual identity. Facta Philosophica, : –. Perner, Josef, Rendl, Bibiane, and Garnham, Alan (). Objects of desire, thought, and reality: problems of anchoring discourse referents in development. Mind and Language, : –. Perry, Clint J. and Barron, Andrew B. (). Honey bees selectively avoid difficult choices. Proceedings of the National Academy of Sciences of the United States of America, (): –. doi: ./pnas. Peters, Stanley (). A truth-conditional formulation of Karttunen’s account of presupposition. Synthese, : –. Peterson, Candida C. (). Development of social-cognitive and communication skills in children born deaf. Scandinavian Journal of Psychology, (): –. doi: ./j.-...x Peterson, Donald M. and Bowler, Dermot M. (). Counterfactual reasoning and false belief understanding in children with autism. Autism, (): –. doi: ./ Pexman, Penny M. (). It’s fascinating research: the cognition of verbal irony. Current Directions in Psychological Science, (): –. Pexman, Penny M. and Olineck, Kara M. (). Understanding irony: how do stereotypes cue speaker intent? Journal of Language and Social Psychology, (): –. doi: ./ X Philip, William (). Spreading in the Acquisition of Universal Quantifiers. In Dawn Bates (ed.), Proceedings of the th West Coast Conference on Formal Linguistics. –. Stanford, CA: CSLI Publications. Philip, William C. H. (). Event quantification in the acquisition of universal quantification. PhD thesis. University of Massachusetts, Amherst.
Philip, William (). Acquiring knowledge of universal quantification. In Jill G. de Villiers and Thomas Roeper (eds.), Handbook of Generative Approaches to Language Acquisition. Dordrecht: Springer. –. Phillips, Colin, Kazanina, Nina, and Abada, Shani H. (). ERP effects of the processing of syntactic long-distance dependencies. Cognitive Brain Research, (): –. Philips, Susan Urmston (). Some sources of cultural variability in the regulation of talk. Language in Society, (): –. Piaget, Jean and Inhelder, Bärbel (). The Child’s Conception of Space. New York: Norton. Piantadosi, Steven T., Tily, Harry, and Gibson, Edward (). Word lengths are optimized for efficient communication. Proceedings of the National Academy of Sciences of the United States of America, (): –. Piantadosi, Steven T., Tily, Harry, and Gibson, Edward (). The communicative function of ambiguity in language. Cognition, (): –. Pickering, Martin J. and Majid, Asifa (). What are implicit causality and consequentiality? Language and Cognitive Processes, : –. Pickering, Martin J., McElree, Brian, Frisson, Steven, Chen, Lillian, and Traxler, Matthew J. (). Underspecification and aspectual coercion. Discourse Processes, (): –. Pierce, Charles S. (). Philosophical Writings of Peirce, ed. Justus Buchler. New York: Dover. Pierce, Russell S., MacLaren, Rick, and Chiappe, Dan (). The role of WM in the metaphor interference effect. Psychonomic Bulletin and Review, (): –. doi: ./PBR... Pierrehumbert, Janet and Hirschberg, Julia (). The meaning of intonational contours in the interpretation of discourse. In Philip R. Cohen, Jerry Morgan, and Martha E. Pollack (eds.), Intentions in Communication. Cambridge, MA: MIT Press. –. Pietroski, Paul (). Events and Semantic Architecture. Oxford: Oxford University Press. Pighin, Stefania and Bonnefon, Jean-François (). Facework and uncertain reasoning in health communication. Patient Education and Counseling, : –. doi: ./j.pec... Pighin, Stefania, Bonnefon, Jean-François, and Savadori, Lucia (). Overcoming number numbness in prenatal risk communication. Prenatal Diagnosis, : –. doi: ./pd. Pijnacker, Judith, Hagoort, Peter, Buitelaar, Jan, Teunisse, Jan-Pieter, and Geurts, Bart (). Pragmatic inferences in high-functioning adults with autism and Asperger syndrome. Journal of Autism and Developmental Disorders, : –. Piñango, Maria Mercedes, Zurif, Edgar, and Jackendoff, Ray (). Real-time processing implications of enriched composition at the syntax-semantics interface. Journal of Psycholinguistic Research, (): –. doi: ./A: Piñango, Maria Mercedes, Winnick, Aaron, Ullah, Rashad, and Zurif, Edgar (). Time-course of semantic composition: the case of aspectual coercion. Journal of Psycholinguistic Research, (): –. doi: ./s---z Pinker, Steven (). Language Learnability and Language Development (with new commentary by the author). Cambridge, MA: Harvard University Press. Pinker, S., Nowak, M. A., & Lee, J. J. (). The logic of indirect speech. Proceedings of the National Academy of Sciences, (): –. Piñon, Christopher (). Achievements in an event semantics. In Aaron Lawson (ed.), Proceedings of Semantics and Linguistic Theory . Ithaca, NY: Cornell University Press. –. Poesio, Massimo (). Discourse interpretation and the scope of operators. PhD thesis. University of Rochester. Pogue, Amanda, Kurumada, Chigusa, and Tanenhaus, Michael K. (). Talker-specific generalization of pragmatic inferences based on under- and over-informative prenominal adjective use. Frontiers in Psychology, : . doi: ./fpsyg.. Polanyi, Livia (). A formal model of the structure of discourse. Journal of Pragmatics, : –. doi: ./-()- Politzer-Ahles, Stephen and Fiorentino, Robert (). The realization of scalar inferences: context sensitivity without processing cost. PLoS ONE, (): e. doi: ./journal.pone.
Politzer-Ahles, Stephen and Gwilliams, Laura (). Involvement of prefrontal cortex in scalar implicatures: evidence from magnetoencephalography. Language, Cognition and Neuroscience, (): –. doi: ./.. Pollard, Carl and Yasavul, Murat (). Anaphoric it-clefts: the myth of exhaustivity. In Ross Burkholder, Carlos Cisneros-Juarez, Emily Coppess et al. (eds.), Proceedings of the th Annual Meeting of the Chicago Linguistic Society. Chicago, IL: Chicago Linguistic Society. Pomerantz, Anita (). Agreeing and disagreeing with assessments: some features of preferred/ dispreferred turn shaped. In J. Maxwell Anderson and John Heritage (eds.), Structures of Social Action. Cambridge: Cambridge University Press. –. Poppels, Till and Levy, Roger (). Resolving quantity and informativeness implicature in indefinite reference. In Thomas Brochhagen, Floris Roelofsen, and Nadine Theiler (eds.), Proceedings of the th Amsterdam Colloquium Amsterdam: University of Amsterdam. –. Posner, Michael I. and Snyder, Charles R. R. (). Attention and cognitive control. In Robert L. Solso (ed.), Information Processing and Cognition. Hillsdale, NJ: Erlbaum. Potts, Christopher (). The Logic of Conventional Implicatures. Oxford: Oxford University Press. Potts, Christopher (). Goal-driven answers in the Cards dialogue corpus. In Nathan Arnett and Ryan Bennett (eds.), Proceedings of the th West Coast Conference on Formal Linguistics. Somerville, MA: Cascadilla Press. –. Potts, Christopher, Lassiter, Daniel, Levy, Roger, and Frank, Michael C. (). Embedded implicatures as pragmatic inferences under compositional lexical uncertainty. Journal of Semantics, (): –. doi: ./jos/ffv Pouscoulous, Nausicaa (). Metaphor—for adults only? Belgian Journal of Linguistics, : –. Pouscoulous, Nausicaa (). ‘The elevator’s buttocks’: metaphorical abilities in children. In Danielle Matthews (ed.), Pragmatic Development in First Language Acquisition. Amsterdam: John Benjamins. –. Pouscoulous, Nausicaa and Tomasello, Michael (submitted). Very early understanding of metaphors. Poster presented at the Meeting of the Society for Research in Child Development, Montreal. Pouscoulous, Nausicaa, Noveck, Ira A., Politzer, Guy, and Bastide, Anne (). A developmental investigation of processing costs in implicature production. Language Acquisition, (): –. Povinelli, Daniel J. and DeBlois, Sandra (). Young children’s (Homo sapiens) understanding of knowledge formation in themselves and others. Journal of Comparative Psychology, (): –. doi: ./-... Prado, Jérôme and Noveck, Ira A. (). How reaction time measures elucidate the matching bias and the way negations are processed. Thinking and Reasoning, (): –. doi: ./ Prasad, Rashmi, Dinesh, Nikhil, Lee, Alan et al. (). The Penn Discourse TreeBank .. Proceedings of the th International Conference on Language Resources and Evaluation. Marrakesh: Palais des Congres Mansour Eddahbi. Prasada, Sadeep and Dillingham, Elaine M. (). Principled and statistical connections in common sense conception. Cognition, : –. Prasada, Sadeep and Dillingham, Elaine M. (). Representation of principled connections: a window onto the formal aspect of common sense conception. Cognitive Science, : –. Prasada, Sadeep, Salajegheh, Anna, Bowles, Anita, and Poeppel, David (). Characterizing kinds and instances of kinds: ERP reflections. Language and Cognitive Processes, : –. Prasada, Sadeep, Khemlani, Sangeet, Leslie, Sarah-Jane, and Glucksberg, Sam (). Conceptual distinctions amongst generics. Cognition, : –. Premack, David and Woodruff, Guy (). Does the chimpanzee have a theory of mind? Behavioral and Brain Sciences, (): –. Priest, Graham (). The logic of paradox. Journal of Philosophical Logic, : –. Prieto, Pilar (). Intonational meaning. Wiley Interdisciplinary Reviews: Cognitive Science, : –. Prince, Ellen F. (). A comparison of it-clefts and Wh-clefts in discourse. Language, : –.
Prince, Ellen (). Toward a taxonomy of given-new information. In Peter Cole (ed.), Radical Pragmatics. New York: Academic Press. –. Prince, Ellen (). Fancy syntax and ‘shared knowledge’. Journal of Pragmatics, : –. Prince, Ellen (). The ZPG letter: subjects, definiteness, and information-status. In William C. Mann and Sandra A. Thompson (eds.), Discourse Description: Diverse Linguistic Analyses of a Fundraising Text. Philadelphia and Amsterdam: John Benjamins. –. Pruden, Shannon M., Hirsh-Pasek, Kathy, Maguire, Mandy J., Meyers, M. M, and Golinkoff, Roberta Michnick (). Foundations of verb learning: infants form categories of path and manner in motion events. In Alejna Brugos, Linnea Micciulla, and Christine E. Smith (eds.), Proceedings of the th Annual Boston University Conference on Language Development. Somerville, MA: Cascadilla Press. –. Pullum, Geoffrey and Scholz, Barbara (). Empirical assessment of stimulus poverty arguments. Linguistic Review, : –. Pulverman, Rachel, Golinkoff, Roberta Michnick, Hirsh-Pasek, Kathy, and Buresh, Jennifer Sootsman (). Infants discriminate manners and paths in non-linguistic dynamic events. Cognition, (): –. doi: ./j.cognition... Pulverman, Rachel, Song, Lulu, Hirsh-Pasek, Kathy, Pruden, Shannon M., and Golinkoff, Roberta Michnick (). Preverbal infants’ attention to manner and path: foundations for learning relational terms. Child Development, (): –. doi: ./cdev. Pulvermüller, Friedemann (). Words in the brain’s language. Behavioral and Brain Sciences, : –. Pulvermüller, Friedemann (). Brain perspective on language mechanisms: from discrete neuronal ensembles to serial order. Progress in Neurobiology, : –. Pustejovsky, James (). The syntax of event structure. Cognition, (): –. Pustejovsky, James (a). The Generative Lexicon. Cambridge, MA: MIT Press. Pustejovsky, James (b). The generative lexicon. Computational Linguistics, (): –. Pyers, Jennie E., Shusterman, Anna, Senghas, Ann, Spelke, Elizabeth S, and Emmorey, Karen (). Evidence from an emerging sign language reveals that language supports spatial cognition. Proceedings of the National Academy of Sciences of the United States of America, (): –. doi: ./pnas. Pyykkönen, Pirita and Järvikivi, Juhani (). Activation and persistence of implicit causality information in spoken language comprehension. Experimental Psychology, (): –. doi: ./-/a Pylkkänen, Liina (). Introducing Arguments. Cambridge, MA: MIT Press. Pylkkänen, Liina and McElree, Brian (). The syntax-semantic interface: on-line composition of sentence meaning. In Matthew J. Traxler and Morton Ann Gernsbacher (eds.), Handbook of Psycholinguistics, nd edn. New York: Elsevier. –. Pylkkänen, Liina and McElree, Brian (). An MEG study of silent meaning. Journal of Cognitive Neuroscience, (): –. Pylkkänen, Liina, Llinás, Rodolfo, and Murphy, Gregory, L. (). The representation of polysemy: MEG evidence. Journal of Cognitive Neuroscience, (): –. Pylkkänen, Liina, Brennan, Jonathan, and Bemis, Douglas K. (). Grounding the cognitive neuroscience of semantics in linguistic theory. Language and Cognitive Processes, (): –. Qing, Ciyang and Franke, Michael (a). Gradable adjectives, vagueness, and optimal language use: a speaker-oriented model. In Todd Snider, Sarah D’Antonio, and Mia Weigand (eds.), Proceedings of the th Semantics and Linguistic Theory Conference. Washington, DC: Linguistic Society of America. –. Qing, Ciyang and Franke, Michael (b). Meaning and use of gradable adjectives: Formal modeling meets empirical data. In Paul Bello, Marcello Guarini, Marjorie McShane, and Brian Scassellati (eds.), Proceedings of the th Annual Conference of the Cognitive Science Society. Austin, TX: Cognitive Science Society. –. Quine, Willard van Ormen (). Ontological relativity. Journal of Philosophy, (): –.
Quinn, Paul C. (). The categorization of above and below spatial relations by young infants. Child Development, (): –. doi: ./ Quinn, Paul C. (). Spatial representation by young infants: categorization of spatial relations or sensitivity to a crossing primitive? Memory and Cognition, (): –. doi: ./BF Quinn, Paul C., Cummins, Maggie, Kase, Jennifer, Martin, Erin, and Weissman, Sheri (). Development of categorical representations for above and below spatial relations in - to -month-old infants. Developmental Psychology, (): –. doi: ./-... Quinn, Paul C., Adams, Adria, Kennedy, Erin, Shettler, Lauren, and Wasnik, Amanda (). Development of an abstract category representation for the spatial relation between in - to -month-old infants. Developmental Psychology, (): –. doi: ./-... Rabagliati, Hugh and Snedeker, Jesse (). The truth about chickens and bats: ambiguity avoidance distinguishes types of polysemy. Psychological Science, (): –. doi: ./ Rabagliati, Hugh, Marcus, Gary F., and Pylkkänen, Liina (). Shifting senses in lexical semantic development. Cognition, : –. Rabagliati, Hugh, Marcus, Gary F., and Pylkkänen, Liina (). Rules, radical pragmatics and restrictions on regular polysemy. Journal of Semantics, (): –. doi: ./jos/ffr Rabovsky, Milena and McRae, Ken (). Simulating the N ERP component as semantic network error: insights from a feature-based connectionist attractor model of word meaning. Cognition, : –. Radó, Janina and Bott, Oliver (). Underspecified representations of quantifier scope? In Maria Aloni, Vadim Kimmelman, Floris Roelofsen, Galit Weidman Sassoon, Katrin Schulz, and Matthijs Westera (eds.), Logic, Language and Meaning: th Amsterdam Colloquium. Springer. Raffman, Diana (). Vagueness without paradox. Philosophical Review, : –. Raffman, Diana (). Vagueness and context relativity. Philosophical Studies, : –. Raffman, Diana (). Unruly Words: A Study of Vague Language. Oxford: Oxford University Press. Raizen, Adina, Vergis, Nikos, and Christianson, Kiel (). Using eye-tracking to examine the reading of texts containing taboo words. In Marina Terkourafi (ed.), Interdisciplinary Perspectives on (Im)politeness. Amsterdam: John Benjamins. –. Rakhlin, Natalia (). A new pragmatic account of quantifier-spreading. Nanzan Linguistics, : –. Rakoczy, Hannes (). In defense of a developmental dogma: children acquire propositional attitude folk psychology around age . Synthese, (): –. Rakoczy, Hannes and Tomasello, Michael (). Done wrong or said wrong? Young children understand the normative directions of fit of different speech acts. Cognition, (): –. Ramiro, Christian, Malt, Barbara C., Srinivasan, Mahesh and Xu, Yang (). Mental algorithms in the historical emergence of word meanings. Proceedings of the th Annual Conference of the Cognitive Science Society. London: Cognitive Science Society. Ramsey, Frank Plumpton (). The Foundations of Mathematics and Other Logical Essays. London: Routledge and Kegan Paul. Rapp, Alexander M., Erb, Michael, Grodd, Wolfgang et al. (). Neural correlates of irony comprehension: the role of schizotypal personality traits. Brain and Language, (): –. doi: ./j. bandl... (). Neural correlates of metonymy resolution. Brain and Language, : –. Rappaport Hovav, Malka (). Lexicalized meaning and the internal. In Susan D. Rothstein (ed.), Theoretical and Crosslinguistic Approaches to the Semantics of Aspect. Amsterdam: John Benjamins. –. Rappaport Hovav, Malka and Levin, Beth (). Building verb meanings. In Miriam Butt and Wilhem Geuder (eds.), The Projection of Arguments: Lexical and Compositional Factors. Stanford, CA: CSLI Publications. –. Rappaport Hovav, Malka and Levin, Beth (). The English dative alternation: the case for verb sensitivity. Journal of Linguistics, (): –.
