Inquiries in Philosophical Pragmatics: Issues in Linguistics (Perspectives in Pragmatics, Philosophy & Psychology, 28) 3030566951, 9783030566951

Citation preview

Perspectives in Pragmatics, Philosophy & Psychology 28

Fabrizio Macagno Alessandro Capone  Editors

Inquiries in Philosophical Pragmatics Issues in Linguistics

Perspectives in Pragmatics, Philosophy & Psychology Volume 28 Editor-in-Chief Alessandro Capone, University of Messina, Italy Consulting Editors Keith Allan, Monash University, Australia Louise Cummings, The Hong Kong Polytechnic University, Hong Kong Wayne A. Davis, Georgetown University, Washington, USA Igor Douven, University of Paris-Sorbonne, France Yan Huang, University of Auckland, New Zealand Istvan Kecskes, State University of New York at Albany, USA Franco Lo Piparo, University of Palermo, Italy Antonino Pennisi, University of Messina, Italy Francesca Santuli, Ca’ Foscari University of Venice, Italy Editorial Board Members Noel Burton-Roberts, University of Newcastle, Newcastle, Australia Brian Butler, University of North Carolina, Asheville, USA Marco Carapezza, University of Palermo, Palermo, Italy Felice Cimatti, Università della Calabria, Cosenza, Italy Eros Corazza, Carleton University, Ottawa, Canada Marcelo Dascal, Tel Aviv University, Tel Aviv, Israel Michael Devitt, City University of New York, New York, USA Frans van Eemeren, University of Amsterdam, Amsterdam, The Netherlands Alessandra Falzone, University of Messina, Messina, Italy Neil Feit, State University of New York, Fredonia, USA Alessandra Giorgi, Ca’ Foscari University of Venice, Venice, Italy Larry Horn, Yale University, New Haven, USA Klaus von Heusinger, University of Stuttgart, Stuttgart, Germany Katarzyna Jaszczolt, University of Cambridge, Cambridge, UK Ferenc Kiefer, Hungarian Academy of Sciences, Budapest, Hungary Kepa Korta, ILCLI, Donostia, Spain Ernest Lepore, Rutgers University, New Brunswick, USA Stephen C. Levinson, Max Planck Institute for Psycholinguistics, Nijmegen,  The Netherlands Fabrizio Macagno, New University of Lisbon, Lisbon, Portugal

Jacob L. Mey, University of Southern Denmark, Odense, Denmark Pietro Perconti, University of Messina, Messina, Italy Francesca Piazza, University of Palermo, Palermo, Italy Roland Posner, Berlin Institute of Technology, Berlin, Germany Mark Richard, Harvard University, Cambridge, USA Nathan Salmon, University of California, Santa Barbara, USA Stephen R. Schiffer, New York University, New York, USA Michel Seymour, University of Montreal, Montreal, Canada Mandy Simons, Carnegie Mellon University, Pittsburgh, USA Timothy Williamson, University of Oxford, Oxford, UK Anna Wierbizcka, Australian National University, Canberra, Australia Dorota Zielinska, Jesuit University of Philosophy and Education Ignatianum, Kraków, Poland More information about this series at http://www.springer.com/series/11797

Fabrizio Macagno  •  Alessandro Capone Editors

Inquiries in Philosophical Pragmatics Issues in Linguistics

Editors Fabrizio Macagno Universidade Nova de Lisboa Lisbon, Portugal

Alessandro Capone University of Messina Messina, Italy

ISSN 2214-3807     ISSN 2214-3815 (electronic) Perspectives in Pragmatics, Philosophy & Psychology ISBN 978-3-030-56695-1    ISBN 978-3-030-56696-8 (eBook) https://doi.org/10.1007/978-3-030-56696-8 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 Chapter 12 is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/). For further details see license information in the chapter. The images or other third party material in this book are included in the book’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the book’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Contents

 Introduction. Inquiries in Philosophical Pragmatics: Linguistic and Theoretical Issues ������������������������������������������������������������������    1 Fabrizio Macagno and Alessandro Capone  Ceteris Paribusiness: On the Power of Salient Exceptions��������������������������    7 Laurence R. Horn I like you may actually implicate ‘I love you’: A reconsideration of some scalar implicatures����������������������������������������������   33 Yan Huang  Pragmatics and Grammar as Sources of Temporal Ordering in Discourse: The Case of And������������������������������������������������������������������������   53 Kasia M. Jaszczolt and Roberto B. Sileo  Inferential Patterns of Emotive Meaning������������������������������������������������������   83 Fabrizio Macagno and Maria Grazia Rossi Seniors, Older People, the Elderly, Oldies, and Old People: What Language Reveals about Stereotypes of Ageing in Australia������������  111 Keith Allan, Réka Benczes, and Kate Burridge  How to Be Impolite (or Worse) in an Artificial Auxiliary Language����������  127 Alan Reed Libert  Pragmemes at the Market Place��������������������������������������������������������������������  133 Alessandro Capone  Pragmatics of Self-Reference Pronouns in Capital Trials����������������������������  155 Krisda Chaemsaithong  Navigating Narrative Subjectivity in Schizophrenia: A Deictic Network Analysis of Narrative Viewpoints of Self and Other����������������������������������������������������������������������������������������������  169 Linde van Schuppen, José Sanders, and Kobie van Krieken v

vi

Contents

 Pragmatic Perspective of Literary Texts for Children ��������������������������������  197 Alla Tsapiv  When Both Utterances and Appearances are Deceptive: Deception in Multimodal Film Narrative������������������������������������������������������  205 Marta Dynel

Introduction. Inquiries in Philosophical Pragmatics: Linguistic and Theoretical Issues Fabrizio Macagno and Alessandro Capone