Rappaport Hovav, Malka and Levin, Beth (). Reflections on manner/result complementarity. In Malka Rappaport Hovav, Edit Doron, and Ivy Sichel (eds.), Syntax, Lexical Semantics, and Event Structure. Oxford: Oxford University Press. –. Rayner, Keith (). Eye movements in reading and information processing: years of research. Psychological Bulletin, : –. Rayner, Keith (). Eye movements and attention in reading, scene perception, and visual search. Quarterly Journal of Experimental Psychology, : –. Rayner, Keith and Duffy, Susan A. (). Lexical complexity and fixation times in reading: effects of word-frequency, verb complexity, and lexical ambiguity. Memory and Cognition, (): –. Rayner, Keith and Well, Arnold D. (). Effects of contextual constraint on eye movements in reading: a further examination. Psychonomic Bulletin and Review, : –. doi: ./BF Rayner, Keith, Warren, Tessa, Juhasz, Barbara J., and Liversedge, Simon P. (). The effects of plausibility on eye movements in reading. Journal of Experimental Psychology: Learning, Memory and Cognition, (): –. doi: ./-... Reagan, Robert T., Mosteller, Frederick, and Youtz, Cleo (). Quantitative meanings of verbal probability expressions. Journal of Applied Psychology, : –. doi: ./-... Recanati, François (). The alleged priority of literal interpretation. Cognitive Science, : –. Recanati, François (). Literal Meaning. Cambridge: Cambridge University Press. Reed, Marjorie A., Pien, Diana L., and Rothbart, Mary K. (). Inhibitory self-control in preschool children. Merrill-Palmer Quarterly, : –. Regel, Stefanie, Gunter, Thomas C., and Friederici, Angela D. (). Isn’t it ironic? An electrophysiological exploration of figurative language processing. Journal of Cognitive Neuroscience, (): –. doi: ./jocn.. Regier, Terry and Zheng, Mingyu (). Attention to endpoints: a cross-linguistic constraint on spatial meaning. Cognitive Science, : –. doi: ./ Regier, Terry, Kemp, Charles, and Kay, Paul (). Word meanings across languages support efficient communication. In Brian MacWhinney and William O’Grady (eds.), Handbook of Language Emergence. Hoboken, NJ: Wiley. –. Reichle, Erik D., Carpenter, Patricia A., and Just, Marcel Adam (). The neural bases of strategy and skill in sentence-picture verification. Cognitive Psychology, (): –. doi: ./ cogp.. Reimer, Marga and Bezuidenhout, Anne (eds.) (). Descriptions and Beyond. Oxford: Oxford University Press. Reinhart, Tanya (). Pragmatics and linguistics: an analysis of sentence topics. Philosophica, : –. Reinhart, Tanya (). Anaphora and Semantic Interpretation. Chicago: University of Chicago Press. Reinhart, Tanya (). Quantifier scope: how labor is divided between QR and choice functions. Linguistics and Philosophy, : –. Reinhart, Tanya (). The processing cost of reference set computation: acquisition of stress shift and focus. Language Acquisition, (): –. Reinhart, Tanya (). Interface Strategies: Optimal and Costly Computations. Cambridge, MA: MIT Press. Reisman, Karl (). Contrapuntal conversations in an Antiguan village. In Richard Bauman and Joel Sherzer (eds.), Language in its Social Setting. Cambridge: Cambridge University Press. Reyna, Valerie F. (). The language of possibility and probability: effects of negation on meaning. Memory and Cognition, : –. doi: ./bf Rialland, Annie and Robert, Stéphane (). The intonational system of Wolof. Linguistics, : –. Rickheit, Gert and Habel, Christopher (eds.) (). Mental Models in Discourse Processing and Reasoning. Amsterdam: North Holland. Rieser, John J. (). Spatial orientation of six-month-old infants. Child Development, (): –. doi: ./
Riest, Carina, Jorschick, Annett B., and De Ruiter, Jan P. (). Anticipation in turn-taking: mechanisms and information sources. Frontiers in Psychology, : . doi: ./fpsyg.. Rigal, Robert (). Right-left orientation: development of correct use of right and left terms. Perceptual and Motor Skills, (): –. doi: ./pms.... Rigal, Robert (). Right-left orientation, mental rotation, and perspective-taking: when can children imagine what people see from their own viewpoint? Perceptual and Motor Skills, (): –. doi: ./pms.... Riggs, Kevin J. and Peterson, Donald M. (). Counterfactual and mental state inferences in preschool children. In Peter Mitchell and Kevin J. Riggs (eds.), Children’s Reasoning and the Mind. Hove: Psychology Press. Riggs, Kevin J., Peterson, Donald M., Robinson, Elizabeth J., and Mitchell, Peter (). Are errors in false belief tasks symptomatic of a broader difficulty with counterfactuality? Cognitive Development, : –. Ripley, David (). Contradictions at the borders. In Rick Nouwen, Robert van Rooij, Uli Sauerland, and Hans-Christian Schmitz (eds.), Vagueness in Communication. Heidelberg: Springer. –. Ripley, David (). Vagueness is a kind of conflation. Logic and Logical Philosophy, : –. Rips, Lance J. and Turnbull, William (). How big is big? Relative and absolute properties in memory. Cognition, : –. Ritov, Ilana and Baron, Jonathan (). Reluctance to vaccinate: omission bias and ambiguity. Journal of Behavioral Decision Making, : –. Roberson, Debi, Davidoff, Jules, Davies, Ian R. L., and Shapiro, Laura R. (). Color categories: evidence for the cultural relativity hypothesis. Cognitive Psychology, (): –. Roberts, Craige (). Information structure in discourse: towards an integrated formal theory of pragmatics. In Jae-Hak Yoon and Andreas Kathol (eds.), Ohio State University Working Papers in Linguistics, vol. . Columbus, OH: Ohio State University Publications. –. Roberts, Craige (). Focus, information flow, and Universal Grammar. In Peter Culicover and Louise McNally (eds.), Syntax and Semantics, Vol. : The Limits of Syntax. New York: Academic Press. –. Roberts, Craige (a). Information structure in discourse: towards an integrated formal theory of pragmatics. Semantics and Pragmatics, (): –. doi: ./sp.. Roberts, Craige (b). Topics. In Klaus von Heusinger, Claudia Maienborn, and Paul Portner (eds.), Semantics: An International Handbook of Natural Language Meaning, vol. . Berlin: De Gruyter. –. Roberts, Craige (). Coherence and anaphora. Presented at the Anaphora and Coherence Workshop at NASSLLI , Rutgers, New Brunswick, NJ. Robinson, E. J., Champion, H., and Mitchell, Peter (). Children’s ability to infer utterance veracity from speaker informedness. Developmental Psychology, (): –. doi: ./ -... Rodd, Jennifer M., Gaskell, M. Gareth, and Marslen-Wilson, William D. (). Making sense of semantic ambiguity: semantic competition in lexical access. Journal of Memory and Language, (): –. Rodd, Jennifer M., Gaskell, M. Gareth, and Marslen-Wilson, William D. (). Modelling the effects of semantic ambiguity in word recognition. Cognitive Science, (): –. doi: ./ scog_ Roeper, Thomas and De Villiers, Jill (). Emergence of bound variable structures. In Eric J. Reuland and Werner Abraham (eds.), Knowledge and Language: From Orwell’s Problem to Plato’s Problem. Dordrecht: Kluwer. –. Roeper, Thomas, Strauss, Uri, and Pearson, Barbara Zurer (). The acquisition path of quantifiers: two kinds of spreading. In Tanja Heizmann (ed.), Current Issues in Language Acquisition, University of Massachusetts Occasional Papers, . Amherst, MA: University of Massachusetts. Roeper, Thomas, Pearson, Barbara Zurer, and Grace, Margaret Grace (). Quantifier spreading is not distributive. In Nick Danis, Kate Mesh, and Hyunsuk Sung (eds.), Proceedings of the th
Annual Boston University Conference on Language Development. Somerville, MA: Cascadilla Press. –. Roese, Neal J., Sanna, Lawrence J., and Galinsky, Adam D. (). The mechanics of imagination: automaticity and control in counterfactual thinking. In Ran R. Hassin, James S. Uleman, and John A. Bargh (eds.), The New Unconscious. New York: Oxford University Press. –. Rohde, Hannah (). Coherence-driven effects in sentence and discourse processing. PhD thesis. University of California, San Diego. Rohde, Hannah and Kehler, Andrew (). Grammatical and information-structural influences on pronoun production. Language, Cognition and Neuroscience, (): –. doi: ./ .. Rohde, Hannah, Kehler, Andrew, and Elman, Jeffrey L. (). Pronoun interpretation as a side effect of discourse coherence. In Danielle McNamara and Greg Trafton (eds.), Proceedings of the th Annual Conference of the Cognitive Science Society. Austin, TX: Cognitive Science Society. –. Rohde, Hannah, Seyfarth, Scott, Clark, Brady, Jäger, Gerhard, and Kaufmann, Stefan (). Communicating with cost-based implicature: a game-theoretic approach to ambiguity. In Sarah BrownSchmidt, Jonathan Ginzburg, and Staffan Larsson (eds.), Proceedings of SemDial (SeineDial): the th Workshop on the Semantics and Pragmatics of Dialogue. Paris: Université Paris-Diderot. –. Romoli, Jacopo (). Soft but strong: neg-raising, soft triggers, and exhaustification. PhD thesis. Harvard University. Romoli, Jacopo (). A problem for the structural characterization of alternatives. Snippets, : –. Romoli, Jacopo (). The presuppositions of soft triggers are obligatory scalar implicatures. Journal of Semantics, (): –. doi: ./jos/fft Romoli, Jacopo and Schwarz, Florian (). An experimental comparison between presuppositions and indirect scalar implicatures. In Florian Schwarz (ed.), Experimental Perspectives on Presuppositions. Cham: Springer International Publishing. –. Romoli, Jacopo, Sudo, Yasutada, and Snedeker, Jesse (). An experimental investigation of presupposition projection in conditional sentences. In Neil Ashton, Anca Chereches, and David Lutz (eds.), Proceedings of Semantics and Linguistic Theory . Ithaca, NY: CLC Publications. –. Romoli, Jacopo, Khan, Manizeh, Snedeker, Jesse, and Sudo, Yasutada (). Resolving temporary referential ambiguity using presupposed content. In Florian Schwarz (ed.), Experimental Perspectives on Presuppositions. Cham: Springer International Publishing. –. Rooth, Mats (). Association with focus. PhD thesis. University of Massachusetts, Amherst. Rooth, Mats (). A theory of focus interpretation. Natural Language Semantics, (): –. Rooth, Mats (). Alternative semantics. In Caroline Féry and Shinichiro Ishihara (eds.), Oxford Handbook of Information Structure. Oxford: Oxford University Press. –. Rosa, Elise C. and Arnold, Jennifer E. (). Predictability affects production: thematic roles can affect reference form selection. Journal of Memory and Language, : –. doi: ./j. jml... Rosch, Eleanor (). Natural categories. Cognitive Psychology, : –. Rosch, Eleanor (). Cognitive representations of semantic categories. Journal of Experimental Psychology: General, : –. Rosch, Eleanor and Mervis, Carolyn B. (). Family resemblance: studies in the internal structure of categories. Cognitive Psychology, : –. Ross, John Robert (). Act. In Donald Davidson and Gilbert Harman (eds.), Semantics in Natural Language. Dordrecht: Reidel. –. Rothbart, Mary K. and Posner, Michael I. (). Temperament and the development of selfregulation. In Lawrence C. Hartlage and Cathy F. Telzrow (eds.), The Neuropsychology of Individual Differences: A Developmental Perspective. New York: Plenum. –. Rothstein, Susan (). Adverbial quantification over events. Natural Language Semantics, (): –.
Rothstein, Susan (). Predicates and Their Subjects. Amsterdam: Springer Netherlands. Rotstein, Carmen and Winter, Yoad (). Total adjectives vs. partial adjectives: scale structure and higher-order modifiers. Natural Language Semantics, : –. Rowland, Caroline F., Pine, Julian M., Lieven, Elena V., and Theakston, Anna L. (). Determinants of acquisition order in wh-questions: re-evaluating the role of caregiver speech. Journal of Child Language, (): –. Rubio-Fernández, Paula (). Suppression in metaphor interpretation: differences between meaning selection and meaning construction. Journal of Semantics, : –. doi: ./jos/m Rubio-Fernández, Paula (). How redundant are redundant color adjectives? An efficiency-based analysis of color overspecification. Frontiers in Psychology, : . doi: ./fpsyg.. Rubio-Fernández, Paula (submitted). Saying too much can be efficient: a reference production/ comprehension study. Rubio-Fernández, Paula and Geurts, Bart (). How to pass the false-belief task before your fourth birthday. Psychological Science, (): –. Rubio-Fernández, Paula and Geurts, Bart (). Don’t mention the marble! The role of attentional processes in false-belief tasks. Review of Philosophy and Psychology, (): –. Rubio-Fernández, Paula and Grassmann, Susanne (). Metaphors as second labels: difficult for preschool children. Journal of Psycholinguistic Research, (): –. doi: ./s--y Rubio-Fernández, Paula, Carston, Robyn and Wearing, Catherine (). Metaphor and hyperbole: testing the continuity hypothesis. Metaphor and Symbol, (): –. Rubio-Fernández, Paula, Cummins, Chris, and Tian, Ye (). Are single and extended metaphors processed differently? A test of two Relevance Theoretic accounts. Journal of Pragmatics, : –. Rubio-Fernández, Paula, Geurts, Bart, and Cummins, Chris (). Is an apple like a fruit? A study on comparison and categorisation statements. Review of Philosophy and Psychology, (): –. doi: ./s--- Ruffman, Ted (). To belief or not belief: children’s theory of mind. Developmental Review, (): –. doi: ./j.dr... Rugg, Michael D. (). The effects of semantic priming and word repetition on event-related potentials. Psychophysiology, (): –. Rumelhart, David E. and McClelland, James L. (). An interactive activation model of context effects in letter perception : the contextual enhancement effect and some tests and extensions of the model. Psychological Review, (): –. Rundblad, Gabriella and Annaz, Dagmara (a). Development of metaphor and metonymy comprehension: receptive vocabulary and conceptual knowledge. British Journal of Developmental Psychology, (): –. Rundblad, Gabriella and Annaz, Dagmara (b). The atypical development of metaphor and metonymy comprehension in children with autism. Autism, (): –. doi: ./ Russell, Benjamin (). Probabilistic reasoning and the computation of scalar implicatures. PhD thesis. Brown University. Russell, Bertrand (). On denoting. Mind, : –. Russell, Bertrand (). Human Knowledge: Its Scope and Limits. New York: Simon and Schuster. Ryskin, Rachel A., Kurumada, Chigusa, and Brown-Schmidt, Sarah (). Information integration in online modulation of pragmatic inferences during language comprehension. Poster presented at the th Annual Meeting of the Psychonomic Society, Boston, MA. Saba, Walid S. and Corriveau, Jean-Pierre (). Plausible reasoning and the resolution of quantifier scope ambiguities. Studia Logica, (): –. doi: ./A:. Sachs, Jacqueline Strunk (). Recopition memory for syntactic and semantic aspects of connected discourse. Perception and Psychophysics, (): –. doi: ./BF Sacks, Harvey, Schegloff, Emanuel A., and Jefferson, Gail (). A simplest systematics for the organization of turn-taking for conversation. Language, : –.
Sacks, Harvey, Schegloff, Emanuel A., and Jefferson, Gail (). A simplest systematics for the organization of turn taking for conversation. In J. Schenkein (Ed.), Studies in the Organization of Conversational Interaction. New York: Academic Press. –. Saddy, Douglas, Drenhaus, Heiner, and Frisch, Stefan (). Processing polarity items: Contrastive licensing costs. Brain and Language, (): –. Sæb, Kjell Johan (). Conversational contrast and conventional parallel: topic implicatures and additive presuppositions. Journal of Semantics, : –. Sag, Ivan (). Deletion and logical form. PhD thesis. MIT, Cambridge, MA. Salverda, Anne Pier and Tanenhaus, Michael K. (). The visual world paradigm. In Annette M. B. de Groot and Peter Hagoort (eds.), Research Methods in Psycholinguistics: A Practical Guide. Malden, MA: Wiley Blackwell. –. Salverda, Anne Pier, Brown, Meredith, and Tanenhaus, Michael K. (). A goal-based perspective on eye movements in visual world studies. Acta Psychologica, (): –. doi: ./j. actpsy... Sanders, Ted J. M., Spooren, Wilbert P. M., and Noordman, Leo G. M. (). Toward a taxonomy of coherence relations. Discourse Processes, : –. Sanford, Alison J. S., Sanford, Anthony J., Molle, Jo, and Emmott, Catherine (). Shallow processing and attention capture in written and spoken discourse. Discourse Processes, (): –. doi: ./sdp_ Sanford, Anthony J. (). Defining embodiment in understanding. In Manuel de Vega, Arthur M. Glenberg, and Arthur C. Graesser (eds.), Symbols, Embodiment and Meaning. Oxford: Oxford University Press. –. Sanford, Anthony J. and Garrod, Simon C. (). Understanding Written Language. New York: Wiley. Sanford, Anthony J. and Graesser, Arthur C. (). Shallow processing and underspecification. Discourse Processes, (): –. doi: ./sdp_ Sanford, Anthony J. and Sturt, Patrick (). Depth of processing in language comprehension: not noticing the evidence. Trends in Cognitive Sciences, (): –. Sanford, Anthony J., Moar, K., and Garrod, Simon C. (). Proper names as controllers of discourse focus. Language and Speech, , –. doi: ./ Santamaría, Carlos, Espino, Orlando, and Byrne, Ruth M. J. (). Counterfactual and semifactual conditionals prime alternative possibilities. Journal of Experimental Psychology: Learning, Memory and Cognition, : –. doi: ./-... Sassoon, Galit (). A slightly modified economy principle: stable properties have nonstable standards. Talk at Workshop on Degree Semantics and its Interfaces, Utrecht. Sassoon, Galit (). The double nature of negative antonymy. In Ana Aguilar Guevara, Anna Chernilovskaya, and Rick Nouwen (eds.), Proceedings of Sinn und Bedeutung . Cambridge, MA: MIT Working Papers in Linguistics. –. Sassoon, Galit W. (a). Vagueness, Gradability and Typicality. Leiden: Brill. Sassoon, Galit (b). A typology of multidimensional adjectives. Journal of Semantics, : –. Sauerland, Uli (). A new semantics for number. In Robert B. Young and Yuping Zhou (eds.), Proceedings of the th Semantics and Linguistic Theory Conference. Ithaca, NY: CLC Publications. –. Sauerland, Uli (). Scalar implicatures in complex sentences. Linguistics and Philosophy, (): –. doi: ./B:LING...db Sauerland, Uli (). Implicated presuppositions. In Anita Steube (ed.), The Discourse Potential of Underspecified Structures. Berlin: De Gruyter. –. Sauerland, Uli (). Vagueness in language: the case against fuzzy logic revisited. In Petr Cintula, Christian G. Fermüller, Lluís Godo, and Petr Hájek (eds.), Understanding Vagueness: Logical, Philosophical, and Linguistic Perspectives. London: College Publications. Sauerland, Uli (a). The computation of scalar implicatures: pragmatic, lexical or grammatical. Language and Linguistics Compass, : –. Sauerland, Uli (b). Where does the Strongest Meaning Hypothesis apply? Snippets, : –.
Sauerland, Uli (to appear). Implicated presuppositions. In Anita Steube (ed.), Sentence and Context. Berlin: De Gruyter. –. Sauerland, Uli and Gotzner, Nicole (). Familial sinistrals avoid exact numbers. PLoS ONE, : e. Sauerland, Uli and Schumacher, Petra (). Pragmatics: theory and experiment growing together. Linguistische Berichte, : –. Sauerland, Uli and Stateva, Penka (). Scalar vs. epistemic vagueness: evidence from approximators. In Masayuki Gibson and Tova Friedman (eds.), Proceedings of the th Semantics and Linguistic Theory Conference. Ithaca, NY: CLC Publications. –. Sauerland, Uli and Stateva, Penka (). Two types of vagueness. In Paul Egré and Nathan Klinedinst (eds.), Vagueness and Language Use. Basingstoke: Palgrave Macmillan. Sauerland, Uli and Yatsushiro, Kazuko (eds.) (). Semantics and Pragmatics: From Experiment to Theory. Berlin: Palgrave Macmillan. Sauerland, Uli, Andersen, Jan, and Yatsushiro, Kazuko (). The plural is semantically unmarked. In Stephan Kepser and Marga Reis (eds.), Linguistic Evidence: Empirical, Theoretical, and Computational Perspectives. Berlin: De Gruyter. –. Sauermann, Antje, Filik, Ruth, and Paterson, Kevin B. (). Processing contextual and lexical cues to focus: evidence from eye movements in reading. Language and Cognitive Processes, (): –. doi: ./.. Saul, Jennifer M. (). What is said and psychological reality: Grice’s project and relevance theorists’ criticisms. Linguistics and Philosophy, : –. Saul, Jennifer (). Are generics especially pernicious? Inquiry. Advance online publication, doi: ./X.. Saur, Eva-Maria (). Clefts: discourse function and the nature of exhaustivity violation effects. Master’s thesis. University of Potsdam. Saxe, Rebecca and Kanwisher, Nancy (). People thinking about thinking people: the role of the temporo-parietal junction in ‘theory of mind’. NeuroImage, (): –. Saxe, Rebecca and Powell, Lindsey J. (). It’s the thought that counts: specific brain regions for one component of theory of mind. Psychological Science, (): –. Saxe, Rebecca and Wexler, Anna (). Making sense of another mind: the role of the right temporoparietal junction. Neuropsychologia, (): –. doi: ./j.neuropsychologia... Saxe, Rebecca, Xiao, D. K., Kovacs, G., Perrett, D. I., and Kanwisher, Nancy (). A region of right posterior superior temporal sulcus responds to observed intentional actions. Neuropsychologia, (): –. doi: ./j.neuropsychologia... Sayeed, Asad (). Representing the effort in resolving ambiguous scope. Presentation at Sinn und Bedeutung , Tübingen. Scha, Remko (). Distributive, collective and cumulative quantification. In Jeroen A. G. Groenendijk, Theo M. V. Janssen, and Martin J. B. Stokhof (eds.), Formal Methods in the Study of Language. Amsterdam: Mathematical Centre Tracts . –. Schafer, Amy (). Focus in relative clause construal. Language and Cognitive Processes, (–): –. Schafer, Amy, Carlson, Katy, Clifton, Charles, Jr., and Frazier, Lyn (). Focus and the interpretation of pitch accent: disambiguating embedded questions. Language and Speech, (): –. doi: ./ Schaffer, Deborah (). The role of intonation as a cue to turn taking in conversation. Journal of Phonetics, (): –. Schegloff, Emanuel A. (). Sequencing in conversational openings. American Anthropologist, (): –. Schegloff, Emanuel A. (). Discourse as an interactional achievement: some uses of ‘uh huh’ and other things that come between sentences. In Deborah Tannen (ed.), Analyzing Discourse: Text and Talk. Washington, DC: Georgetown University Press. –.
Schegloff, Emanuel A. (). Reflections on quantification in the study of conversation. Research on Language and Social Interaction, (): –. Schegloff, Emanuel A. (). Overlapping talk and the organization of turn-taking for conversation. Language in Society, (): –. Schein, Barry (). Events and the semantic content of thematic relations. In Gerhard Preyer and Georg Peter (eds.), Logical Form, Language and Semantic Content: On Contemporary Developments in the Philosophy of Language and Linguistics. Oxford: Oxford University Press. –. Schein, Barry (). Event semantics. In Gillian K. Russell and Delia Graff Fara (eds.), Routledge Companion to Philosophy of Language. Oxford: Routledge. –. Schlenker, Philippe (a). Be articulate: a pragmatic theory of presupposition projection. Theoretical Linguistics, (): –. Schlenker, Philippe (b). Presupposition projection: explanatory strategies. Theoretical Linguistics, (): –. Schlenker, Philippe (). Local contexts. Semantics and Pragmatics, (): –. doi: ./sp.. Schlenker, Philippe (a). Local contexts and local meanings. Philosophical Studies, (): –. Schlenker, Philippe (b). Presuppositions and local contexts. Mind, (): –. Schlenker, Philippe (). Presupposition projection: two theories of local contexts part II. Language and Linguistics Compass, (): –. doi: ./j.-X...x Schlenker, Philippe (). Maximize presupposition and Gricean reasoning. Natural Language Semantics, (): –. Schmidt, Lauren, Goodman, Noah, Barner, David, and Tenenbaum, Joshua (). How tall is tall? Compositionality, statistics, and gradable adjectives. In Niels A. Taatgen and Hedderik van Rijn (eds.), Proceedings of the st Annual Conference of the Cognitive Science Society. Austin, TX: Cognitive Science Society. –. Schöller, Anthea and Franke, Michael (). Semantic values as latent parameters: surprising few and many. In Sarah D’Antonio, Mary Moroney, and Carol-Rose Little (eds.), Proceedings of the th Semantics and Linguistic Theory Conference. Washington, DC: Linguistic Society of America. –. Schöller, Anthea and Franke, Michael (). How many manys? Exploring semantic theories with data-driven computational models. In Nadine Bade, Polina Berezovskaya, and Anthea Schöller (eds.), Proceedings of Sinn und Bedeutung . Stuttgart: University of Stuttgart. –. Schöller, Anthea and Franke, Michael (). Semantic values as latent parameters: testing a fixed threshold hypothesis for cardinal readings of few and many. Linguistics Vanguard, (). doi: ./ lingvan-- Schulz, Katrin and Van Rooij, Robert (). Pragmatic meaning and nonmonotonic reasoning: the case of exhaustive interpretation. Linguistics and Philosophy, (): –. doi: ./s--. Schulz, Laura E. (). Finding new facts; thinking new thoughts. Advances in Child Development and Behavior, : –. Schulz, Petra (). Factivity: Its Nature and Acquisition. Berlin: De Gruyter. Schulze, Cornelia, Grassmann, Susanne, and Tomasello, Michael (). -year-old children make relevance inferences in indirect verbal communication. Child Development, (): –. Schumacher, Petra B. (). The hepatitis called . . . : electrophysiological evidence for enriched composition. In Jörg Meibauer and Markus Steinbach (eds.), Experimental Pragmatics/Semantics. Amsterdam/Philadelphia: John Benjamins. –. Schumacher, Petra B. (). When combinatorial processing results in reconceptualization: towards a new approach of compositionality. Frontiers in Psychology, : . doi: ./fpsyg.. Schumacher, Petra B. (). Content and context in incremental processing: ‘the ham sandwich’ revisited. Philosophical Studies, : –. Schwarz, Bernhard (). At least and quantity implicature: choices and consequences. In Maria Aloni, Maria Franke, and Floris Roelofsen (eds.), Proceedings of the th Amsterdam Colloquium. Amsterdam: University of Amsterdam.
Schwarz, Bernhard, Buccola, Brian, and Hamilton, Michael (). Two types of class B numeral modifiers: a reply to Nouwen . Semantics and Pragmatics, : –. Schwarz, Florian (). Processing presupposed content. Journal of Semantics, (): –. doi: ./jos/ffm Schwarz, Florian (). Presuppositions are fast, whether hard or soft: evidence from the visual world paradigm. In Todd Snider, Sarah D’Antonio, and Mia Weigand (eds.), Proceedings of the th Semantics and Linguistic Theory Conference. Washington, DC: Linguistic Society of America. –. Schwarz, Florian (a). Introduction: presuppositions in context—theoretical issues and experimental perspectives. In Florian Schwarz (ed.), Experimental Perspectives on Presuppositions. Cham: Springer International Publishing. –. Schwarz, Florian (b). Presuppositions vs. asserted content in online processing. In Florian Schwarz (ed.), Experimental Perspectives on Presuppositions. Cham: Springer International Publishing. –. Schwarz, Florian (c). Symmetry and incrementality in conditionals. In Florian Schwarz (ed.), Experimental Perspectives on Presuppositions. Cham: Springer International Publishing. –. Schwarz, Florian (a). Experimental work in presupposition and presupposition projection. Annual Review of Linguistics, : –. Schwarz, Florian (b). False but slow: evaluating statements with non-referring definites. Journal of Semantics, (): –. doi: ./jos/ffu Schwarz, Florian and Tiemann, Sonja (). Presupposition processing: the case of German wieder. In Maria Aloni, Vadim Kimmelmann, Floris Roelofsen, Galit W. Sassoon, Katrin Schulz, and Matthijs Westera (eds.), Proceedings of the th Amsterdam Colloquium. Berlin: Springer. –. Schwarz, Florian and Tiemann, Sonja (). Presupposition projection in online processing. Journal of Semantics, (): –. doi: ./jos/ffw Schwarz, Florian, Clifton, Charles, Jr., and Frazier, Lyn (to appear). Strengthening ‘or’: effects of focus and downward entailing contexts on scalar implicatures. In Jan Anderssen, Keir Moulton, Florian Schwarz, and Cherlon Ussery (eds.), Semantics and Processing (University of Massachusetts Occasional Papers ). Amherst, MA: GLSA Publications. Schwarzschild, Roger (). Plurals, presuppositions, and the sources of distributivity. Natural Language Semantics, : –. Schwarzschild, Roger (). Pluralities. Dordrecht: Kluwer. Schwarzschild, Roger (). Givenness, AvoidF and other constraints on the placement of accent. Natural Language Semantics, : –. Schwarzschild, Roger (). Stubborn distributivity, multiparticipant nouns and the count/mass distinction. In Suzi Lima, Kevin Mullin, and Brian Smith (eds.), Proceedings of the th Meeting of the North East Linguistic Society, Vol. . Amherst, MA: GLSA. –. Schwertel, Uta (). Plural semantics for natural language understanding: a computational prooftheoretic approach. PhD thesis. University of Zürich. Schwoebel, John, Dews, Shelly, Winner, Ellen, and Srinivas, Kavitha (). Obligatory processing of the literal meaning of ironic utterances: further evidence. Metaphor and Symbol, (–): –. doi: ./SMS&_ Scollon, Ronald and Scollon, Suzanne B. K. (). Narrative, Literacy and Face in Interethnic Communication. Norwood, NJ: Ablex Publishing Corporation. Scontras, Gregory and Goodman, Noah D. (). Resolving uncertainty in plural predication. Cognition, : –. Scott, Fiona and Baron-Cohen, Simon (). Imagining real and unreal objects: an investigation of imagination in autism. Journal of Cognitive Neuroscience, : –. doi: ./jocn.... Scott, Rose M. and Baillargeon, Renée (). Early false-belief understanding. Trends in Cognitive Sciences, (): –. Searle, John R. (). Speech Acts: An Essay in the Philosophy of Language. Cambridge: Cambridge University Press.
Searle, John R. (). Metaphors. In Andrew Ortony (ed.), Metaphor and Thought. Cambridge: Cambridge University Press. –. Searle, John R. (). Intentionality: An Essay in the Philosophy of Mind. New York: Cambridge University Press. Sedivy, Julie C. (). Invoking discourse-based contrast sets and resolving syntactic ambiguities. Journal of Memory and Language, : –. Sedivy, Julie C. (). Pragmatic versus form-based accounts of referential contrast: evidence for effects of informativity expectations. Journal of Psycholinguistic Research, (): –. Sedivy, Julie C. and Carlson, Gregory (). Sold on Language: How Advertisers Talk to You and What This Says About You. Chichester: John Wiley and Sons. Sedivy, Julie C., Tanenhaus, Michael K., Chambers, Craig G., and Carlson, Gregory N. (). Achieving incremental semantic interpretation through contextual representation. Cognition, : –. Sekerina, Irina A. and Sauermann, Antje (). Visual attention and quantifier-spreading in heritage Russian bilinguals. Second Language Research, (): –. Sekerina, Irina A. and Sauermann, Antje (). Quantifier spreading in child eye movements: a case of the Russian quantifier kazhdyj ‘every’. Glossa: A Journal of General Linguistics, (): . doi: ./gjgl. Sekerina, Irina A., Brooks, Patricia J., Campanelli, Luca, and Schwartz, Anna M. (). Quantifier spreading in school-age children: an eye-tracking study. In Katalin É. Kiss (ed.), Linguistic and Cognitive Aspects of Quantification. New York: Springer. Selkirk, Elizabeth (). Phonology and Syntax: The Relation Between Sound and Structure. Cambridge, MA: MIT Press. Selkirk, Elizabeth O. (). Sentence prosody: Intonation, stress, and phrasing. In John A. Goldsmith (ed.), Handbook of Phonological Theory. Oxford: Blackwell. –. Selkirk, Elizabeth (). Contrastive focus, givenness and the unmarked status of ‘discourse-new’. Acta Linguistica Hungarica, : –. Serchuk, Phil, Hargreaves, Ian, and Zach, Richard (). Vagueness, logic, and use: four experimental studies on vagueness. Mind and Language, : –. Sereno, Sara C., O’Donnell, Patrick J., and Rayner, Keith (). Eye movements and lexical ambiguity resolution: investigating the subordinate-bias effect. Journal of Experimental Psychology: Human Perception and Performance, (): –. Serratrice, Ludovica and Allen, Shanley E. M. (). The Acquisition of Reference. Amsterdam: John Benjamins. Serratrice, Ludovica, Sorace, Antonella, Filiaci, Francesca, and Baldo, Michela (). Bilingual children’s sensitivity to specificity and genericity: evidence from metalinguistic awareness. Bilingualism: Language and Cognition, : –. Seuren, Pieter A. M. (). Eubulides as a th-century semanticist. Language Sciences, : –. Shamay-Tsoory, Simone G. and Aharon-Peretz, Judith (). Dissociable prefrontal networks for cognitive and affective theory of mind: a lesion study. Neuropsychologia, (): –. doi: ./j.neuropsychologia... Shamay-Tsoory, Simone G., Aharon-Peretz, Judith, and Levkovitz, Yechiel (). The neuroanatomical basis of affective mentalizing in schizophrenia: comparison of patients with schizophrenia and patients with localized prefrontal lesions. Schizophrenia Research, (–): –. doi: ./j.schres... Shannon, Claude E. and Weaver, Warren (). The Mathematical Theory of Communication. Urbana, IL: University of Illinois Press. Shanon, Benny (). On the two kinds of presuppositions in natural language. Foundations of Language, : –. Shao, Jenny and Neville, Helen (). Analyzing semantic processing using event-related brain potentials. Newsletter of the Center for Research in Language, (): –. Shattuck-Hufnagel, Stephanie and Turk, Alice E. (). A prosody tutorial for investigators of auditory sentence processing. Journal of Psycholinguistic Research, : –.
Shatz, Marilyn, Wellman, Henry M., and Silber, Sharon (). The acquisition of mental verbs: a systematic investigation of the first reference to mental state. Cognition, (): –. Sheldon, Amy (). The role of parallel function in the acquisition of relative clauses in English. Journal of Verbal Learning and Verbal Behavior, : –. doi: ./S-()- Sher, Shlomi and McKenzie, Craig R. M. (). Information leakage from logically equivalent frames. Cognition, : –. doi: ./j.cognition... Shetreet, Einat, Chierchia, Gennaro, and Gaab, Nadine (). Linguistic inability or poor performance: dissociating scalar implicature generation and mismatch in the developing brain. Developmental Psychology, (): –. Shibata, Midori, Toyomura, Akira, Itoh, Hiroaki, and Abe, Jun-ichi (). Neural substrates of irony comprehension: a functional MRI study. Brain Research, : –. doi: ./j. brainres... Shibatani, Masayoshi (). Passives and related constructions: a prototype analysis. Language, : –. Shiffrin, Richard M. and Schneider, Walter (). Controlled and automatic human information processing: II. Perceptual learning, automatic attending and a general theory. Psychological Review, : –. doi: ./-X... Shusterman, Anna and Li, Peggy (a). A framework for work on frames of reference. In David Barner and Andrew Scott Baron (eds.), Core Knowledge and Conceptual Change. Oxford: Oxford University Press. –. Shusterman, Anna and Li, Peggy (b). Frames of reference in spatial language acquisition. Cognitive Psychology, , –. doi: ./j.cogpsych... Shusterman, Anna, Lee, Sang Ah, and Spelke, Elizabeth S. (). Cognitive effects of language on human navigation. Cognition, : –. doi: ./j.cognition... Sidnell, Jack (). Conversational turn-taking in a Caribbean English Creole. Journal of Pragmatics, : –. Sigurd, Bengt (). Round numbers. Language in Society, : –. Simons, Mandy (). On the conversational basis of some presuppositions. In Rachel Hastings, Brendan Jackson, and Zsofia Zvolenszky (eds.), Proceedings of the th Semantics and Linguistic Theory Conference. Ithaca, NY: CLC Publications. –. Simons, Mandy (). Presupposition and accommodation: understanding the Stalnakerian picture. Philosophical Studies, (): –. Simons, Mandy (). Dividing things up: the semantics of or and the modal/or interaction. Natural Language Semantics, : –. Simons, Mandy, Tonhauser, Judith, Beaver, David, and Roberts, Craige (). What projects and why. In Nan Li and David Lutz (eds.), Proceedings of the th Conference on Semantics and Linguistic Theory. Ithaca, NY: CLC Publications. –. Simons, Mandy, Beaver, David, Roberts, Craige, and Tonhauser, Judith (). The best question: explaining the projection behavior of factive verbs. Discourse Processes, : –. Singh, Raj (). Maximize presupposition! and local contexts. Natural Language Semantics, : –. Singh, Raj, Fedorenko, Evelina, Mahowald, Kyle, and Gibson, Edward (). Accommodating presuppositions is inappropriate in implausible contexts. Cognitive Science, (): –. Singh, Raj, Wexler, Ken, Astle-Rahim, Andrea, Kamawar, Deepthi, and Fox, Danny (). Children interpret disjunction as conjunction: consequences for the theories of implicature and child development. Natural Language Semantics, (): –. doi: ./s--- Sirota, Miroslav and Juanchich, Marie (). To what extent do politeness expectations shape risk perception? Even numerical probabilities are under the spell! Acta Psychologica, : –. doi: ./j.actpsy... Sirota, Miroslav and Juanchich, Marie (). A direct and comprehensive test of two postulates of politeness theory applied to uncertainty communication. Judgment and Decision Making, (): –.