Together with the volume “Inquiries in philosophical pragmatics: Theoretical developments,” this book collects selected contributions to the conference Pragmasophia II held in Lisbon in 2018. In the first volume, twelve chapters outlined a path that characterizes pragmatic research, starting from the philosophical grounds and leading to new explorations and new interconnections with other disciplines, in particular argumentation theory and discourse analysis. From a thematic perspective, the first volume moved from the classical topics of quantifiers, intentions, and common knowledge to the theoretical challenges related to the phenomena of pure indexicals, deferred reference, explicatures and indirect reports, and finally to the inquiry into classical topics using mixed theoretical approaches. Thus, metaphors are considered from the perspective of their explanatory or argumentative effects, relevance is analyzed taking into account its measurability, and stereotypes are studied as vehicles and content of implicit strategies. The path from theory to practice runs through the twelve chapters from the strictly philosophical discussions of the first papers to the applications of the pragmatic insights to specific fields of practice, such as medical discourse, artificial intelligence, and political discourse and advertising. The eleven essays that this second volume collects follow a similar thread. The analytical emphasis is now placed on how a linguistic structure or expression manifests a pragmatic phenomenon, and not the foundational and essential question of

F. Macagno () Instituto de Filosofia da Nova (IFILNOVA), Facultade de Ciências Sociais e Humanas, Universidade Nova de Lisboa, Lisbon, Portugal e-mail: [email protected] A. Capone Department of Cognitive Sciences, Educational and Cultural Studies (CSECS), University of Messina, Messina, Italy © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 F. Macagno, A. Capone (eds.), Inquiries in Philosophical Pragmatics, Perspectives in Pragmatics, Philosophy & Psychology 28, https://doi.org/10.1007/978-3-030-56696-8_1

1

2

F. Macagno and A. Capone

what a specific pragmatic phenomenon is. This different approach runs through all the chapters, which are ordered considering their relationship with the specificity of the context. The first three papers, devoted to the topics of prototype-based generalizations, scalar implicatures, and temporal ordering, propose new insights into pragmatic phenomena considering linguistic behavior independent of the social activity in which it is placed. In the last papers, the analytical and empirical goals become predominant. Here, the theoretical advances are considered as instruments for analyzing specific texts, such as marketplace interactions, courtroom speech, schizophrenic discourse, narratives of literary texts for children or deception in movies. In between  these two groups of contributions, three papers devoted to emotive or loaded words bridge the passage from more theoretical to more empirical and context-­oriented works. These three chapters address the theoretical topic of derogative force of slurs or ethical terms (Stevenson, 1937) considering specific social contexts (political discourse, national surveys, and artificial languages) and investigating the strategies of manipulation and quasi-definition of the “emotive meaning” of specific terms in political discourse, ageing stereotypes manifested in the use of specific terminology, and impoliteness expressed in Esperanto. In the first chapter, Horn analyzes through linguistic lenses a problem that is essential to pragmatics, argumentation, and logic, namely the so-called “plausible” or prototype-based generalizations. In logic and argumentation, this topic is at the basis of the structure of the so-called natural inferences, describing our everyday reasoning (Walton, 1995, 2001). By describing the behavior of personal pronouns having sex-neutral references, the prototype gender of some words (such as the doctor or the informant), and bare plurals, Horn brings to light the cancellability conditions of such generalizations, contributing to the debate on their presuppositional or inferential nature. In particular, he builds on the principles of salience and relevance, which rule out some default inferences when they are made incoherent with the context to propose a new perspective on the phenomenon. Framed in the words of the popular disctum that “it is easier to ask forgiveness than to ask permission,” Horn formulates the principle that governs prototype-based generalization: the discourse sequencing and processing of generalization and exception follows a left-to-­ right asymmetry. A ceteris paribus generalization can be drawn even if later on a subsequently acknowledged exception narrows the domain of that generalization retroactively, but if the salient exception comes first, the default generalization it apparently contradicts cannot be asserted. The principles governing scalar implicatures are the subject matter of Huang’s chapter, dedicated to a very little investigated issue, marked scalar implicatures. Huang describes this type of implicature as dependent on the non-use of the semantically or informationally stronger alternatives that could have been used in the context (“some students got an A” instead of “all the students got an A”), which results in the negation of the (semantically, informationally) stronger option (“not all the students got an A”). However, this type of implicature fails to account for some inferences, called “non-canonical scalar implicatures” that the author exemplifies through excerpts from different languages. He observes that in some contexts “I like you” implies the stronger “I love you” in Chinese, a general noun (“person”) is used

Introduction. Inquiries in Philosophical Pragmatics: Linguistic and Theoretical Issues