Sitnikova, Tatiana, Kuperberg, Gina R., and Holcomb, Phillip J. (). Semantic integration in videos of real-world events: an electrophysiological investigation. Psychophysiology, (): –. Skantze, Gabriel and Schlangen, David (). Incremental dialogue processing in a micro-domain. Proceedings of the th Conference of the European Chapter of the Association for Computational Linguistics. Stroudsburg, PA: Association for Computational Linguistics. Skolnick, Deena and Bloom, Paul (). What does Batman think about SpongeBob? Children’s understanding of the fantasy/fantasy distinction. Cognition, : B–B. Skopeteas, Stavros and Fanselow, Gisbert (). Focus and the exclusion of alternatives: on the interaction of syntactic structure with pragmatic inference. Lingua, : –. Skordos, Dimitrios and Papafragou, Anna (). Lexical, syntactic, and semantic-geometric factors in the acquisition of motion predicates. Developmental Psychology, (): –. doi: ./ a Skordos, Dimitrios and Papafragou, Anna (). Children’s derivation of scalar implicatures: alternatives and relevance. Cognition, : –. Skordos, Dimitrios, Myers, Allyson, and Barner, David (under review). Access to alternatives and universal quantification in language acquisition. Slabakova, Roumyana (). Learnability in the second language acquisition of semantics: a bidirectional study of a semantic parameter. Second Language Research, : –. Slobin, Dan I. (). Two ways to travel: verbs of motion in English and Spanish. In Masayoshi Shibatani and Sandra A. Thompson (eds.), Grammatical Constructions: Their Form and Meaning. Oxford: Oxford University Press. –. Slobin, Dan I. (). Language and thought online: cognitive consequences of linguistic relativity. In Dedre Gentner and Susan Goldin-Meadow (eds.), Language in Mind. Cambridge, MA: MIT Press. –. Slugoski, Ben R. (). Mindless processing of requests? Don’t ask twice. British Journal of Social Psychology, (): –. Slugoski, Ben R. and Turnbull, William (). Cruel to be kind and kind to be cruel: sarcasm, banter and social relations. Journal of Language and Social Psychology, (): –. Smith, Carlota S. (). The Parameter of Aspect. Dordrecht: Kluwer. Smith, Carol L. (). Quantifiers and question answering in young children. Journal of Experimental Child Psychology, : –. doi: ./-()- Smith, E. Allyn and Hall, Kathleen C. (). Projection diversity: experimental evidence. In Gregory Kierstead (ed.), Proceedings of the ESSLLI Workshop on Projective Content. Columbus, OH: Ohio State University. Smith, J. David, Shields, Wendy E., and Washburn, David A. (). The comparative psychology of uncertainty monitoring and metacognition. Behavioral and Brain Sciences, : –. Smith, J. David, Coutinho, Mariana V. C., Church, Barbara A., and Beran, Michael J. (). Executive-attentional uncertainty responses by rhesus macaques (Macaca mulatta). Journal of Experimental Psychology: General, : –. doi: ./a Smith, J. W. A. (). Children’s comprehension of metaphor: a Piagetian interpretation. Language and Speech, : –. doi: ./ Smith, Nathaniel J. and Levy, Roger (). The effect of word predictability on reading time is logarithmic. Cognition, (): –. doi: ./j.cognition... Smith, Nicholas J. J. (). Vagueness and Degrees of Truth. Oxford: Oxford University Press. Smithson, Michael, Budescu, David V., Broomell, Stephen B., and Por, Han-Hui (). Never say ‘not’: impact of negative wording in probability phrases on imprecise probability judgments. International Journal of Approximate Reasoning, : –. doi: ./j.ijar... Smyth, Ron (). Grammatical determinants of ambiguous pronoun resolution. Journal of Psycholinguistic Research, : –. doi: ./BF Sneed, Elisa (). The role of input in the acquisition of generic NPs. In Alejna Brugos, Manuella R. Clark-Cotton, and Seungwan Ha (eds.), BUCLD Online Proceedings Supplement. Boston, MA: Boston University.
Sneed, Elisa (). Input in the acquisition of genericity. PhD thesis. Northwestern University. Soames, Scott (). How presuppositions are inherited: a solution to the projection problem. Linguistic Inquiry, : –. Soames, Scott (). Presupposition. In Dov Gabbay and Franz Guenther (eds.), Handbook of Philosophical Logic, vol. IV. Dordrecht: Riedel. –. Sodian, Beate, Thoermer, Claudia, and Dietrich, Nadine (). Two- to four-year-old children’s differentiation of knowing and guessing in a non-verbal task. European Journal of Developmental Psychology, (): –. doi: ./ Soja, Nancy N., Carey, Susan, and Spelke, Elizabeth S. (). Ontological categories guide young children’s inductions of word meaning: object terms and substance terms. Cognition, (): –. doi: ./-()- Solt, Stephanie (). Few more and many fewer: complex quantifiers based on many and few. In Rick Nouwen and Jakub Dotlačil (eds.), Proceedings of the ESSLLI Workshop on Quantifier Modification. Dublin: Trinity College Dublin. Solt, Stephanie (). Notes on the comparison class. In Rick Nouwen, Robert van Rooij, Uli Sauerland, and Hans-Christian Schmitz (eds.), Vagueness in Communication. Heidelberg: Springer. –. Solt, Stephanie (). Vagueness and imprecision: empirical foundations. Annual Review of Linguistics, : –. Solt, Stephanie (). On measurement and quantification: the case of most and more than half. Language, : –. Solt, Stephanie (). Multidimensionality, subjectivity and scales: experimental evidence. In Elena Castroviejo, Louise McNally, and Galit Sassoon (eds.), Gradability, Scale Structure and Vagueness: Experimental Perspectives. Cham: Springer. Solt, Stephanie and Gotzner, Nicole (). Experimenting with degree. In Anca Chereches (ed.), Proceedings of the nd Semantics and Linguistic Theory Conference. Ithaca, NY: CLC Publications. –. Solt, Stephanie, Cummins, Chris, and Palmović, Marijan (). The preference for approximation. International Review of Pragmatics, : –. Sorensen, Roy (). The sorites and the Generic Overgeneralization Effect. Analysis, (): –. Sorensen, Roy (). Vagueness. In Edward N. Zalta (ed.), Stanford Encyclopedia of Philosophy, winter edn. Stanford, CA: Metaphysics Research Lab, Stanford University. Southgate, Victoria and Hamilton, Antonia F. de C. (). Unbroken mirrors: challenging a theory of Autism. Trends in Cognitive Sciences, (): –. doi: ./j.tics... Southgate, Victoria, Senju, Atsushi, and Csibra, Gergely (). Action anticipation through attribution of false belief by -year-olds. Psychological Science, (): –. Southgate, Victoria, Chevallier, Coralie, and Csibra, Gergely (). Sensitivity to communicative relevance tells young children what to imitate. Developmental Science, : –. Southgate, Victoria, Chevallier, Coralie, and Csibra, Gergely (). Seventeen-month-olds appeal to false beliefs to interpret others’ referential communication. Developmental Science, (): –. doi: ./j.-...x Spalek, Katharina, Gotzner, Nicole, and Wartenburger, Isabell (). Not only the apples: focus sensitive particles improve memory for information-structural alternatives. Journal of Memory and Language, : –. doi: ./j.jml... Spector, Benjamin (). Aspects of the pragmatics of plural morphology: on higher-order implicatures. In Uli Sauerland and Penka Stateva (eds.), Presupposition and Implicature in Compositional Semantics. Houndmills: Palgrave Macmillan. –. Speelman, Craig P. and Kirsner, Kim (). The representation of text-based and situation-based information in discourse comprehension. Journal of Memory and Language, : –. doi: ./-X()-P Speer, Shari R., Crowder, Robert G., and Thomas, Lisa M. (). Prosodic structure and sentence recognition. Journal of Memory and Language, : –. doi: ./jmla..
Speer, Shari R., Warren, Paul, and Schafer, Amy J. (). Situationally independent prosodic phrasing. Laboratory Phonology, : –. Spelke, Elizabeth S. and Tsivkin, Sanna (). Initial knowledge and conceptual change: space and number. In Melissa Bowerman and Stephen C. Levinson (eds.), Language Acquisition and Conceptual Development. Cambridge: Cambridge University Press. –. Spenader, Jennifer (). Presuppositions in spoken discourse. PhD thesis. University of Stockholm. Computational Linguistics. Sperber, Dan and Wilson, Deirdre (). Ordered entailments: an alternative to presuppositional theories. In Choon-Kyu Oh and David A. Dinneen (eds.), Presupposition, Syntax and Semantics, vol. XI. New York: Academic Press. –. Sperber, Dan and Wilson, Deirdre (). Irony and the use-mention distinction. In Peter Cole (ed.), Radical Pragmatics. New York: Academic Press. –. Sperber, Dan and Wilson, Deirdre (/). Relevance: Communication and Cognition. Cambridge, MA: Harvard University Press (nd edn ). Sperber, Dan and Wilson, Deirdre (). A deflationary account of metaphors. In Raymond W. Gibbs Jr. (ed.), Cambridge Handbook of Metaphor and Thought. New York: Cambridge University Press. –. Spotorno, Nicola and Noveck, Ira A. (). When is irony effortful? Journal of Experimental Psychology: General, (): –. doi: ./a Spotorno, Nicola, Koun, Eric, Prado, Jérôme, Van Der Henst, Jean-Baptiste, and Noveck, Ira A. (). Neural evidence that utterance-processing entails mentalizing: the case of irony. NeuroImage, (): –. doi: ./j.neuroimage... Spotorno, Nicola, Cheylus, Anne, Van Der Henst, Jean-Baptiste, and Noveck, Ira A. (). What’s behind a P? Integration operations during irony processing. PLoS ONE, (): e. doi: ./journal.pone. Spranca, Mark, Minsk, Elisa, and Baron, Jonathan (). Omission and commission in judgment and choice. Journal of Experimental Social Psychology, : –. Sprouse, Jon, Schütze, Carson T., and Almeida, Diogo (). A comparison of informal and formal acceptability judgments using a random sample from Linguistic Inquiry –. Lingua, : –. Spychalska, Maria, Kontinen, Jarmo, and Werning, Markus (). Investigating scalar implicatures in a truth-value judgment task: evidence from event-related brain potentials. Language, Cognition and Neuroscience, (): –. doi: ./.. Srinivasan, Mahesh and Barner, David (). The Amelia Bedelia effect: world knowledge and the goal bias in language acquisition. Cognition, (): –. Srinivasan, Mahesh and Rabagliati, Hugh (). How concepts and conventions structure the lexicon: cross-linguistic evidence from polysemy. Lingua, : –. Srinivasan, Mahesh and Snedeker, Jesse (). Judging a book by its cover and its contents: the representation of polysemous and homophonous meanings in four-year-old children. Cognitive Psychology, : –. Srinivasan, Mahesh and Snedeker, Jesse (). Polysemy and the taxonomic constraint: children’s representation of words that label multiple kinds. Language Learning and Development, (): –. Srinivasan, Prakash and Yates, Alexander (). Quantifier scope disambiguation using extracted pragmatic knowledge: preliminary results. Proceedings of the Conference on Empirical Methods in Natural Language Processing, Vol. . Stroudsburg, PA, USA: Association for Computational Linguistics. –. Srinivasan, Mahesh, Al-Mughairy, Sara, Foushee, Ruthe, and Barner, David (). Learning language from within: children use semantic generalizations to infer word meanings. Cognition, : –. St. George, Marie, Kutas, Marta, Martinez, A., and Sereno, Martin I. (). Semantic integration in reading: engagement of the right hemisphere during discourse processing. Brain, (): –.
Stallard, David (). Two kinds of metonymy. Proceedings from the st Annual Meeting of the Association for Computational Linguistics. Stroudsburg, PA: Association for Computational Linguistics. –. Stalnaker, Robert (). Presuppositions. Journal of Philosophical Logic, (): –. doi: ./ bf. Stalnaker, Robert (). Pragmatic presuppositions. In Milton K. Munitz and Peter K. Unger (eds.), Semantics and Philosophy. New York: New York University Press. –. Stalnaker, Robert (). Assertion. In Peter Cole (ed.), Syntax and Semantics : Pragmatics. New York: Academic Press. –. Stalnaker, Robert (). Common ground. Linguistics and Philosophy, (): –. Stanfield, Robert A. and Zwaan, Rolf A. (). The effect of implied orientation derived from verbal context on picture recognition. Psychological Science, : –. doi: ./-. Stanley, Jason and Szabó, Zoltán G. (). On quantifier domain restriction. Mind and Language, (–): –. Stateva, Penka, Andreetta, Sara, and Stepanov, Arthur (). On the nature of the plurality inference: ladybugs for Anne. In Ludivine Dupuy, Adrianna Grabizna, Nadège Foudon, and Pierre SaintGermier (eds.), Papers Dedicated to Anne Reboul. Lyon: CNRS. http://reboul.isc.cnrs.fr. Staub, Adrian (). The parser doesn’t ignore intransitivity, after all. Journal of Experimental Psychology: Learning, Memory, and Cognition, (): –. doi: ./-... Staudte, Maria, Crocker, Matthew W., Heloir, Alexis, and Kipp, Michael (). The influence of speaker gaze on listener comprehension: contrasting visual versus intentional accounts. Cognition, (): –. doi: ./j.cognition... Steedman, Mark (). Information structure and the syntax-phonology interface. Linguistic Inquiry, : –. Steedman, Mark (). Information-structural semantics for English intonation. In Chungmin Lee, Matthew Gordon, and Daniel Büring (eds.), Topic and Focus: Crosslinguistic Perspectives on Meaning and Intonation. Dordrecht: Springer. –. Steedman, Mark (). The surface-compositional semantics of English intonation. Language, : –. Steinhauer, Karsten, Drury, John E., Portner, Paul, Walenski, Matthew, and Ullman, Michael T. (). Syntax, concepts, and logic in the temporal dynamics of language comprehension: evidence from event-related potentials. Neuropsychologia, (): –. doi: ./j.neuropsychologia... Stephan, Elena, Liberman, Nira, and Trope, Yaacov (). Politeness and psychological distance: a construal level perspective. Journal of Personality and Social Psychology, (): –. doi: ./a Stephenson, Tamina (). Judge dependence, epistemic modals, and predicates of personal taste. Linguistics and Philosophy, : –. Sterken, Rachel Katharine (). Leslie on generics. Philosophical Studies, (): –. Sternberg, Robert J. and Gastel, Joyce (). Coping with novelty in human intelligence: an empirical investigation. Intelligence, (): –. Stevens, Jon Scott, De Marneffe, Marie-Catherine, Speer, Shari R., and Tonhauser, Judith (). Rational use of prosody predicts projectivity in manner adverb utterances. In Glenn Gunzelmann, Andrew Howes, Thora Tenbrink, and Eddy J. Davelaar (eds.), Proceedings of the th Annual Meeting of the Cognitive Science Society. Austin, TX: Cognitive Science Society. Stevenson, Rosemary J. and Over, David E. (). Deduction from uncertain premises. Quarterly Journal of Experimental Psychology, A: –. doi: ./ Stevenson, Rosemary J. and Over, David E. (). Reasoning from uncertain premises: effects of expertise and conversational context. Thinking and Reasoning, : –. Stevenson, Rosemary J., Crawley, Rosalind A., and Kleinman, David (). Thematic roles, focus, and the representation of events. Language and Cognitive Processes, : –. doi: ./
Stevenson, Rosemary, Knott, Alistair, Oberlander, Jon, and McDonald, Sharon (). Interpreting pronouns and connectives: interactions among focusing, thematic roles, and coherence relations. Language and Cognitive Processes, : –. doi: ./ Stewart, Andrew J., Pickering, Martin J., and Sanford, Anthony J. (). Implicit consequentiality. In Morton Ann Gernsbacher and Sharon J. Derry (eds.), Proceedings of the th Annual Conference of the Cognitive Science Society. Mahwah, NJ: Lawrence Erlbaum. –. Stewart, Andrew J., Pickering, Martin J., and Sanford, Anthony J. (). The time course of the influence of implicit causality information: focusing versus integration accounts. Journal of Memory and Language, (): –. doi: ./jmla.. Stewart, Andrew J., Haigh, Matthew, and Kidd, Evan (). An investigation into the online processing of counterfactual and indicative conditionals. Quarterly Journal of Experimental Psychology, (): –. doi: ./ Stiller, Alex J., Goodman, Noah D., and Frank, Michael C. (). Ad-hoc scalar implicature in adults and children. In Laura Carlson, Christoph Hölscher, and Thomas F. Shipley (eds.), Proceedings of the rd Annual Conference of the Cognitive Science Society. Austin, TX: Cognitive Science Society. Stiller, Alex J., Goodman, Noah D., and Frank, Michael C. (). Ad-hoc implicature in preschool children. Language Learning and Development, (): –. Stites, Lauren J. and Özçaliskan, Şeyda (). Developmental changes in children’s comprehension and explanation of spatial metaphors for time. Journal of Child Language, (): –. Stivers, Tanya, Enfield, Nick J., Brown, Penelope et al. (). Universals and cultural variation in turn-taking in conversation. Proceedings of the National Academy of Sciences of the United States of America, (): –. Stockall, Linnaea and Husband, E. Matthew (). Processing (the) events: lexical and structural ingredients of inner aspect. In Carson T. Schütze and Linnaea Stockall (eds.), Connectedness: Papers by and for Sarah VanWagenen. UCLA Working Papers in Linguistics . London: UCLA. Stolterfoht, Britta, Friederici, Angela D., Alter, Kai, and Steube, Anita (). Processing focus structure and implicit prosody during reading: differential ERP effects. Cognition, : –, . doi: ./j.cognition... Stone, Valerie E., Baron-Cohen, Simon, Calder, Andrew, Keane, Jill, and Young, Andrew (). Acquired theory of mind impairments in individuals with bilateral amygdala lesions. Neuropsychologia, (): –. Stowe, Laurie A., Tanenhaus, Michael K., and Carlson, Gregory N. (). Filling gaps on-line: use of lexical and semantic information in sentence processing. Language and Speech, (): –. doi: ./ Strawson, Peter F. (). On referring. Mind, : –. Streefkerk, Barbertje, Pols, Louis C. W., and Ten Bosch, Louis F. M. (). Prominence in read aloud sentences, as marked by listeners and classifiers automatically. Proceedings of the Institute of Phonetic Sciences. Amsterdam: University of Amsterdam, : –. Street, James A. and Dąbrowska, Ewa (). More individual differences in language attainment: how much do adult native speakers of English know about passives and quantifiers? Lingua, (): –. doi: ./j.lingua... Sturt, Patrick, Sanford, Anthony J., Stewart, Andrew, and Dawydiak, Eugene (). Linguistic focus and good-enough representations: an application of the change-detection paradigm. Psychonomic Bulletin and Review, (): –. Sudo, Yasutada (). On the semantics of phi features on pronouns. PhD thesis. Cambridge, MA: Massachusetts Institute of Technology. Sudo, Yasutada and Spathas, Giorgos (). Gendered nouns and nominal ellipsis in Greek. Ms. UCL. Sugawara, Ayaka, Kotek, Hadas, Hackl, Martin, and Wexler, Ken (). Long vs. short QR: evidence from the acquisition of ACD. In Sarah Baiz, Nora Goldman, and Rachel Hawkes (eds.), Proceedings of the th Annual Boston University Conference on Language Development. Somerville, MA: Cascadilla Press. –.
Surian, Luca, Baron-Cohen, Simon, and Van der Lely, Heather (). Are children with autism deaf to Gricean maxims? Cognitive Neuropsychiatry, : –. doi: ./ Swinney, David A. (). Lexical access during sentence comprehension: (re)consideration of context effects. Journal of Verbal Learning and Verbal Behavior, (): –. Syrett, Kristen (a). QR out of a tensed clause: evidence from Antecedent-Contained Deletion. Ratio, : –. Syrett, Kristen (b). Experimental support for inverse scope readings of finite-clause embedded Antecedent-Contained Deletion sentences. Linguistic Inquiry, : –. Syrett, Kristen (c). Events and agents in the acquisition of universal quantification. Theoretical Linguistics, : –. Syrett, Kristen (d). Mapping properties to individuals in language acquisition. In Elizabeth Grillo and Kyle Jepson (eds.), Proceedings of the th Annual Boston University Conference on Language Development. Somerville, MA: Cascadilla Press. –. Syrett, Kristen and Lidz, Jeffrey (). Children want to access every interpretation adults do. In Leah Bateman and Cherlon Ussery (eds.), Proceedings of the th Annual Meeting of the North East Linguistic Society. Charleston, SC: Booksurge Publishing. –. Syrett, Kristen and Lidz, Jeffrey (). QR in child grammar: evidence from Antecedent-Contained Deletion. Language Acquisition: A Journal of Developmental Linguistics, : –. Syrett, Kristen and Lidz, Jeffrey (). Competence, performance and the locality of Quantifier Raising: Evidence from -year-old children. Linguistic Inquiry, : –. Syrett, Kristen and Musolino, Julien (). Collectivity, distributivity, and the interpretation of numerical expressions in child and adult language. Language Acquisition: A Journal of Developmental Linguistics, : –. Syrett, Kristen and Musolino, Julien (). All together now: disentangling semantics and pragmatics with ‘together’ in child and adult language. Language Acquisition: A Journal of Developmental Linguistics, : –. doi: ./.. Syrett, Kristen, Bradley, Evan, Kennedy, Christopher, and Lidz, Jeffrey (). Shifting standards: children’s understanding of gradable adjectives. In Kamil ud Deen, Jun Nomura, Barbara Schulz, and Bonnie D. Schwartz (eds.), Proceedings of the Inaugural Conference on Generative Approaches to Language Acquisition—North America (GALANA), vol. . Cambridge, MA: UCONN Occasional Papers in Linguistics. –. Syrett, Kristen, Kennedy, Christopher, and Lidz, Jeffrey (). Meaning and context in children’s understanding of gradable adjectives. Journal of Semantics, : –. Szabó, Zoltán Gendler (ed.) (). Semantics versus Pragmatics. Oxford: Clarendon Press. Szabolcsi, Anna (). Compositionality in focus. Acta Linguistica Societatis Linguistice Europaeae, : –. Szabolcsi, Anna (). All quantifiers are not equal: the case of focus. Acta Linguistica Hungarica, : –. Szabolcsi, Anna (). Ways of Scope Taking. Dordrecht: Kluwer. Szabolcsi, Anna (). Quantification and ACD: what is the evidence from real-time processing evidence for? A response to Hackl et al. (). Journal of Semantics, : –. Taler, Vanessa, Klepousniotou, Ekaterini, and Phillips, Natalie A. (). Comprehension of lexical ambiguity in healthy aging, mild cognitive impairment, and mild Alzheimer’s disease. Neuropsychologia, : –. Talmy, Leonard (). Figure and ground in complex sentences. In Cathy Cogen, Henry Thompson, Graham Thurgood, Kenneth Whistler, and James Wright (eds.), Proceedings of the First Annual Meeting of the Berkeley Linguistics Society. Berkeley, CA: Berkeley Linguistics Society. –. Talmy, Leonard (). How language structures space. In Herbert L. Pick and Linda P. Acredolo (eds.), Spatial Orientation: Theory, Research and Application. New York, NY: Plenum Press. –. Talmy, Leonard (). Lexicalization patterns: semantic structure in lexical forms. In Timothy Shopen (ed.), Language Typology and Syntactic Description. New York, NY: Cambridge University Press. –.
Tanenhaus, Michael K. and Trueswell, John C. (). Sentence comprehension. In Joanne L. Miller and Peter D. Eimas (eds.), Speech, Language, and Communication. San Diego, CA: Academic Press. –. Tanenhaus, Michael K., Leiman, James M., and Seidenberg, Mark S. (). Evidence for multiple stages in the processing of ambiguous words in syntactic contexts. Journal of Verbal Learning and Verbal Behavior, (): –. Tanenhaus, Michael K., Spivey-Knowlton, Michael J., Eberhard, Kathleen M., and Sedivy, Julie C. (). Integration of visual and linguistic information in spoken language comprehension. Science, (): –. doi: ./science. Tanz, Christine (). Studies in the Acquisition of Deictic Terms. Cambridge: Cambridge University Press. Tarenskeen, Sammie, Broersma, Mirjam, and Geurts, Bart (). ‘Hand me the yellow stapler’ or ‘Hand me the yellow dress’: colour overspecification depends on object category. In Christine Howes and Staffan Larsson (eds.), Proceedings of the th Workshop on the Semantics and Pragmatics of Dialogue (SemDial). Gothenburg: University of Gothenburg. –. Taylor, Ann E. and Saint-Cyr, Jean A. (). The neuropsychology of Parkinson’s disease. Brain and Cognition, (), –. Teichman, Matt (). The sophisticated kind theory. Inquiry, –. doi: ./X.. Teigen, Karl Halvor and Brun, Wibecke (). Yes, but it is uncertain: direction and communicative intention of verbal probabilistic terms. Acta Psychologica, : –. doi: ./-() E- Teigen, Karl Halvor and Brun, Wibecke (). The directionality of verbal probability expressions: effects on decisions, predictions, and probabilistic reasoning. Organizational Behavior and Human Decision Processes, : –. doi: ./obhd.. Teigen, Karl Halvor and Brun, Wibecke (). Ambiguous probabilities: when does p = . reflect a possibility, and when does it express a doubt? Journal of Behavioral Decision Making, : –. doi: ./-(/):..CO;-U Teigen, Karl Halvor and Brun, Wibecke (a). Verbal expressions of uncertainty and probability. In David Hardman and Laura Macchi (eds.), Thinking: Psychological Perspectives on Reasoning, Judgment and Decision-Making. Chichester: John Wiley and Sons. –. Teigen, Karl Halvor and Brun, Wibecke (b). Verbal probabilities: a question of frame? Journal of Behavioral Decision Making, : –. doi: ./bdm. Teigen, Karl Halvor, Evensen, Pia C., Samoilow, Dimitrij K., and Vatne, Karin B. (). Good luck and bad luck: how to tell the difference. European Journal of Social Psychology, : –. Teigen, Karl Halvor, Juanchich, Marie, and Riege, Anine H. (). Improbable outcomes: infrequent or extraordinary? Cognition, : –. doi: ./j.cognition... Teigen, Karl Halvor, Juanchich, Marie, and Filkuková, Petra (). Verbal probabilities: an alternative approach. Quarterly Journal of Experimental Psychology, (): –. doi: ./ .. Teixeira, Cláudia and Silva, Ana Fialho (). The interpretation of verbal probability expressions used in the IAS/IFRS: some Portuguese evidence. Tekhné: Revista de Estudos Politécnicos/Polytechnical Studies Review, : –. Ten Bosch, Louis, Oostdijk, Nelleke, and De Ruiter, Jan P. (). Durational aspects of turn-taking in spontaneous face-to-face and telephone dialogues. Paper presented at the th International Conference on Text, Speech and Dialogue (TSD), Brno. Tenny, Carol L. (). Aspectual Roles and the Syntax-Semantics Interface. Dordrecht: Springer. Ter Meulen, Alice (). Generic information, conditional contexts and constraints. In Elizabeth C. Traugott, Alice ter Meulen, Judy Snitzer Reilly, and Charles A. Ferguson (eds.), On Conditionals. Cambridge: Cambridge University Press. –. Terken, Jacques and Hirschberg, Julia (). Deaccentuation of words representing ‘given’ information: effects of persistence of grammatical function and surface position. Language and Speech, : –.
Terken, Jacques and Nooteboom, Sieb G. (). Opposite effects of accentuation and deaccentuation on verification latencies for given and new information. Language and Cognitive Processes, : –. Terkourafi, Marina (). Politeness in Cypriot Greek: a frame-based approach. PhD thesis. University of Cambridge. Terkourafi, Marina (). Beyond the micro-level in politeness research. Journal of Politeness Research, : –. Tesan, Graciela, Johnson, Blake W., and Crain, Stephen (). How the brain responds to ‘any’: an MEG study. Brain and Language, (): –. doi: ./j.bandl... Tettamanti, Marco, Manenti, Rosa, Della Rosa, Pasquale A. et al. (). Negation in the brain: modulating action representations. NeuroImage, (): –. doi: ./j.neuroimage... Thatcher, Robert W. (). Cyclic cortical reorganization during early childhood. Brain and Cognition, : –. doi: ./-()-Y Theil, Michael (). The role of translations of verbal into numerical probability expressions in risk management: a meta-analysis. Journal of Risk Research, : –. doi: ./ Thibodeau, Paul H. and Durgin, Frank H. (). Metaphor aptness and conventionality: a processing fluency account. Metaphor and Symbol, (): –. Thornton, Rosalind (). Referentiality and wh-movement in child English: juvenile d-linkuency. Language Acquisition, : –. Thorward, Jennifer (). The interaction of contrastive stress and grammatical context in child English speakers’ interpretations of existential quantifiers. BA thesis. Ohio State University. Tian, Ye and Breheny, Richard (under review). A new look at negative sentence verification. Tian, Ye, Breheny, Richard, and Ferguson, Heather J. (). Why we simulate negated information: a dynamic pragmatic account. Quarterly Journal of Experimental Psychology, (): –. Tian, Ye, Ferguson, Heather J., and Breheny, Richard (). Processing negation without context— why and when we represent the positive argument. Language, Cognition and Neuroscience, (): –. doi: ./.. Tiemann, Sonja (). The processing of wieder (‘again’) and other presupposition triggers. PhD thesis. Eberhard Karls Universität Tübingen. Tiemann, Sonja, Schmid, Mareike, Bade, Nadine et al. (). Psycholinguistic evidence for presuppositions: on-line and off-line data. In Ingo Reich, Eva Horch, and Dennis Pauly (eds.), Proceedings of Sinn und Bedeutung . Saarbrücken: Universitätsverlag des Saarlandes. –. Tiemann, Sonja, Kirsten, Mareike, Beck, Sigrid, Hertrich, Ingo, and Rolke, Bettina (). Presupposition processing and accommodation: an experiment on wieder (‘again’) and consequences for other triggers. In Florian Schwarz (ed.), Experimental Perspectives on Presuppositions. Cham: Springer International Publishing. –. Tieu, Lyn (). Logic and grammar in child language: how children acquire the semantics of polarity sensitivity. PhD thesis. University of Connecticut. Tieu, Lyn (). Input vs. output in the acquisition of negative polarity: the curious case of any. In Pierre Larrivée and Chungmin Lee (eds.), Negation and Polarity: Experimental Perspectives. Springer. –. Tieu, Lyn and Lidz, Jeffrey (). NPI licensing and beyond: children’s knowledge of the semantics of ‘any’. Language Acquisition, (): –. Tieu, Lyn, Bill, Cory, Romoli, Jacopo, and Crain, Stephen (). Plurality inferences are scalar implicatures: evidence from acquisition. In Todd Snider, Sarah D’Antonio, and Mia Weigand (eds.), Proceedings of the th Semantics and Linguistic Theory Conference. Washington, DC: Linguistic Society of America. –. Tieu, Lyn, Bill, Cory, Romoli, Jacopo, and Crain, Stephen (). Plurality inferences are scalar implicatures: further evidence from acquisition. Presentation at the Workshop on Experimental and Crosslinguistic Evidence for the Distinction between Implicatures and Presuppositions. ZAS, Berlin. Tieu, Lyn, Romoli, Jacopo, Zhou, Peng, and Crain, Stephen (). Children’s knowledge of free choice inferences and scalar implicatures. Journal of Semantics, (): –.
Tieu, Lyn, Bill, Cory, Romoli, Jacopo, and Crain, Stephen (). Testing theories of plural meanings: insights from child language. Ms., Macquarie University. Tieu, Lyn, Yatsushiro, Kazuko, Cremers, Alexandre, Romoli, Jacopo, Sauerland, Uli, and Chemla, Emmanuel (). On the role of alternatives in the acquisition of simple and complex disjunctions in French and Japanese. Journal of Semantics, (): –. Tily, Harry and Piantadosi, Steven (). Refer efficiently: use less informative expressions for more predictable meanings. In Kees van Deemter, Albert Gatt, Roger P. G. van Gompel, and Emiel Krahmer (eds.), Proceedings of the Workshop on the Production of Referring Expressions: Bridging the Gap Between Computational and Empirical Approaches to Reference (PRE-CogSci ). Amsterdam: Cognitive Science Society. Todorova, Marina, Straub, Kathy, Badecker, William, and Frank, Robert (). Aspectual coercion and the online computation of sentential aspect. In Lila R. Gleitman and Aravind K. Joshi (eds.), Proceedings of the nd Annual Conference of the Cognitive Science Society. Austin, TX: Cognitive Science Society. –. Tomasello, Michael (). Joint attention as social cognition. In Chris Moore and Philip J. Dunham (eds.), Joint Attention: Its Origins and Role in Development. Hillsdale, NJ: Erlbaum. –. Tomasello, Michael (). Constructing a Language: A Usage-Based Account of Language Acquisition. Cambridge, MA: Harvard University Press. Tomasello, Michael (). Origins of Human Communication. Cambridge, MA: MIT Press. Tomasello, Michael and Haberl, Katharina (). Understanding attention: - and -month-olds know what is new for other persons. Developmental Psychology, (): –. doi: ./... Tomasello, Michael, Carpenter, Malinda, Call, Josep, Behne, Tanya, and Moll, Henrike (). Understanding and sharing intentions: the origins of cultural cognition. Behavioral and Brain Sciences, (): –. doi: ./SX Tomasino, Barbara, Weiss, Peter H., and Fink, Gereon R. (). To move or not to move: imperatives modulate action-related verb processing in the motor system. Neuroscience, (): –. doi: ./j.neuroscience... Tomlinson, John M., Bailey, Todd M., and Bott, Lewis (). Possibly all of that and then some: scalar implicatures are understood in two steps. Journal of Memory and Language, (): –. doi: ./j.jml... Tonhauser, Judith (). On the heterogeneity of projective content. Presentation at the Workshop on Experimental and Crosslinguistic Evidence for the Distinction between Implicatures and Presuppositions, ZAS, Berlin. Tonhauser, Judith (). Prosodic cues to speaker commitment. In Mary Moroney, Carol-Rose Little, Jacob Collard, and Dan Burgdorf (eds.), Proceedings of the th Semantics and Linguistic Theory Conference. Ithaca, NY: CLC Publications. –. Tonhauser, Judith, Beaver, David, Roberts, Craige, and Simons, Mandy (). Towards a taxonomy of projective content. Language, (): –. Tottie, Gunnel (). Negation in English Speech and Writing: A Study in Variation. San Diego: Academic. Townsend, David J. (). Aspectual coercion in eye movements. Journal of Psycholinguistic Research, (): –. doi: ./s--- Townsend, David J. and Bever, Thomas G. (). Sentence Comprehension: The Integration of Habits and Rules. Cambridge, MA: MIT Press. Trabasso, Tom, Rollins, Howard, and Shaughnessy, Edward (). Storage and verification stages in processing concepts. Cognitive Psychology, (): –. doi: ./-()- Tracy, Karen (). The many faces of facework. In Howard Giles and W. Peter Robinson (eds.) Handbook of Language and Social Psychology. Chichester: Wiley. –. Traugott, Elizabeth Closs and Dasher, Richard B. (). Regularity in Semantic Change. Cambridge: Cambridge University Press.