3

in Malagasy to refer to the more specific (informational) speaker’s husband, while in face-threatening contexts weaker expressions are used to imply stronger ones in Chinese, Japanese, and English. Huang explains such non-canonical scalar implicatures by introducing an epistemic dimension not only in the nature of the implicature, but also in the reasons governing its triggering. Thus, the choice of an alternative can be determined by either epistemic reasons, or non-epistemic ones (it would be impolite, indiscrete, unethical, immoral, and/or politically/ideologically incorrect to use a semantically stronger alternative), which leads to an implicature “from weak to stronger.” The third chapter addresses the disputed source of the default temporal interpretation of sentential and coordination (“Mary broke her leg and went to hospital”). The temporal meaning of conjunction is associated by various theories with different factors, including aspectual class of the connected eventualities, grammatical tenses, or Gricean implicatures. Jaszczolt and Sileo describe and compare critically the dominant views considering the factors leading to a temporal interpretation in a non-biased context, and the role that assumed scenarios and biasing contexts play in such a reading. By combining theoretical analyses with the results of five case studies, the authors propose an integrated perspective centered on the notion of eventhood and the degrees thereof. The theoretical linguistic analyses proposed in the first three papers are followed by three contributions that investigate different aspects of emotive meaning, each taking into account a specific communicative context. Macagno and Rossi introduce this thematic section by addressing the notion of emotive meaning from an argumentation theory perspective. They propose an inferential approach to this phenomenon, where the “expressive force” of “loaded” or “emotive” words such as slurs, pejorative words, or ethical terms is conceived as a defeasible and automatic or automatized evaluative and intended inference commonly associated with their use. Such inferences are described and distinguished through argumentation schemes (Walton, Reed, & Macagno, 2008), representations of the common forms of natural inference. The automatic evaluative inferences are shown to be aspects of the connotation of such loaded terms, which can be modified and manipulated by recontextualization strategies. Through the analysis of the past US presidential campaign, the authors bring to light the relationship between emotive meaning, inferences, and contexts of use. In the following chapter, Allan, Benczes, and Burridge map the evaluative inferences and the stereotypes associated with the nominal phrases that describe ageing in Australia. By running an online survey, the authors analyze the common associations triggered by the use of the NP “seniors,” “older people,” “old people,” “oldies,” and “the elderly.” They highlight how, while the first two expressions lead to positive inferences, related to positive personal characteristics of health, experience, and wisdom, the referents of “old people” and “oldies” are not described in an evaluative way. Finally, “the elderly” is found to carry negative stereotypes connected to frail health or incompetence. The last contribution in this thematic section concerns the existence and the development of slurs and pejoratives in artificial auxiliary languages. Libert

4

F. Macagno and A. Capone

observes that slurs can be hardly expected to exist in such languages, which were artificially created to facilitate international communication. The author argues that auxiliary languages were not designed to insult (on the contrary, they were thought to prevent impoliteness); moreover, the expressive force of a pejorative would be a coded component thereof (bad-x), which would defuse its derogatory effect. However, this chapter provides some clear examples of slurs and pejoratives drawn from Esperanto corpora, which suggests that the use of a language for human communication naturally results in the expression of human behavior, which includes verbal aggression and attacks. The relationship between pragmatics and human behavior (and conversational settings) becomes predominant in the last section of the volume, which collects papers in which pragmatics is regarded as an instrument for analyzing a specific context and human activity. The seventh chapter, “Pragmemes at the marketplace,” develops this transition from theoretical advances to empirical analyses. Capone builds on the theory of pragmemes, or “culturally situated speech acts” (Mey, 2001, 2016), to capture not only the illocutionary force of utterances expressed in a specific context, but higher order communicative intentions that presuppose communicative competence, and cultural and textual habits and rules. Capone observes that at the marketplace different types of pragmemes can be found, each characterized by a mix of poetic, conative, argumentative, and phatic functions. The author points out how the understanding of the communicative intention expressed by a pragmeme requires pragmatic competence that involves an accurate knowledge of the context, background assumptions, and ethical norms. By describing the specific pragmeme “selling fish at the marketplace,” he illustrates these complex interrelations in its different manifestations. The use of pragmatic instruments for accounting for linguistic strategies in specific conversational settings is the purpose of Chaemsaithong’s chapter, in which he provides a corpus-based study on the use of self-reference pronouns in the penalty phase of capital trials. The author regards self-reference as a persuasive instrument through which the speaker can manipulate the conversational situation, involving other actors (primarily the jurors) in his or her viewpoint or dissociating themselves from a certain group. For example, in closing statements, defense attorneys tend to use more the personal pronoun “we” than “I,” both considering their absolute frequencies and the relative ones (compared to the prosecution). On this view, pronouns become instruments for framing jurors’ perceptions, creating solidarity with or distance from the defendant and signal or foreground differences. Pragmatics, and in particular the relationship between deixis and perception in schizophrenic discourse, is the focus of van Schuppen, Sanders and van Krieken’s paper. The basic assumption is that language provides an instrument for understanding how the self, the other, and the world are experienced by the people diagnosed with this disorder. By analyzing the markers of perspective-taking in schizophrenic patients’ narratives, the authors investigate the nature and complexity of their perspectivization issues. Through a model for the analysis of perspective in conversational discourse called “Deictic Navigation Network,” the authors capture how the use of tense shifts, spatial deixis, and speech and thought reports constructs the

Introduction. Inquiries in Philosophical Pragmatics: Linguistic and Theoretical Issues

5

speakers’ projections of themselves and the others in the narrative domain. The study of a corpus based on this methodology shows how people affected by schizophrenia tend to fail to signal through their language that they acknowledge the hearer’s separate and different perspective (implying a difficulty to recognize the others as bearers of independent viewpoints), and to maintain viewpoint stability, separating between viewpoints anchored in different domains. The linguistic strategies for creating perlocutionary effects in literary texts for children are the focus of Tsapiv’s contribution. Her starting point is the assumption that a specific perlocutionary effect can be produced through a definite narrative strategy, developed at the lexical, grammatical, semiotic, and narrative level. In this chapter, the author shows how different effects are pursued through different types of narratives and metaphors. The last chapter in this twofold collection of essays is devoted to one of the frontiers of pragmatics, multimodal film analysis, situating the discussion and analysis of classical philosophical issues within the study of a very specific communicative activity. In “When both utterances and appearances are deceptive: Deception in multimodal film narrative,” Dynel takes into account a classical philosophical topic frequently discussed in pragmatics and linguistics, lying and deception (Meibauer, 2014). However, deception becomes an artistic phenomenon in movies, which needs to be analyzing considering not only the verbal and pragmatic dimension, but also non-verbal communication, multimodal analysis, and studies of fictional narrative. The author proposes an eclectic approach to this multidisciplinary and multifaceted phenomenon, showing how deception is the result not only of the characters’ messages (verbal and not verbal), but also of the production crew (constricting a fictional world that invites false make beliefs in viewers), the intradiegetic narrator that reports on the characters’ interactions, and the interaction between the extradiegetic and the intradiegetic narrator. Through the analysis of multimodal deceptions in some episodes of Dr. House and Hitchcock’s Stage Fright, the author discusses the philosophical distinction between covert ambiguity and deceptively withholding information, and the relationship between objective facts and individual mental experiences. This journey in two volumes from philosophical approaches to pragmatics to its new frontiers illustrate the many directions and methods that characterize this field of study, each characterized by a different role of and focus on the context, and a different conception of the context itself. Thus, the conditions, the constraints, and the regularities of the language in use become methods for studying how language is used and what its use in specific circumstances can tell us. Acknowledgments  This work was supported by the Fundação para a Ciência e a Tecnologia (research grant no. PTDC/FER-FIL/28278/2017).

6

F. Macagno and A. Capone

References Meibauer, J. (2014). Lying at the semantics-pragmatics interface. Berlin, Germany: De Gruyter. Mey, J. (2001). Pragmatics. An introduction. Oxford, UK: Blackwell. Mey, J. (2016). Why we need the pragmeme, or: Speech acting and its peripeties. In K. Allan, C. Alessandro, & K. Istvan (Eds.), Pragmemes and theories of language use (pp. 133–140). Cham, Switzerland: Springer. Stevenson, C. (1937). The emotive meaning of ethical terms. Mind, XLVI, 14–31. https://doi. org/10.1093/mind/XLVI.181.14 Walton, D. (1995). Argumentation schemes for presumptive reasoning. Mahwah, NJ: Routledge. Walton, D. (2001). Abductive, presumptive and plausible arguments. Informal Log, 21, 141–169. https://doi.org/10.22329/il.v21i2.2241 Walton, D., Reed, C., & Macagno, F. (2008). Argumentation schemes. New York, NY: Cambridge University Press.

Ceteris Paribusiness: On the Power of Salient Exceptions Laurence R. Horn

Constraints on he/man Language: McConnell-Ginet’s Generalization For over four decades feminist linguists and philosophers of language have addressed the semantic, cognitive, and political factors associated with gender asymmetries in nominal and pronominal choice (see Miller & Swift, 1977 for an early survey and Saul, 2017 for a recent overview). The sociolinguistic spotlight has focused on the history, extent, and implications of the prescriptively sanctioned use of man and he for sex-neutral reference—he/man language in Martyna (1980)’s term. Bare singular and simple indefinite man in (1a, b) exemplify this use, while the bare singulars in (2) yield the male-specific meaning exhibited by the man or that man. 1. a. Man is mortal. Man is the animal that laughs (cries, speaks, etc.) The proper study of mankind is man. [Pope] b. What a piece of work is a man. [Hamlet II.ii.317] No man is an island. [Donne] 2. Man is enemy to virginity.  [All’s Well That Ends Well, I.i.125] Man’s heart must be in his head. Woman’s head must be in her heart.  [Coleridge] Higgamus hoggamus, woman’s monogamous/Hoggamus higgamus, man is polygamous.  [attributed variously to William James, Dorothy Parker, and Ogden Nash]

L. R. Horn () Yale University, New Haven, CT, USA e-mail: [email protected] © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 F. Macagno, A. Capone (eds.), Inquiries in Philosophical Pragmatics, Perspectives in Pragmatics, Philosophy & Psychology 28, https://doi.org/10.1007/978-3-030-56696-8_2

7

8

L. R. Horn

It has been argued by McConnell-Ginet (1979), Nunberg (1992), and Horn and Kleinedler (2000), inter alia, that the occurrences of man in (1) are neither truly sex-­neutral nor exclusively male-referential but correspond instead to a Roschian prototype (Rosch, 1973, 1978). Given the traditional status of adult males as the prototype or default members of the category HUMAN in our culture (and others), it follows that when and only when an adult male exemplar can be used to verify a proposition, the putatively sex-neutral man is relatively (if variably) acceptable in non-­episodic statements. As shown by the contrasts between (3) and (4), however, when a predicated property rules out verification by (adult) male exemplars, no inclusive interpretation for bare singular man subjects is available (whence the double-take effect notated by the ! diacritic). (Note also the mutually reinforcing effect of purportedly sex-neutral man and he.) 3. Man is the only animal that commits suicide. Man is the only mammal capable of rape. Man is the only mammal embarrassed by his nakedness. 4. !Man is virtually the only animal that menstruates. !Man, being a mammal, breast-feeds his young. (Miller & Swift, 1977: 29) !Man is the only mammal embarrassed by his pregnancy. In fact, though, the prototype status of adult male is not essentially a fact about the word man as such but extends to the doctor, the reader, the informant, and other nominals whose denotation is not “female by definition” (Miller and Swift 1991: 41) or—as with teachers, secretaries, prostitutes, or nurses—female by default. Thus consider: 5. a. !Man’s vital interests are life, food, access to females, etc.  (Erich Fromm, “The Erich Fromm Theory of Aggression”, New York Times Magazine, February 27, 1972) b. !We must somehow become witnesses to the everyday speech which the informant will use as soon as the door is closed behind us: the style in which he argues with his wife, scolds his children, or passes the time of day with his friends.  (William Labov (1972), Sociolinguistic Patterns, p. 85) The purportedly sex-neutral avatar, man or informant, is revealed to harbor a Y chromosome, a fellow who seeks females or has a wife with whom to argue.1

 The same applies to plurals, as in the case of people:

1

Sen. Chuck Grassley defended his party’s tax plan this weekend by saying that plans to reduce or eliminate the estate tax mean that people will use their money more wisely. “I think not having the estate tax recognizes the people that are investing”, Grassley (R-Iowa) told the Des Moines Register, “as opposed to those that are just spending every darn penny they have, whether it’s on booze or women or movies.” https://www.politico.com/story/2017/12/03/grassley-tax-booze-women-movies-277764