Traxler, Matthew J., Pickering, Martin J., and McElree, Brian (). Coercion in sentence processing: evidence from eye-movements and self-paced reading. Journal of Memory and Language, (): –. Tribushinina, Elena (). Once again on norms and comparison classes. Linguistics, : –. Tribushinina, Elena (). Adjective semantics, world knowledge and visual context: comprehension of size terms by - to -year-old Dutch-speaking children. Journal of Psycholinguistic Research, : –. Trinh, Tue and Haida, Andreas (). Constraining the derivation of alternatives. Natural Language Semantics, (): –. doi: ./s---y Troiani, Vanessa, Peelle, Jonathan E., Clark, Robin, and Grossman, Murray (). Is it logical to count on quantifiers? Dissociable neural networks underlying numerical and logical quantifiers. Neuropsychologia, : –. Trope, Yaacov and Liberman, Nira (). Temporal construal. Psychological Review, (): –. doi: ./-X... Trope, Yaacov and Liberman, Nira (). Construal-level theory of psychological distance. Psychological Review, (): –. Troster, Alexander I. and Woods, Steven Paul (). Neuropsychological aspects of Parkinson’s disease and parkinsonian syndromes. In Rajesh Pahwa, Kelly E. Lyons, and William C. Koller (eds.), Handbook of Parkinson’s Disease, rd edn. New York: Marcel Dekker. –. Truckenbrodt, Hubert (). Semantics of intonation. In Claudia Maienborn, Klaus von Heusinger, and Paul Portner (eds.), Semantics: An International Handbook of Natural Language Meaning, vol. . Berlin: De Gruyter. –. Trueswell, John C. and Papafragou, Anna (). Perceiving and remembering events crosslinguistically: evidence from dual-task paradigms. Journal of Memory and Language, (): –. doi: ./j.jml... Trueswell, John C. and Tanenhaus, Michael K. (). Toward a lexicalist framework for constraintbased syntactic ambiguity resolution. In Charles Clifton Jr., Lyn Frazier, and Keith Rayner (eds.), Perspectives in Sentence Processing. Hillsdale, NJ: Lawrence Erlbaum Associates. –. Trueswell, John C., Tanenhaus, Michael K., and Garnsey, Susan M. (). Semantic influences on parsing: use of thematic role information in syntactic ambiguity resolution. Journal of Memory and Language, : –. Tunstall, Susanne (). The interpretation of quantifiers: semantics and processing. PhD thesis. University of Massachusetts, Amherst. Turnbull, Rory, Royer, Adam J., Ito, Kiwako, and Speer, Shari R. (). Prominence perception in and out of context. Social and Linguistic Speech Prosody: Proceedings of the th International Conference on Speech Prosody. Dublin: Trinity College Dublin. –. Turnbull, Rory, Burdin, Rachel Steindel, Clopper, Cynthia G., and Tonhauser, Judith (). Contextual predictability and the prosodic realization of focus: a cross-linguistic comparison. Language, Cognition and Neuroscience, : –. doi: ./.. Tye, Michael (). Sorites paradoxes and the semantics of vagueness. In James E. Tomberlin (ed.), Philosophical Perspectives: Logic and Language. Atascadero, CA: Ridgeview. –. Uchiyama, Hitoshi T., Seki, Ayumi, Kageyama, Hiroko et al. (). Neural substrates of sarcasm: a functional magnetic-resonance imaging study. Brain Research, (): –. doi: ./j. brainres... Uchiyama, Hitoshi T., Saito, Daisuke N., Tanabe, Hiroki C. et al. (). Distinction between the literal and intended meanings of sentences: a functional magnetic resonance imaging study of metaphor and sarcasm. Cortex, (): –. doi: ./j.cortex... Ullman, Tomer D., Xu, Yang, and Goodman, Noah D. (). The pragmatics of spatial language. In Anna Papafragou, Daniel Grodner, Daniel Mirman, and John C. Trueswell (eds.), Proceedings of the th Annual Conference of the Cognitive Science Society. Austin, TX: Cognitive Science Society. –.
Ünal, Ercenur and Papafragou, Anna (). Interactions between language and mental representations. Language Learning, (): –. doi: ./lang. Urrutia, Mabel, Moreno, Vicente, De Vega, Manuel, and Morera, Yurena (). A double task study of embodiment in counterfactuals. Paper presented at the Garachico Workshop, Symbols, Embodiment and Meaning: A debate. University of La Laguna, Tenerife. Urrutia, Mabel, De Vega, Manuel, and Bastiaansen, Marcel C. M. (). Understanding counterfactuals in discourse modulates ERP and oscillatory gamma rhythms in the EEG. Brain Research, : –. Utsumi, Akira (). Verbal irony as implicit display of ironic environment: distinguishing ironic utterances from nonirony. Journal of Pragmatics, : –. doi: ./S-()- Vaish, Amrisha, Demir, Ozlem Ece, and Baldwin, Dare (). Thirteen- and eighteen-month-old infants recognize when they need referential information. Social Development, (): –. Vallduví, Enric and Maria Vilkuna (). On rheme and kontrast. In Peter Culicover and Louise McNally (eds.), Syntax and Semantics, Vol. : The Limits of Syntax. New York: Academic Press. –. Valle Arroyo, Francisco (). Negatives in context. Journal of Verbal Learning and Verbal Behavior, (): –. doi: ./S-()-X Van Berkum, Jos J. A. (). Sentence comprehension in a wider discourse: can we use ERPs to keep track of things? In Manuel Carreiras and Charles Clifton Jr. (eds.), The On-line Study of Sentence Comprehension: Eyetracking, ERPs and Beyond. New York: Psychology Press. –. Van Berkum, Jos J. A., Brown, Colin M., Hagoort, Peter, and Zwitserlood, Pienie (). Eventrelated brain potentials reflect discourse-referential ambiguity in spoken language comprehension. Psychophysiology, : –. Van Berkum, Jos J. A., Zwitserlood, Pienie, Hagoort, Peter, and Brown, Colin M. (). When and how do listeners relate a sentence to the wider discourse? Evidence from the N effect. Cognitive Brain Research, (): –. Van Berkum, Jos J. A., Brown, Colin M., Zwitserlood, Pienie, Kooijman, Valesca, and Hagoort, Peter (). Anticipating upcoming words in discourse: evidence from ERPs and reading times. Journal of Experimental Psychology: Learning, Memory, and Cognition, (): –. doi: ./... Van Berkum, Jos J. A., Koornneef, Arnout W., Otten, Marte, and Nieuwland, Mante S. (). Establishing reference in language comprehension: an electrophysiological perspective. Brain Research, : –. Van Berkum, Jos J. A., Van den Brink, Danielle, Tesink, Cathelijne M. J. Y., Kos, Miriam, and Hagoort, Peter (). The neural integration of speaker and message. Journal of Cognitive Neuroscience, (): –. doi: ./jocn.. Van Berkum, Jos J. A., Holleman, Bregje, Nieuwland, Mante S., Otten, Marte, and Murre, Jaap (). Right or wrong? The brain’s fast response to morally objectionable statements. Psychological Science, : –. doi: ./j.-...x Van Deemter, Kees (). Not Exactly: In Praise of Vagueness. New York: Oxford University Press. Van der Henst, Jean-Baptiste, Carles, Laure, and Sperber, Dan (). Truthfulness and relevance in telling the time. Mind and Language, : –. Van der Sandt, Rob (). Presupposition projection as anaphora resolution. Journal of Semantics, : –. doi: ./jos/.. Van der Wege, Mija M. (). Lexical entrainment and lexical differentiation in reference phrase choice. Journal of Memory and Language, (: –. doi: ./j.jml... Van Dijk, Teun A. and Kintsch, Walter (). Strategies of Discourse Comprehension. New York: Academic Press. Van Dyke, Julie A. and Lewis, Richard L. (). Distinguishing effects of structure and decay on attachment and repair: a cue-based parsing account of recovery from misanalysed ambiguities. Journal of Memory and Language, (): –.
Van Dyke, Julie A. and McElree, Brian (). Retrieval interference in sentence comprehension. Journal of Memory and Language, (): –. Van Fraassen, Bas C. (). Singular terms, truth-value gaps, and free logic. Journal of Philosophy, : –. Van Fraassen, Bas C. (). Presupposition, implication, and self-reference. Journal of Philosophy, (): –. Van Gompel, Roger P. G. and Pickering, Martin J. (). Lexical guidance in sentence processing: a note on Adams, Clifton, and Mitchell (). Psychonomic Bulletin and Review, (): –. Van Gompel, Roger P. G., Gatt, Albert, Krahmer, Emiel, and Van Deemter, Kees (). Overspecification in reference: modelling size contrast effects. Poster presented at the th Architectures and Mechanisms for Language Processing Conference, University of Edinburgh. Van Herwegen, Jo, Dimitriou, Dagmara, and Rundblad, Gabriella (). Development of novel metaphor and metonymy comprehension in typically developing children and Williams syndrome. Research in Developmental Disabilities, (): –. Van Hoeck, Nicole, Begtas, Elizabet, Steen, Johan, Kestemont, Jenny, Vandekerckhove, Marie, and Van Overwalle, Frank (). False belief and counterfactual reasoning in a social environment. NeuroImage, : –. Van Kuppevelt, Jan (). Discourse structure, topicality, and questioning. Journal of Linguistics, : –. Van Petten, Cyma and Kutas, Marta (). Interactions between sentence context and word frequency in event-related brain potentials. Memory and Cognition, (): –. Van Petten, Cyma and Luka, Barbara J. (). Prediction during language comprehension: benefits, costs, and ERP components. International Journal of Psychophysiology, (): –. doi: ./ j.ijpsycho... Van Rij, Jacolien (). Pronoun processing: computational, behavioral, and psychophysiological studies in children and adults. PhD thesis. University of Groningen. Van Rij, Jacolien, Van Rijn, Hedderik, and Hendriks, Petra (). How WM load influences linguistic processing in adults: a computational model of pronoun interpretation in discourse. Topics in Cognitive Science, (): –. Van Rooij, Robert (). Questioning to resolve decision problems, Linguistics and Philosophy, : –. Van Rooij, Robert (). Vagueness, tolerance, and non-transitive entailment. Ms. Van Rooij, Robert, and Schulz, Katrin (). Exhaustive interpretation of complex sentences. Journal of Logic, Language and Information, (): –. doi: ./tops. Van Rooij, Robert and Schulz, Katrin (). Only: meaning and implicature. In Maria Aloni, Alastair Butler, and Paul Dekker (eds.), Questions in Dynamic Semantics. Amsterdam: Elsevier. –. Van Tiel, Bob and Schaeken, Walter (). Processing conversational implicatures: alternatives and counterfactual reasoning. Cognitive Science, (S): –. doi: ./cogs. Van Tiel, Bob, Van Miltenburg, Emiel, Zevakhina, Natalia, and Geurts, Bart (). Scalar diversity. Journal of Semantics, : –. Van Valin, Robert D., Jr. (). Semantic macroroles in role and reference grammar. In Rolf Kailuweit and Martin Hummel (eds.), Semantische Rollen. Tübingen: Narr. –. VanLehn, Kurt (). Determining the Scope of English Quantifiers. Technical report AI-TR-, MIT. Vasishth, Shravan and Drenhaus, Heiner (). Locality in German. Dialogue and Discourse, (): –. Vasishth, Shravan, Drenhaus, Heiner, Saddy, Douglas, and Lewis, Richard L. (). Processing negative polarity. Presentation at the th CUNY Sentence Processing Conference, University of Arizona. Vasishth, Shravan, Brüssow, Sven, Lewis, Richard L., and Drenhaus, Heiner (). Processing polarity: how the ungrammatical intrudes on the grammatical. Cognitive Science, (): –. doi: ./
Vazquez, Eduard, Gevers, Theo, Lucassen, Marcel, Van de Weijer, Joost, and Baldrich, Ramon (). Saliency of color image derivatives: a comparison between computational models and human perception. Journal of the Optical Society of America, : –. Velleman, Dan, Beaver, David, Bumford, Dylan, Destruel, Emilie, and Onea, Edgar (). ‘Yes, but . . . ’: exhaustivity and at-issueness across languages. Poster presented at PEPA , Rutgers, NJ. Velleman, Dan, Beaver, David, Destruel, Emilie, Bumford, Dylan, Onea, Edgar, and Coppock, Elizabeth (). It-clefts are IT (Inquiry Terminating) constructions. In Anca Chereches (ed.), Proceedings of the nd Semantics and Linguistic Theory Conference. Ithaca, NY: CLC Publications. –. Vendler, Zeno (). Verbs and times. Philosophical Review, (): –. Vendler, Zeno (). Each and every, any and all. Mind, : –. Vendler, Zeno (). Linguistics in Philosophy. Ithaca, NY: Cornell University Press. Vergis, Nikos and Terkourafi, Marina (). The role of the speaker’s emotional state in im/ politeness assessments. Journal of Language and Social Psychology, : –. Verkuyl, Henk J. (). On the Compositional Nature of the Aspects. Dordrecht: Reidel. Verkuyl, Henk J. (). Aspectual classes and aspectual composition. Linguistics and Philosophy, : –. Verkuyl, Henk J. (). A Theory of Aspectuality. Cambridge: Cambridge University Press. Verkuyl, Henk (). Some issues in the analysis of multiple quantification with plural NPs. In Fritz Hamm and Erhard W. Hinrichs (eds.), Plurality and Quantification. Dordrecht: Kluwer. –. Verkuyl, Henk and Van der Does, Jaap (). The semantics of plural noun phrases. In Jaap van der Does and Jan van Eijck (eds.), Quantifiers, Logic, and Language. Stanford, CA: CSLI Publications. Viberg, Åke (). The verbs of perception: a typological study. Linguistics, (): –. Vigliocco, Gabriella, Vinson, David P., Druks, Judit, Barber, Horacio, and Cappa, Stefano F. (). Nouns and verbs in the brain: a review of behavioural, electrophysiological, neuropsychological and imaging studies. Neuroscience and Biobehavioral Reviews, (): –. doi: ./j.neubiorev... Villalta, Elisabeth (). The role of context in the resolution of quantifier scope ambiguities. Journal of Semantics, : –. Villejoubert, Gaëlle, Almond, Louise, and Alison, Laurence (). Interpreting claims in offender profiles: the role of probability phrases, base-rates and perceived dangerousness. Applied Cognitive Psychology, : –. doi: ./acp. Von Fintel, Kai (). NPI-licensing, Strawson-entailment, and context-dependency. Journal of Semantics, : –. Von Fintel, Kai (). Conditional strengthening: a study in implicature. Ms. MIT. Von Fintel, Kai (). Would you believe it? The King of France is back! (Presuppositions and truthvalue intuitions.) In Marga Reimer and Anne Bezuidenhout (eds.), Descriptions and Beyond. Oxford: Oxford University Press. –. Von Fintel, Kai (). What is presupposition accommodation, again? Philosophical Perspectives, (): –. Von Heusinger, Klaus (). Reference and representation of pronouns. In Horst J. Simon and Heike Wiese (eds.), Pronouns: Grammar and Representation. Amsterdam: Benjamins. –. Von Stechow, Arnim (). Comparing semantic theories of comparison. Journal of Semantics, : –. Von Stechow, Arnim (). Focusing and backgrounding operators. In Werner Abraham (ed.), Discourse Particles. Amsterdam: John Benjamins. –. Von Stechow, Arnim (). The temporal degree adjectives früh(er)/spät(er) ‘early(er)’/‘late(r)’ and the semantics of the positive. In Anastasia Giannakidou and Monika Rathert (eds.), Quantification, Definiteness, and Nominalization. Oxford: Oxford University Press. –. Vosniadou, Stella (). Children and metaphors. Child Development, : –. Vosniadou, Stella, Ortony, Andrew, Reynolds, Ralph E., and Wilson, Paul T. (). Sources of difficulty in children’s understanding of metaphorical language. Child Development, : –.
Wagers, Matthew W., Lau, Ellen F., and Phillips, Colin (). Agreement attraction in comprehension: representations and processes. Journal of Memory and Language, (): –. Waggoner, John E. and Palermo, David S. (). Betty is a bouncing bubble: children’s comprehension of emotion-descriptive metaphors. Developmental Psychology, (): –. doi: ./ -... Wagner, Michael and Watson, Duane G. (). Experimental and theoretical advances in prosody: a review. Language and Cognitive Processes, : –. Wagner, Michael, McClay, Elise, and Mak, Lauren (). Incomplete answers and the rise-fall-rise contour. In Raquel Fernández and Amy Isard (eds.), Proceedings of the th Workshop on the Semantics and Pragmatics of Dialogue. Amsterdam: University of Amsterdam. –. Wakusawa, Keisuke, Sugiura, Motoaki, Sassa, Yuko et al. (). Comprehension of implicit meanings in social situations involving irony: a functional MRI study. NeuroImage, (): –. doi: ./j.neuroimage... Wales, R. J. and Grieve, R. (). What is so difficult about negation? Perception and Psychophysics, (): –. doi: ./BF Walker, Marilyn, Iida, Masayo, and Cote, Sharon (). Japanese discourse and the process of centering. Computational Linguistics, : –. Wallsten, Thomas S., Budescu, David V., Rapoport, Amnon, Zwick, Rami, and Forsyth, Barbara (). Measuring the vague meanings of probability terms. Journal of Experimental Psychology: General, : –. doi: ./-... Wallsten, Thomas S., Fillenbaum, Samuel, and Cox, James A. (). Base rate effects on the interpretation of probability and frequency expressions. Journal of Memory and Language, : –. doi: ./-X()- Wallsten, Thomas S., Budescu, David V., Zwick, Rami, and Kemp, Steven M. (). Preferences and reasons for communicating probabilistic information in verbal or numerical terms. Bulletin of the Psychonomic Society, : –. doi: ./BF Walsh, Clare R. and Byrne, Ruth M. J. (). Counterfactual thinking: the temporal order effect. Memory and Cognition, : –. Wang, A. Ting, Lee, Susan S., Sigman, Marian, and Dapretto, Mirella (). Developmental changes in the neural basis of interpreting communicative intent. Social Cognitive and Affective Neuroscience, (): –. doi: ./scan/nsl Ward, Gregory (). Equatives and deferred reference. Language, (): –. Ward, Gregory and Birner, Betty (). Information structure and non-canonical syntax. In Laurence R. Horn and Gregory Ward (eds.), Handbook of Pragmatics. Oxford: Blackwell. –. Ward, Peter and Sturt, Patrick (). Linguistic focus and memory: an eye movement study. Memory and Cognition, (): –. doi: ./BF Wardlow-Lane, Liane and Ferreira, Victor S. (). Speaker-external versus speaker-internal forces on utterance form: do cognitive demands override threats to referential success? Journal of Experimental Psychology: Learning, Memory, and Cognition, : –. doi: ./a Washburn, Mary Byram, Kaiser, Elsi, and Zubizarreta, Maria Luisa (). The English it-cleft: no need to get exhausted. In Klaus von Heusinger, Edgar Onea, and Malte Zimmermann (eds.), Questions in Discourse. London: Brill. Wason, Peter C. (). Response to affirmative and negative binary statements. British Journal of Psychology, : –. Wason, Peter C. (). The contexts of plausible denial. Journal of Verbal Learning and Verbal Behavior, (), –. Wason, P. C. and Johnson-Laird, P. N. (). Psychology of Reasoning: Structure and Content. London: Batsford. Watson, Duane G. (). The many roads to prominence: understanding emphasis in conversation. Psychology of Learning and Motivation, : –. Watson, Duane G., Gunlogson, Christine A., and Tanenhaus, Michael K. (). Online methods for the investigation of prosody. In Stefan Sudhoff, Denisa Lenertová, Roland Meyer et al. (eds.), Methods in Empirical Prosody Research. New York: De Gruyter. –.
Watson, Duane G., Arnold, Jennifer, and Tanenhaus, Michael K. (). Tic tac TOE: effects of predictability and importance on acoustic prominence in language production. Cognition, : –. Watts, Richard J. (). Linguistic politeness and politic verbal behaviour: reconsidering claims for universality. In Richard J. Watts, Sachiko Ide, and Konrad Ehlich (eds.), Politeness in Language. Berlin: De Gruyter. –. Watts, Richard J. (). Politeness. Cambridge: Cambridge University Press. Weber, Andrea, Braun, Bettina, and Crocker, Matthew W. (). Finding referents in time: eyetracking evidence for the role of contrastive accents. Cognition, (): –. Weber, Elke U. and Hilton, Denis J. (). Contextual effects in the interpretations of probability words: perceived base rate and severity of events. Journal of Experimental Psychology: Human Perception and Performance, : –. doi: ./-... Weber, Zach (). A paraconsistent model of vagueness. Mind, : –. Weiland, Hanna, Bambini, Valentina, and Schumacher, Petra (). The role of literal meaning in figurative language comprehension: evidence from masked priming ERP. Frontiers in Human Neuroscience, (): –. doi: ./fnhum.. Weiland-Breckle, Hanna and Schumacher, Petra B. (). Artist-for-work metonymy: type clash or underspecification? Mental Lexicon, (): –. Weissenborn, Jürgen (). L’acquisition des prépositions spatiales. Problèmes cognitifs et linguistiques [The acquisition of spatial prepositions: cognitive and linguistic problems]. In Christoph Schwarze (ed.), Analyse des prépositions. IIIme Colloque franco-allemand de linguistique théorique. Tubingen: Max Niemeyer Verlag. –. Welby, Pauline (). Effects of pitch accent position, type, and status on focus projection. Language and Speech, (): –. Wellman, Henry M. (). Making Minds: How Theory of Mind Develops. Oxford: Oxford University Press. Wellman, Henry M., Cross, David, and Watson, Julanne (). Meta-analysis of Theory-of-Mind development: the truth about false belief. Child Development, (): –. Wells, Bill and Peppé, Sue (). Ending up in Ulster: prosody and turn-taking in English dialects. In Elizabeth Couper-Kuhlen and Margret Selting (eds.), Prosody in Conversation: Interactional Studies. Cambridge: Cambridge University Press. –. Wellwood, Alexis, Xiaoxue He, Angela, Lidz, Jeffrey, and Williams, Alexander (). Participant structure in event perception: towards the acquisition of implicitly -place predicates. University of Pennsylvania Working Papers in Linguistics, (): . Wesseling, Wieneke and Van Son, Rob J. J. H. (). Timing of experimentally elicited minimal responses as quantitative evidence for the use of intonation in projecting TRPs. Presentation at Interspeech, Lisbon. Westera, Matthijs and Brasoveanu, Adrian (). Ignorance in context: the interaction of modified numerals and QUDs. In In Todd Snider, Sarah D’Antonio, and Mia Weigand (eds.), Proceedings of the th Semantics and Linguistic Theory Conference. Washington, DC: Linguistic Society of America. –. Westerbeek, Hans, Koolen, Ruud, and Maes, Alfons (). Stored object knowledge and the production of referring expressions: the case of color typicality. Frontiers in Psychology, : . ./fpsyg.. Westerståhl, Dag (). Determiners and context sets. In Johan van Benthem and Alice G. B. ter Meulen (eds.), Generalized Quantifiers in Natural Language. Dordrecht: Foris Publications. –. Wierzbicka, Anna (). Bodies and their parts: an NSM approach to semantic typology. Language Sciences, (): –. Wilder, Chris (). Antecedent containment and ellipsis. In Artemis Alexiadou, Nana Fuhrop, Paul Law, and Sylvia Lökken (eds.), FAS Papers in Linguistics. Berlin: FAS. –. Wilder, Chris (). Phrasal movement in LF: de re readings, VP-ellipsis and binding. In Kiyomi Kusumoto (ed.), Proceedings of the North East Linguistic Society . Amherst: Graduate Linguistic Student Association, University of Massachusetts. –.
Williams, Alexander (). Arguments in Syntax and Semantics. Cambridge: Cambridge University Press. Williamson, Timothy (). Vagueness. London: Routledge. Wilson, Deirdre (). Presuppositions and Non-Truth-Conditional Semantics. New York: Academic Press. Wilson, Deirdre (). Relevance and lexical pragmatics. Italian Journal of Linguistics, : –. Wilson, Deirdre (). The pragmatics of verbal irony: echo or pretence? Lingua, (): –. Wilson, Deirdre (). Irony and metarepresentation. UCL Working Papers in Linguistics, : –. Wilson, Deirdre and Carston, Robyn (). Metaphor, relevance and the emergent property issue. Mind and Language, (): –. doi: ./j.-...x Wilson, Deirdre and Carston, Robyn (). A unitary approach to lexical pragmatics: relevance, inference and ad hoc concepts. In Noel Burton-Roberts (ed.), Pragmatics. Basingstoke: Palgrave Macmillan. –. Wilson, Deirdre and Sperber, Dan (). On Grice’s theory of conversation. In Paul Werth (ed.), Conversation and Discourse. London: Croom Helm. –. Wilson, Deirdre and Sperber, Dan (). Relevance theory. In Laurence R. Horn and Gregory Ward (eds.), Handbook of Pragmatics. Oxford: Blackwell. –. Wilson, Deirdre and Sperber, Dan (). Explaining irony. In Deirdre Wilson and Dan Sperber (eds.), Meaning and Relevance. Cambridge: Cambridge University Press. –. Wilson, Frances, Papafragou, Anna, Bunger, Ann, and Trueswell, John C. (). Rapid extraction of event participants in caused motion events. In Laura Carlson, Christoph Hölscher, and Thomas F. Shipley (eds.), Proceedings of the rd Annual Conference of the Cognitive Science Society. Austin, TX: Cognitive Science Society. –. Wimmer, Heinz and Perner, Josef (). Beliefs about beliefs: representation and constraining function of wrong beliefs in young children’s understanding of deception. Cognition, (): –. Winitz, Harris, Sanders, Robert, and Kort, Joan (). Comprehension and production of the/–ez/ plural allomorph. Journal of Psycholinguistic Research, (): –. Winner, Ellen (/). The Point of Words: Children’s Understanding of Metaphor and Irony. Cambridge, MA: Harvard University Press. Winograd, Terry (a). Natural Language Understanding. New York: Academic Press. Winograd, Terry (b). Understanding natural language. Cognitive Psychology, (): –. doi: ./-()- Winter, Yoad (). Choice functions and the scopal semantics of indefinites. Linguistics and Philosophy, : –. Winter, Yoad (). Flexibility Principles in Boolean Semantics. Cambridge, MA: MIT Press. Wittgenstein, Ludwig (). Philosophische Untersuchungen [Philosophical Investigations]. Oxford: Blackwell. Wolbers, Thomas and Hegarty, Mary (). What determines our navigational abilities? Trends in Cognitive Sciences, (): –. doi: ./j.tics... Wolf, Florian, Gibson, Edward, and Desmet, Timothy (). Discourse coherence and pronoun resolution. Language and Cognitive Processes, : –. doi: ./ Wolff, Phillip and Gentner, Dedre (). The time course of metaphor comprehension. Proceedings of the Fourteenth Annual Conference of the Cognitive Science Society. Hillsdale, NJ: Lawrence Erlbaum. –. Wolff, Phillip and Gentner, Dedre (). Evidence for role-neutral initial processing of metaphors. Journal of Experimental Psychology: Learning, Memory, and Cognition, : –. Wolff, Phillip and Gentner, Dedre (). Structure-mapping in metaphor comprehension. Cognitive Science, : –. Wolter, Lynsey, Gorman, Kristen Skovbroten, and Tanenhaus, Michael K. (). Scalar reference, contrast and discourse: separating effects of linguistic discourse from availability of the referent. Journal of Memory and Language, (): –.
Wood, Justin N., Kouider, Sid, and Carey, Susan (). Acquisition of singular-plural morphology. Developmental Psychology, (): –. Wood, Linda A. and Kroger, Rolf O. (). Politeness and forms of address. Journal of Language and Social Psychology, (): –. Woodward, Amanda L. and Markman, Ellen M. (). Early word learning. In William Damon (ed.), Handbook of Child Psychology: Vol. : Cognition, Perception and Language. Hoboken, NJ: John Wiley and Sons. –. Wright, Crispin (). On the coherence of vague predicates. Synthese, : –. Wright, Crispin (). Language-mastery and the sorites paradox. In Gareth Evans and John McDowell (eds.), Truth and Meaning: Essays in Semantics. Oxford: Clarendon Press. Reprinted in Keefe and Smith (a). Wright, Crispin (). Further reflections on the sorites paradox. Philosophical Topics, : –. Reprinted (with omissions) in Keefe and Smith (a). Wynn, Karen (). Children’s acquisition of the number words and the counting system. Cognitive Psychology, : –. Xiang, Ming, Dillon, Brian, and Phillips, Colin (). Illusory licensing effects across dependency types: ERP evidence. Brain and Language, (): –. doi: ./j.bandl... Xiang, Ming, Grove, Julian, and Giannakidou, Anastasia (). Dependency-dependent interference: NPI interference, agreement attraction, and global pragmatic inferences. Frontiers in Psychology, : . Xiang, Ming, Grove, Julian, and Giannakidou, Anastasia (). Semantic and pragmatic processes in the comprehension of negation: an event related potential study of negative polarity sensitivity. Journal of Neurolinguistics, : –. Xu, Yang, Malt, Barbara C., and Srinivasan, Mahesh (). Evolution of word meanings through metaphorical mapping: systematicity over the past millennium. Cognitive Psychology, : –. doi: ./j.cogpsych... Xu, Yi and Xu, Ching X. (). Phonetic realization of focus in English declarative intonation. Journal of Phonetics, : –. Xue, Jingyang and Onea, Edgar (). Correlation between presupposition projection and atissueness: an empirical study. In Gregory Kierstead (ed.), Proceedings of the ESSLLI Workshop on Projective Content. Columbus, OH: Ohio State University. Yanılmaz, Aydoğan and Drury, John E. (). Licensing in Turkish: an ERP study. Poster presented at the th Cognitive Neuroscience Society Meeting, San Francisco. Yap, Foong Ha, Chu, Patrick Chun Kau, Yiu, Emily Sze Man et la. (). Aspectual asymmetries in the mental representation of events: role of lexical and grammatical aspect. Memory and Cognition, (): –. Yasavul, Murat (). Prosody of focus and contrastive topic in K’iche’. Ohio State University Working Papers in Linguistics, : –. Yildirim, Ilker, Degen, Judith, Tanenhaus, Michael K., and Jaeger, T. Florian (). Talker-specificity and adaptation in quantifier interpretation. Journal of Memory and Language, : –. doi: ./j.jml... Yngve, Victor H. (). On getting a word in edgewise. In Mary Ann Campbell, James Lindholm, Alice Davison et al. (eds.), Papers from the Sixth Regional Meeting, Chicago Linguistic Society. Chicago, IL: Chicago Linguistic Society. –. Yoon, J. M. D., Johnson, M. H., and Csisbra, G. (). Communication-induced memory biases in preverbal infants. Proceedings of the National Academy of Sciences of the United States of America, : –. Yoon, Si On, Koh, Sungryong, and Brown-Schmidt, Sarah (). Influence of perspective and goals on reference production in conversation. Psychonomic Bulletin and Review, (): –. doi: ./s--- Young, R. and Chase, William G. (). Additive stages in the comparison of sentences and pictures. Presentation to the Midwestern Psychological Association, Chicago.
Yurchenko, Anna, Den Ouden, Dirk-Bart, Hoeksema, Jack, Dragoy, Olga, Hoeks, John C. J., and Stowe, Laurie A. (). Processing polarity: ERP evidence for differences between positive and negative polarity. Neuropsychologia, (): –. Zadeh, Lotfi (a). The concept of a linguistic variable and its application to approximate reasoning—I. Information Sciences, :–. Zadeh, Lotfi (b). Fuzzy logic and approximate reasoning. Synthese, : –. Zadeh, Lotfi (). PRUF—a meaning representation language for natural languages. International Journal of Man-Machine Studies, : –. Zapf, Jennifer A. and Smith, Linda B. (). Meaning matters in children’s plural productions. Cognition, (): –. Zeevat, Henk (). Presupposition and accommodation in update semantics. Journal of Semantics, (): –. doi: ./jos/.. Zehr, Jérémy (). Vagueness, presupposition and truth value judgments. PhD thesis. Institut Jean Nicod, École Normale Superieure de Paris. Zehr, Jérémy and Schwarz, Florian (to appear). Entailed vs. non-entailed presuppositions: an experimental assessment. In Christopher Hammerly and Brandon Prickett (eds.), Proceedings of the th Annual Meeting of the North East Linguistic Society. Amherst, MA: GLSA Publications. Zehr, Jérémy, Bill, Cory, Tieu, Lyn, Romoli, Jacopo, and Schwarz, Florian (). Existential presupposition projection from ‘none’? An experimental investigation. In Thomas Brochhagen, Floris Roelofsen, and Nadine Theiler (eds.), Pre-Proceedings of the th Amsterdam Colloquium. Amsterdam: University of Amsterdam. –. Zehr, Jérémy, Bill, Cory, Tieu, Lyn, Romoli, Jacopo, and Schwarz, Florian (). Presupposition projection from the scope of None: universal, existential, or both? In Mary Moroney, Carol-Rose Little, Jacob Collard, and Dan Burgdorf (eds.), Proceedings of the th Semantics and Linguistic Theory Conference. Ithaca, NY: CLC Publications. –. Zehr, Jérémy, Hirsch, Aron, Bacovcin, Hezekiah, and Schwarz, Florian (). Priming local accommodation of hard triggers in disjunction. Ms. University of Pennsylvania/MIT. Zelazo, Philip David, Carter, Alice, Reznick, J. Steven, and Frye, Douglas (). Early development of executive function: a problem-solving framework. Review of General Psychology, : –. Zerbian, Sabine (). Expression of information structure in the Bantu language Northern Sotho. PhD thesis. Humboldt University, Berlin. Zerkle, Sandra A. and Arnold, Jennifer E. (). Discourse attention during utterance planning affects referential form choice. Linguistics Vanguard, (s): . doi: ./lingvan- Zheng, Mingyu and Goldin-Meadow, Susan (). Thought before language: how deaf and hearing children express motion events across cultures. Cognition, (): –. doi: ./S()- Zhu, Huichun (). A cognitive and cross-linguistic approach to polysemy. PhD thesis. Lehigh University, Bethlehem, PA. Zhu, Huichun and Malt, Barbara C. (). Cross-linguistic evidence for cognitive foundations of polysemy. In Paul Bello, Marcello Guarini, Marjorie McShane, and Brian Scassellati (eds.), Proceedings of the th Annual Conference of the Cognitive Science Society. Austin, TX: Cognitive Science Society. –. Zimmermann, Ede (). Free choice disjunction and epistemic possibility. Natural Language Semantics, : –. Zimmermann, Malte (). Contrastive focus and emphasis. Acta Linguistica Hungarica, : –. Zimmermann, Malte and Onea, Edgar (). Focus marking and focus interpretation. Lingua, : –. Zondervan, Albert Jan (). Scalar implicatures or focus: An experimental approach. PhD thesis. University of Utrecht.