Ceteris Paribusiness: On the Power of Salient Exceptions

9

In certain discourse contexts, however, access to the putatively gender-neutral use of man or he tends to be restricted or blocked even for oblivious speakers and writers. The contrast in (6), from Black & Coward (1981: 109), is designed to illustrate this effect, although given the societal changes over the last several decades that have affected the demographics of the helping professions, perhaps a sounder minimal pair is that in (7): 6. a. A doctor’s first duty is to {his/?her} patients. b. A nurse’s first duty is to {her/?his} patients. 7. a. A football coach’s first duty is to {his/!her} players. b. A wet nurse’s first duty is to {her/!his} clients. The constraint in question is more striking in certain linguistic frames and contexts. Usage entries exploring putatively sex-neutral he vs. singular they often foreground such cases, such as this one from MWDEU (1994: 902): [T]he insistence on the masculine singular has its limitations. Sometimes its results are downright silly: Everyone should be able to decide for himself whether or not to have an abortion. — Albert Blumenthal New York state assembly member

The MWDEU entry goes on to point out that “the masculine pronoun is awkward at best used in reference to antecedents of both sexes”, as in these two cases with conjoined antecedents cited in the same entry, both taken from a 1983 issue of Reader’s Digest: 8. a. She and Louis had a game—who could find the ugliest photograph of himself.  —Joseph P. Lash, Eleanor and Franklin b. …the ideal that every boy and girl should be so equipped that he shall not be handicapped in his struggle for social progress.  —C. C. Fries (1940), American English Grammar Equally dramatic are contexts in which the pronoun is linked to disjoined antecedents (McConnell-Ginet, 1979, 2011: 197, 1988: 93). 9. a. Any {pupil/!boy or girl} who thinks that he knows the answer… b. !Either my mother or my father will come next week, and he will stay until my birthday. As McConnell-Ginet observes (2011: 197), “A disjunction with a specifically female disjunct simply cannot take he as an anaphoric replacement”. Similar cases are cited by Huddleston & Pullum (2002: 492) and Pullum (2004) with reflexives and possessive pronouns: 10. a. {One of the witnesses/!Either the husband or the wife} has perjured himself. b. !Was it your father or your mother who broke his leg on a ski trip? We can safely assume that, rather than including profligate lesbians among the “people” in question, Sen. Grassley was merely exhibiting his default blinders.

10

L. R. Horn

And it is not just subjects that can yields “downright silly” results, as seen in this minimal pair from Horn and Kleinedler (2000): 11. a. Every child should be taught how to wash his genitalia. b. !Every child should be taught how to wash his penis or vagina. Nor must an actual conjunction or disjunction be present for this effect to be triggered. McConnell-Ginet (2011: 197) adds the examples in (12), where he is decidedly less likely than he or she or they: 12. a. If either spouse in a marriage wants a divorce, {he or she/!he} should consult a lawyer. b. If either of my parents comes, {he or she/!he} will bring a friend. The status of spouse and parent as hyperordinate category labels (cf. Rosch, 1978) renders the basic level hyponyms (husband/wife; mother/father) salient, although some speakers judge these examples as less anomalous than the cases in (8)–(11) with structurally overt female antecedents. What these cases have in common is that they conform to what we will call McConnell-Ginet’s generalization: 13. MGG: McConnell-Ginet’s generalization In contexts where femaleness has been made explicit or is especially salient, it is difficult to use he even where there is no reference to a specific individual.  (McConnell-­ Ginet, 1988: 93)

As we saw in connection with the examples in (4), a similar constraint applies to the use of man. And as we are about to see, the spirit of MGG can be extended to a wide range of diverse phenomena.

Un-accommodation: The Szabó-von Fintel Constraint Our extension of MGG begins with the case of accommodation. While the basic idea underlying presupposition accommodation can be traced back to Strawson, Grice, and Stalnaker, it is first formalized under that name by David Lewis. For Stalnaker (1974), pragmatic presuppositions are “propositions whose truth [the speaker] takes for granted, or seems to take for granted, in making his statement” (Stalnaker, 1974: 198); presupposed material can be communicated as new information by a speaker who “tells his auditor something …by pretending that his auditor already knows it” (Stalnaker, 1974: 202). Whether or not we accept the pretense explanation, it seems plausible that presuppositions are linked to what is potentially non-controversial, rather than to what is mutually known: I am asked by someone who I have just met, “Are you going to lunch?” I reply, “No, I've got to pick up my sister.” Here I seem to presuppose that I have a sister even though I do not assume that the speaker knows this. Yet the statement is clearly acceptable. (Stalnaker, 1974: 202, citing an observation of Jerry Sadock (p.c.))

Ceteris Paribusiness: On the Power of Salient Exceptions

11

On a dynamic view of context, a speaker can act as if a proposition is part of the common ground by getting her addressee to shift his view of the context to accommodate it. 14. Rule of accommodation for presuppositions If at time t something is said that requires presupposition P to be acceptable, and if P is not presupposed just before t, then—ceteris paribus and within certain limits—presupposition P comes into existence at t.  (D. Lewis, 1979: 340, emphasis added)

Ceteris paribus, the speaker of (15) 15. I can’t make it for lunch—I have to pick up my sister at the airport. counts on the addressee to accommodate the presupposition that she has a sister (assuming he doesn’t already know she does) and the addressee will be willing to oblige. But what if cetera aren’t paria? As Kai von Fintel points out, citing Zoltán Szabó (p.c.), a sister whose existence is explicitly put into question by the addressee is not so easy to accommodate. Thus, Szabó observes, if A’s objects to B’s presupposition in (15), B cannot blithely count on A to accommodate: 16. A: Don’t lie to me, I know for a fact that you don’t have a sister. B1: #(In fact) I have to pick her up at the airport. B2: But I do. In fact, I have to pick up her up at the airport. As von Fintel (2008: 163) summarizes the point, “In a context like A’s outburst in (16) where it is clear that it is controversial whether B has a sister, presupposing it is out of bounds despite the existence of presupposition accommodation”. Von Fintel further illustrates the resistance to accommodating controversial information by the example in (17), uttered as an aside by a daughter who knows that her father was hitherto unaware of her engagement. 17. #Oh Dad, I forgot to tell you that my fiancé and I are moving to Seattle next week. In the terms of MGG, we might say that in contexts where the controversiality of a given assumption has been made explicit or is especially salient, it is difficult (“out of bounds”) to expect the hearer to accommodate that assumption even though ceteris paribus accommodation proceeds without a hitch. We now broaden our inquiry to consider a range of other cases that have been discussed in a variety of different approaches and that will enable us to sharpen our description of the relevant principle.

12

L. R. Horn

Conditionals: Sobel Sequences and Other Irreversibilia Consider David Lewis’s remarks on the instability of counterfactual kangaroos: ‘If kangaroos had no tails, they would topple over’ seems to me to mean something like this: in any possible state of affairs in which kangaroos have no tails, and which resembles our actual state of affairs as much as kangaroos having no tails permits it to, the kangaroos topple over. (D. Lewis, 1973: 1) A counterfactual ϕ ψ is true at a world i if and only if ψ holds at certain ϕ-worlds; but certainly not all ϕ-worlds matter. ‘If kangaroos had no tails, they would topple over’ is true (or false, as the case may be) at our world, quite without regard to those possible worlds where kangaroos walk around on crutches, and stay upright that way…What is meant by the counterfactual is that, things being pretty much as they are—the scarcity of crutches for kangaroos being pretty much as it actually is, the kangaroos’ inability to use crutches being pretty much as it actually is, and so on—if kangaroos had no tails they would topple over. (D. Lewis, 1973: 8-9, emphasis added)

Thus, the truth of (18a) consistent with the falsity of (18b). 18. a. If kangaroos had no tails they would topple over. b. If kangaroos had no tails and they walked around on crutches they would topple over. The move from (18a) to (18b) is an instance of Strengthening the Antecedent, a valid inference law for the material conditional: 19. p ⊃ q p & r ⊃ q (for any arbitrary proposition r) But, as recognized inter alia by Stalnaker (1968), natural language conditionals— whether counterfactual/subjunctive2 as in (18) or indicative—are not monotonic in this way. Lewis (1973: 10, referencing Sobel 1970) provides a dramatic illustration of this through “Sobel sequences” like (20): 20. a. If Otto had come, it would have been a lively party. b. …but if Otto and Anna had come, it would have been a dreary [= nonlively] party. c. …but if Waldo had come as well, it would have been a lively party.

 It has long been recognized that “counterfactual” conditionals need not actually be counterfactual, given the evidence from contexts of constructed criminology: 2

(i) If Jones had taken arsenic, he would have shown just exactly those symptoms which he does in fact show.  (Anderson, 1951: 37) (ii) If Mary were allergic to penicillin, she would have exactly the symptoms she is showing.  (Karttunen & Peters, 1979: 6) We cannot address this issue here but will retain the standard terminology for exposition.

Ceteris Paribusiness: On the Power of Salient Exceptions

13

…and so on ad infinitum, even keeping the context constant. (See Appendix A for analogues from popular culture.) More generally, Sobel sequences exhibit the schema in (20′), where > represents the Stalnaker/Lewis conditional: 20′. ϕ1 > ψ ϕ1 & ϕ2 > ¬ψ ϕ1 & ϕ2 & ϕ3 > ψ On this basis, Lewis argues that no theory based on the strict conditional can model natural language counterfactual conditionals; what is needed instead is a variably strict conditional. (For a different conclusion, in three flavors, see von Fintel, 2001; Gillies, 2007; Willer, 2017.) We focus here on the failure of Strengthening the Antecedent for natural language if-then statements—not just subjunctive but also indicative conditionals. (As noted by Lycan, 2001; Willer, 2017, among others, apparent Sobel paradoxes arise for indicative conditionals as well as for subjunctives.) A classic example, revisited by Stalnaker (1968) and D. Lewis (1973), is the contrast between (21a) and (21b). 21. a. If I strike this match, it will light. (T) b. If I strike this match and {it’s wet/I have first soaked it in water}, it will light. (F) c. If I strike this match and it’s (either) wet or not wet, it will light. (F) Even more striking is the fact that (21c), from Warmbrod (1981: 269), in which the antecedent is “strengthened” by a tautology, is as unacceptable as (21b) (see also Herburger & Mauck, 2011; Nute, 1980). Similar paradigms are available for subjunctive conditionals: 22. a. If you had dropped that pencil, it would have fallen. (T) b. If you had dropped that pencil and had first attached a helium balloon to it, it would have fallen. (F) c. If you had dropped that pencil and had either first attached a helium balloon to it or not done so, it would have fallen. (F) Thus, alongside the invalid schema in (23a), we have the more puzzlingly invalid schema in (23b), where the antecedent appears to have been strengthened vacuously. 23. a.

p > q          b. p > q p & r > q         p & (r v ¬r) > q

Furthermore, when confronted with either (22b) or (22c), we have a tendency to revisit our initially confident assessment of the truth of (22a), as Nute observes:3 [A]fter rejecting the conditional ‘If we struck this match and it were wet, then it would light’, there is an inclination to change our evaluation of the original conditional ‘If we struck this match it would light’.  (Nute, 1980: 161)

3  While both Nute and Warmbrod restrict their discussion to subjunctive conditionals, as illustrated by the syntax of this passage, examples like (21) show that no such restriction is necessary.