Zondervan, Arjen (). Effects of Question Under Discussion and focus on scalar implicatures. In Marlies Kluck and Eric-Jan Smits (eds.), Proceedings of the Fifth Semantics in the Netherlands Day. Groningen: University of Groningen. –. Zondervan, Arjen (). Experiments on QUD and focus as a contextual constraint on scalar implicature calculation. In Uli Sauerland and Kazuko Yatsushiro (eds.), Semantics and Pragmatics: From Experiment to Theory. Basingstoke: Palgrave Macmillan. Zubin, David A. (). Discourse function of morphology: the focus system in German. In Talmy Givón (ed.), Syntax and Semantics : Discourse and Syntax. New York: Academic Press. Zubizarreta, Maria Luisa and Oh, Eunjeong (). On the Syntactic Composition of Manner and Motion. Cambridge, MA: MIT Press. Zwaan, Rolf A. (). The immersed experiencer: toward an embodied theory of language comprehension. Psychology of Learning and Motivation, , –. Zwaan, Rolf A. and Radvansky, Gabriel A. (). Situation models in language comprehension and memory. Psychological Bulletin, (): –. Zwaan, Rolf A. and Yaxley, Richard H. (a). Hemispheric differences in semantic-relatedness judgements. Cognition, : B–B. Zwaan, Rolf A. and Yaxley, Richard H. (b). Spatial iconicity affects semantic-relatedness judgments. Psychonomic Bulletin and Review, : –. Zwaan, Rolf A., Madden, Carol J., and Whitten, Shannon N. (). The presence of an event in the narrated situation affects its activation. Memory and Cognition, : –. Zwaan, Rolf A., Stanfield, Robert A., and Yaxley, Richard H. (). Language comprehenders mentally represent the shapes of objects. Psychological Science, : –. Zwarts, Frans (). Nonveridical contexts. Linguistic Analysis, : –. Zwarts, Frans (). A hierarchy of negative expressions. In Heinrich Wansing (ed.), Negation: A Notion in Focus. Berlin: De Gruyter, –. Zwarts, Frans (). Three types of polarity. In Fritz Hamm and Erhard W. Hinrichs (eds.), Plurality and Quantification. Dordrecht: Kluwer. –. Zweig, Eytan (). Number neutral bare plurals and the multiplicity agreement. Linguistics and Philosophy, : –.
I .................. A
abductive inference , absolute (frame of reference) – acceptability judgement , , –, n., –, –, –, –, , –, , , , –, –, –, , –, , –, –, – accessibility , , –, –, n., , , , –, , accommodation –, , –, –, –, –, – local accommodation –, –, – achievement n., n., n. acquisition first language , , –, , , , –, –, –, , –, , –, , –, –, , , –, –, , –, –, , –, –, , –, , , , , , –, –, , –, , –, , second language –, , ad hoc concepts , –, –, – ad hoc scales –, , adjectives of comparison – Agent – alternative semantics , alternatives –, , –, , , , , , , –, , , , , , , , –, Alzheimer’s disease , ambiguity , , –, , , , , –, –, –, , , –, , , –, –, –, –, , –, , –, analogy , , anaphora n., , , , , –, , –, , –, Antecedent Contained Deletion – anti-presupposition , – antonyms , aphasia , aptness (of metaphor) – aspect – aspectual verbs , asymmetric loss at issue , , , , , –, , –, – attitude ascription , , , , attitude verbs , n. attributive use audience design
Autism-spectrum Quotient , , –, , Autism Spectrum Disorder , , , , –, , , automatic processes
B
backchannel background –, , base rate Bayesian inference , , , , –, – BECOME –, n. bilingualism –, – binding –, –, , , – bivalence –, – borderline contradictions –, , –n. bridging inference – bunny spreading –
C
Career of Metaphor hypothesis , CAUSE , c-command , , , , , , , , – Centering Theory , , , , –, characterizing sentences , charity , child language –, –, , , , , , , –, , –, , –, , , –, , , , –, –, , –, –, –, –, , , , –, , , –n., , , , , –, –, , –, , –, , child-directed speech , , class inclusion class inclusion model of metaphor – in quantificational tasks , , –, classifiers – clefts , , , , , , , –, –, –, – coercion –, Cognitive Linguistics cognitive load , coherence , , –, –, coherence judgement – collectivity n., , , , – common ground , , , , , , , , , , –, , , , communicative lexicon hypothesis –
comparison class – competence competence assumption conceptual metaphor theory , – conceptual pacts , , conceptual spaces conditional perfection –, conditionals , , –, , –, –, , –, –, , –, –, , , , – conjunction fallacy connectionism , , constraints –, , , , , – Construal Level Theory – containment relations –, –, context change potential , continuer contrastive inference , , , contrastive stress , , , , , –, , – conventionality –, , – Conversation Analysis , conversion errors , – cooperativity , –, , co-presence , , , , corpus linguistics , , , , , –, , , , , –, , –, – cost , , , , , – countability – count nouns –, – counterfactuals – counterhysteresis – cues , –, –, , –, , , , –, –, , –, , –, –, , , , –, – CULMINATE
D
deafness , , defaults , , –, –, , –, definites –, , –, , , –, , –, –, degree modifiers –, , denominal verbs – diachronic change , –, – Direct Access –, directionality – Discourse Representation Theory , n., , , , – disjunction , , , –, , , , –, , –, , , –, distributivity n., –, –, , –, , –, DO –, domain restriction , , , , , domain widening , – downward entailing see entailment direction durativity –,
dynamic semantics , –, , dynamicity –
E
each –, –, , , , –, , echoic use –, n., ecological validity , , – EEG , , , – egocentrism , – ellipsis , –, embodiment –, emergent properties enhanced contrast entailment direction , , –, –, , –, epistemic state , , , , , –, –, epistemic step , , epistemic theory of vagueness –, , , , equatives , ERPs , , , , –, , –, –, , , –, , , events –, , –, n., , every –, –, –, , , –, –, , excluded middle executive functions , , , –, , , Exemplar theory exhaustification , , , –, exhaustivity , , , –, –, –, –, , , –, existential closure , Expectancy Hypothesis , expectation –, , , , –, , eye gaze –, eye-tracking , , , , , , –, n., , –, , , , , –, , , –, –, –, –, n., –, , , , , , –, , , –, , , , , , , , , , , ,
F
face , –, – factives , , , , , , –, – false belief , –, –, –, – family of sentences , faultless disagreement – figure (location) – fMRI , , , , , , –, focus , , , , , , , , –, –, –, –, –, – focus projection – Frames of Reference , , –, – framework theories free choice , , –, , , n. fuzzy logic , , , –, , –,
G
game theory – garden path , , , – Generalized Quantifier Theory generative models – genericity –, , , gesture , , , –, , , – goal (path) gradability , – Graded Salience hypothesis –, , , –, grammatical theory of scalar implicature , Gricean pragmatics –, , , , , –, –, n., –, , –, –, –, , –, –, Cooperative Principle , , , Maxim of Quality , Maxim of Quantity , , , , –, maxims , , , , , , ground (location) –
H
habitual , , ‘Hey, wait a minute’ test historical linguistics , –, – homonymy –, –, –, – hyperbole , , , –, hysteresis n., n.
I
idioms , , –, ignorance inferences , –, –, –, – illusory licensing – implicature: –, , , –, , –, , –, , , –, –, –, – conventional implicature , , , embedded implicatures , , , –, –, generalized conversational implicatures –, politeness implicatures , relevance implicatures – scalar implicatures –, , –, –, , –, , , –, –, , , –, , , , , , –, , –, –, , –, , implicit causality –, –, –, – implicit consequentiality indirectness informational privilege –, – information structure , , –, –, , – informativeness , , , , , , , , –, –, , inhibitory control , –, , inquisitive semantics n. Instrument , – intonation see prosody intention , , , –, –, , , –, , , , –, , , , –, ,
interactive activation intrinsic (frame of reference) – irony , –, , , isomorphism , –, iteratives , , –,
J
joint attention , –,
K
kinds
L
Latent Semantic Analysis lexical decision task , , , –, , , lexical flexibility – Lexicalization Constraint lexical precedents lexical rules – literal-first interpretation , , –, –, –, – Local Contexts Logic of Paradox n. loose use , , , ,
M
manner (of motion) –, manner verbs – markedness –, , , , n., –, , , , Marr’s levels of analysis mass nouns , – matching bias – Maximize Presupposition –, measure functions membership functions –, –, mental models –, –, , mental spaces – metaphor , , , , , , –, , , extended metaphor –, – metonymy , , –, , –, , , – modality , –, , –, , , n., n., –, , , , , , sensory modality , modularity monotonicity , , n., , , –, – mouse-tracking , , movement (syntax) , –, – multiplicity inference –, , – multivalence , – mutual knowledge ,
N
N , –, , , –, , –, , , , –, , , –,
negation , , , , –, –, , , n.–, , n., n., , , , , , , –, – double negation – metalinguistic negation , – negative polarity items (NPIs) , , –, – neo-Gricean , , newness –, –, –, non-contradiction non-veridicality , – numerals –, –, –, –, –, –, –, –, , modified numerals n., –
Principle of Interpretive Economy Principle of Scope Interpretation probabilistic pragmatics –, , –, –, –, – probability –, –, , –, – Processing Scope Economy proffered content , pronouns , , –, –, , –, –, , , –, prosody , , , –, , , –, –, –, –, , –, , –, prototypes , , , , , pseudoclefts –
O
quantification –, , –, , , –, , , , , , , , , n., –, –, , , , –, , , quantifier floating Quantifier Raising –, , – quantifier spreading –, n., – questions , –, , , –, , question-answer congruence , Question Under Discussion (QUD) , , –, –, , –, , , , –, , , –, –, , –, –, , –, , –, –, –
obligatory local effect , – only –, , , –, , , , –, , –, , order effects ,
P
P , , , , –, parallelism (of processing) Parkinson’s disease , passives –, –, –, – path (of motion) –, Patient , – perspective taking –, , , –, –, picture matching task , , , , –, –, plausible deniability plausible denial/dissent contexts –, , –, , plurality , –, –, , –, –, polarity , , –, , , politeness , , , – polysemy , , , –, , possessives , , , – possible worlds , , –, – poverty of the stimulus – pragmatic tolerance , predicate transfer prediction , , , – presupposition: , , –, , , , , , , , , –, , , –, conditional presupposition , informative presupposition obligatory presupposition – projection , , –, –, –, , , trigger classification –, – triggering –, , – Pretense Theory , n. priming –, –, , n., , , , , –, , , , –, , , , Principle B , Principle C , –
Q
R
Ramsey test range requirement Rational Speech Act model , – rationale clauses – recency –, – reference –, , –, , –, –, , , reference resolution , –, – registrations relative (frame of reference) –, – relevance , , , , , , –, , , –, , , Relevance Theory , , , , , Repeated Name Penalty , , –, result verbs – Right Frontier Constraint –
S
salience , , –, , –, , , –, , –, , scalar adjectives , –, , , scope , –, –, , nn.–, –, , , , , – inverse scope –, self-paced reading , n., , , , n., , , , –, –, , –, , –, , n., –, –, –, , , –, – semelfactives n.
severity effect – simile , , , – situation model – sketch –n. sluicing – sociolinguistics , , – sorites paradox , , –, , , – source (path) spatial terms – Specific Language Impairment (SLI) speech acts , , , , , , , , , squeamishness Stalnaker’s Bridge stereotypes , , , , stops-making-sense task –, , –, , –, – strategic processes strict/tolerant evaluation –, –, , – Strongest Meaning Hypothesis –, –, , –, n., , subitizing – subjectivity –, subset problem – subvaluation , , –, –, , , , supervaluation , –, –, –, , , , suppositional theory surprisal , , syllogisms – symmetry problem
thematic roles –, , , –, Theory of Mind , , –, , –, , , – Theory theory together – topicality , , –, –, Transition Relevance Place trivalent logic –, , – truth-conditional meaning , , –, , , , , , , , , , , –, , –, , , , , , , , –, , truth-value judgement , , –, –, , , –, –, , , , , , , –, –, –, , –, Turn Construction Unit turn-taking – type–shifting
T
W
taboo words , telicity – temporality – tense , , , , testimonial injustice
U
underspecification –, , understatement , upward entailing see entailment direction
V
vagueness , n., –, –, , , –, n. vagueness-as-ignorance –, Vagueness Principle variation inference – verbal root – Which Outcome approach – Williams syndrome , Winograd schema , – working memory , , –, , , , ,
OXF ORD HANDBO O K S I N LI N G U I S T I C S THE OXFORD HANDBOOK OF AFRICAN AMERICAN LANGUAGE Edited by Sonja Lanehart
THE OXFORD HANDBOOK OF APPLIED LINGUISTICS Second edition Edited by Robert B. Kaplan
THE OXFORD HANDBOOK OF ARABIC LINGUISTICS Edited by Jonathan Owens
THE OXFORD HANDBOOK OF CASE Edited by Andrej Malchukov and Andrew Spencer
THE OXFORD HANDBOOK OF CHINESE LINGUISTICS Edited by William S-Y. Wang and Chaofen Sun
THE OXFORD HANDBOOK OF COGNITIVE LINGUISTICS Edited by Dirk Geeraerts and Hubert Cuyckens
THE OXFORD HANDBOOK OF COMPARATIVE SYNTAX Edited by Gugliemo Cinque and Richard S. Kayne
THE OXFORD HANDBOOK OF COMPOSITIONALITY Edited by Markus Werning, Wolfram Hinzen, and Edouard Machery
THE OXFORD HANDBOOK OF COMPOUNDING Edited by Rochelle Lieber and Pavol Štekauer
THE OXFORD HANDBOOK OF COMPUTATIONAL LINGUISTICS Edited by Ruslan Mitkov
THE OXFORD HANDBOOK OF CONSTRUCTION GRAMMAR Edited by Thomas Hoffman and Graeme Trousdale
THE OXFORD HANDBOOK OF CORPUS PHONOLOGY Edited by Jacques Durand, Ulrike Gut, and Gjert Kristoffersen
THE OXFORD HANDBOOK OF DERIVATIONAL MORPHOLOGY Rochelle Lieber and Pavol Štekauer
THE OXFORD HANDBOOK OF DEVELOPMENTAL LINGUISTICS Edited by Jeffrey Lidz, William Snyder, and Joe Pater
THE OXFORD HANDBOOK OF ELLIPSIS Edited by Jeroen van Craenenbroeck and Tanja Temmerman
THE OXFORD HANDBOOK OF ERGATIVITY Edited by Jessica Coon, Diane Massam, and Lisa deMena Travis
THE OXFORD HANDBOOK OF EVIDENTIALITY Edited by Alexandra Y. Aikhenvald
THE OXFORD HANDBOOK OF EXPERIMENTAL SEMANTICS AND PRAGMATICS Edited by Chris Cummins and Napoleon Katsos
THE OXFORD HANDBOOK OF GRAMMATICALIZATION Edited by Heiko Narrog and Bernd Heine
THE OXFORD HANDBOOK OF HISTORICAL PHONOLOGY Edited by Patrick Honeybone and Joseph Salmons
THE OXFORD HANDBOOK OF THE HISTORY OF ENGLISH Edited by Terttu Nevalainen and Elizabeth Closs Traugott
THE OXFORD HANDBOOK OF THE HISTORY OF LINGUISTICS Edited by Keith Allan
THE OXFORD HANDBOOK OF INFLECTION Edited by Matthew Baerman
THE OXFORD HANDBOOK OF INFORMATION STRUCTURE Edited by Caroline Féry and Shinichiro Ishihara
THE OXFORD HANDBOOK OF JAPANESE LINGUISTICS Edited by Shigeru Miyagawa and Mamoru Saito
THE OXFORD HANDBOOK OF LABORATORY PHONOLOGY Edited by Abigail C. Cohn, Cécile Fougeron, and Marie Hoffman
THE OXFORD HANDBOOK OF LANGUAGE AND LAW Edited by Peter Tiersma and Lawrence M. Solan
THE OXFORD HANDBOOK OF LANGUAGE AND SOCIETY Edited by Ofelia García, Nelson Flores, and Massimiliano Spotti
THE OXFORD HANDBOOK OF LANGUAGE EVOLUTION Edited by Maggie Tallerman and Kathleen Gibson
THE OXFORD HANDBOOK OF LANGUAGE POLICY AND PLANNING Edited by James W. Tollefson and Miguel Pérez-Milans
THE OXFORD HANDBOOK OF LEXICOGRAPHY Edited by Philip Durkin
THE OXFORD HANDBOOK OF LINGUISTIC ANALYSIS Second edition Edited by Bernd Heine and Heiko Narrog
THE OXFORD HANDBOOK OF LINGUISTIC FIELDWORK Edited by Nicholas Thieberger
THE OXFORD HANDBOOK OF LINGUISTIC INTERFACES Edited by Gillian Ramchand and Charles Reiss
THE OXFORD HANDBOOK OF LINGUISTIC MINIMALISM Edited by Cedric Boeckx
THE OXFORD HANDBOOK OF LINGUISTIC TYPOLOGY Edited by Jae Jung Song
THE OXFORD HANDBOOK OF LYING Edited by Jörg Meibauer
THE OXFORD HANDBOOK OF MODALITY AND MOOD Edited by Jan Nuyts and Johan van der Auwera
THE OXFORD HANDBOOK OF MORPHOLOGICAL THEORY Edited by Jenny Audring and Francesca Masini
THE OXFORD HANDBOOK OF NAMES AND NAMING Edited by Carole Hough
THE OXFORD HANDBOOK OF PERSIAN LINGUISTICS Edited by Anousha Sedighi and Pouneh Shabani-Jadidi
THE OXFORD HANDBOOK OF POLYSYNTHESIS Edited by Michael Fortescue, Marianne Mithun, and Nicholas Evans
THE OXFORD HANDBOOK OF PRAGMATICS Edited by Yan Huang
THE OXFORD HANDBOOK OF REFERENCE Edited by Jeanette Gundel and Barbara Abbott
THE OXFORD HANDBOOK OF SOCIOLINGUISTICS Second Edition Edited by Robert Bayley, Richard Cameron, and Ceil Lucas
THE OXFORD HANDBOOK OF TABOO WORDS AND LANGUAGE Edited by Keith Allan
THE OXFORD HANDBOOK OF TENSE AND ASPECT Edited by Robert I. Binnick
THE OXFORD HANDBOOK OF THE WORD Edited by John R. Taylor
THE OXFORD HANDBOOK OF TRANSLATION STUDIES Edited by Kirsten Malmkjaer and Kevin Windle
THE OXFORD HANDBOOK OF UNIVERSAL GRAMMAR Edited by Ian Roberts
THE OXFORD HANDBOOK OF WORLD ENGLISHES Edited by Markku Filppula, Juhani Klemola, and Devyani Sharma