14

L. R. Horn

Along the same lines, consider the Sobel sequence in (24), an indicative variant of Lewis’s counterfactual schema in (20). We can imagine a context in which Jay, typically an effervescent type who dances with lampshade on head, is prone in Larry’s presence to retreat to the corner of the room to engage in metaphysical debates on the nature of negation, which tends to put a damper on the revelry. But what if the Sobel sequence is reversed as (25)? 24. a. If Jay shows up the party will be festive. b. But if Jay and Larry show up the party will be boring. 25. a. If Jay and Larry show up the party will be boring. b. But if Jay shows up the party will be festive. There may or may not be, in Nute’s words, “an inclination to change our evaluation” of (24a) once we are confronted with the observation in (24b), depending on how likely the state of affairs expressed therein is to arise. But there is a palpable tendency to reject (25b) once (25a) has been asserted. It is evidently easier to move from a ceteris paribus generalization (e.g. about what happens if Jay shows up) to the admission of a special exception to that generalization (e.g. about what happens if Jay and Larry show up) than to state the special case first and then blithely maintain the generalization it would invalidate. The asymmetry is clear: reverse Sobel sequences, which explicitly proceed from a narrow domain (with salient exceptions to the general pattern) to a wider domain (which ignores the possibility of such exceptions), present a problem that regular Sobel sequences, moving from the wider to the narrower domain, do not. The anomaly of sequences like that in (25) has prompted a number of approaches,4 of which we shall single out one in particular: 26. Epistemic irresponsibility (Moss, 2012) (EI) It is epistemically irresponsible to utter sentence S in context C if there is some proposition ϕ and possibility μ such that when the speaker utters S: (i) S expresses ϕ in C (ii) ϕ is incompatible with μ (iii) μ is a salient possibility (iv) the speaker of S cannot rule out μ. [I]f a speaker cannot rule out a possibility made salient by some utterance, then it is irresponsible of her to assert a proposition incompatible with this possibility. (Moss, 2012: 568)

4  The problem posed by reverse Sobel sequences was first noted in print by von Fintel (2001), citing an unpublished observation of Irene Heim, and further discussed by Gillies (2007), Moss (2012), Klecha (2014), Nichols (2017), Lewis (2018), and Ippolito (2020). The role of salience (or “envisaging”) in the asymmetry of Sobel sequences is foreshadowed in the framework of Lycan (2001): “because (25a) already forces envisaging of Larry, a Larry event will linger into the reference class presupposed by (25b)”, thereby rendering the latter unassertible (W. Lycan, p.c., 17 May 2019).

Ceteris Paribusiness: On the Power of Salient Exceptions

15

Moss’s take on epistemically irresponsible utterances is in essence a generalization of von Fintel’s “out of bounds” call for failed accommodation as in (16) above. But there is an earlier precursor worth revisiting: If you say “Italy is boot-shaped” and get away with it, low standards are required and the standards fall if need be; thereafter “France is hexagonal” is true enough. But if you deny that Italy is boot-shaped, pointing out the differences, what you have said requires high standards under which “France is hexagonal” is far from true enough. I take it that the rule of accommodation can go both ways. But for some reason raising of standards goes more smoothly than lowering. If the standards have been high, and something is said that is true enough only under lowered standards, and nobody objects, then indeed the standards are shifted down. But what is said, although true enough under the lowered standards, may still seem imperfectly acceptable. Raising of standards, on the other hand, manages to seem commendable even when we know that it interferes with our conversational purpose. (D. Lewis, 1979: 172, emphasis added)

While the asymmetry Lewis sketches here is motivated by contrasts between acceptable and unacceptable vagueness, it applies directly to the cases reviewed above of acceptable and unacceptable presupposition accommodation and of ordinary and reverse Sobel sequences. In moving from (24a) to (24b), for example, we effectively raise the standards for parties at which Jay is present; as we have seen, conceding the truth of (24b) doesn’t prevent us from maintaining (24a) as a general truth. But in moving from (25a) to (25b), we lower the standards by disregarding the Jay-and-­Larry parties, and under such circumstances (25b) is indeed “imperfectly acceptable” at best. If this is the general description of the pattern exhibited in ordinary and reverse Sobel sequences, we expect to find it attested in other circumstances in which a generalization can be felicitously followed by an exception to it but not vice versa. One natural place to look is in a kindred domain to that of conditionals, namely generics, but to do so we must return to the zoo, this time leaving the poor tailless kangaroos toppling away behind us and moving on to the enclosure of maneless lions and albino ravens.

 lurals, Bare and Definite: (Non-)Maximality and (Non-) P Homogeneity Since Carlson (1977), the semantics of generic bare plurals like (27a) has been a central puzzle for semanticists; the key issues are discussed by Carlson & Pelletier (eds., Carlson & Pelletier, 1995), Liebesman (2011), Leslie and Lerner (2016), Nickel (2017), and references therein. (We set aside episodic bare plurals with indefinite readings like Lions are roaring.) A frequent observation is the irreducibility of (27a) to statements with explicit quantification like those in (27b). Further, as Nickel and others have noted, the suggestion that a covert operator Gen might be employed in the representation of (27a) runs up against the fact that (27c) can be regarded as true even though no lion both has a mane (females don’t) and gives birth to live young (non-females don’t).

16

L. R. Horn

27. a. Lions have manes. (T) b. {All lions/Most lions/Almost all lions} have manes. (F) c. Lions have manes and give birth to live young. Our focus will be on the relation of generic generalizations like (27a), repeated here as (28a), to statements like those in (28b) and (28c), parallel to the corresponding conditionals in the paradigms of (21) and (22). Here again, admitting even a vacuous proviso can turn a felicitous generalization unhappy or, in Moss’s terms, incapable of being responsibly asserted. 28. a. Lions have manes. (T) b. Lions that are females have manes. (F) c. Lions, whether they’re male or female, have manes. (F) Unlike the case of pre-soaked matches and helium-ascending pens, however, non-­ mane-­bearing lions do not actually represent an exception to a ceteris paribus generalization.5 Indeed, given that cubs of either sex along with female lions of any age lack manes, the majority of lions are maneless, although this fact evidently does not interfere with the intuitive truth of (28a). Further, as Leslie and others have pointed out, some generic bare plurals (e.g. Ticks carry Lyme disease) are accepted as true when few members of the relevant set have the property in question, while others (e.g. Books are paperbacks, Presidents are Protestant) are judged false even when the property holds for a strong majority of individuals in the domain. The locus classicus for a generic bare plural that does express a near-universal truth is (29a); the salient appearance of exceptions, even within a tautologous proviso clause as in (29c), makes it impossible to express the generalization responsibly. 29. a. Ravens are black. (T) b. Ravens that are albino aren’t black. (F) c. Ravens, whether or not they’re albino, are black. (F) The asymmetry diagnosed earlier, as attested in the acceptability of ordinary as opposed to reverse Sobel sequences, reappears with generics. Thus compare the felicitous (30) with the infelicitous (31): 30. a. Birds fly. But some birds are penguins and they don’t fly. b. Birds fly. Mumble is a penguin, though, so Mumble doesn’t fly. 31. a. Some birds are penguins and they don’t fly. #But birds fly. b. Mumble is a penguin, so Mumble doesn’t fly. #But birds fly. The ill-formed bare plural continuations in (31) can of course be rescued by modifiers: But birds usually fly, But most birds fly, etc. But without modification,

5  We are bypassing some important issues here. On the difference between (i) and (ii), see Nickel (2010).

(i) Ceteris paribus, all ravens are black. (ii) All ravens are black, unless they are albino.

Ceteris Paribusiness: On the Power of Salient Exceptions

17

there is a clear asymmetry between following up a generalization with a particular case that counterexemplifies it and following up a particular case with the generalization it counterexemplifies.6 But while sequences like those in (31) may be unsalvageable, it is not obvious that the truth of (29a) is irreparably affected by an assertion of (29b), even in the absence of explicit modification. Similarly, consider (32), from Moss (2012: 575): 32. a. If kangaroos didn’t have tails, they’d topple over. (T) b. If kangaroos didn’t have tails but walked with crutches, they’d topple over. (F) c. OK, but seriously, if kangaroos didn’t have tails, they’d topple over. (T) Similar “retraction sequences” are insightfully discussed by Nichols (2017) and K. Lewis (2018), providing alternate wording for objecting to marginal exceptions: “Come on, give me a break”, “but really now”, “picky, picky”, and so on. Evidently, reverse Sobel sequences are felicitous when the speaker can reasonably rule out the possibility that has been raised to salience, as envisaged in step (iv) of Moss’s epistemic irresponsibility recipe in (26). The key distinction is that between salient exceptions to true ceteris paribus generalizations as in (32) or (33) and rejections of claims based on properties of prototypical set members as in (34). 33. A: Ravens are black. B: But of course albino ravens aren’t black. A: Yeah, picky, picky, but come on, ravens are black. 34. A: Lions have manes. B: But of course female lions and male lion cubs don’t have manes. A: #Yeah, picky, picky, but come on, lions have manes. You can’t easily unsee what you’ve seen or unhear what you’ve heard—except when you can. While generic bare plurals like those in (33A) and (34A) are—at least at first glance—taken with (quasi-)universal force, this does not prevent them from allowing exceptions or outliers to which the predicated property fails to apply, as we have seen. In this respect, bare plurals are importantly similar to definite plurals, whatever the other differences may separate the two constructions. Another similarity between bare and definite plurals that has received considerable attention in recent years is the homogeneity effect, formulated by Janet Fodor (1970: 158–168) as the “all-or-none” presupposition. That is, while bare or definite plurals have quasi-­ universal force, the predicate in such cases—as opposed to those with explicit quantification—is taken to be true of all the relevant individuals or of none of them: 35. a. #I didn’t see the boys, but I did see some of them. b. I didn’t see all the boys, but I did see some of them. 6  Outliers need not be overtly mentioned to be rendered sufficiently salient to affect subsequent generalizations. “Birds fly” is more likely to be judged false (or at least not true) when subjects are shown pictures of penguins and ostriches along with songbirds than when they are offered a photo array of cardinals, robins, and eagles.

18

L. R. Horn

Similarly, given the question in (36a), it “would be odd to reply” with (36c), although the same sentence is entirely natural as a response to the question in (36b) 36. a. Are the boys we met orphans? b. Are all the boys we met orphans? c. No, some of them are. In short, “all can be used to contrast with some, but the cannot…[A] simple definite noun phrase in the plural not only does not contrast with some, but does not even admit the possibility that the sentence might be true of some but not all things of the kind described” (Fodor, 1970: 160, 162). And the same all-or-none effect holds for bare plurals as well. Thus, the generic bare plurals in (37a), unlike the quantified subjects in (37b), “leave no room for disagreements about different women.”7 37. a. Women {enjoy/do not enjoy} washing dishes. b. All women {enjoy/do not enjoy} washing dishes. [where all…not = ‘not all’] Fodor’s all-or-none for bare and definite plurals resurfaces as the “generic excluded middle” of von Fintel (1997: 31): “When a kind is denied to have a generic property Pk, then any of its individuals cannot have the corresponding individual-level property Pi.” As von Fintel notes, his principle is a direct descendant of the “homogeneity” or “uniformity” presupposition of Löbner (1985, 2000), as illustrated by the interpretation of negative responses to questions like Do mammals lay eggs or Are the children asleep? Löbner’s homogeneity principle—“If the predicate P is false for the NP, its negation not-P is true for the NP”—is later reformulated as the “presupposition of indivisibility”: “Whenever a predicate is applied to one of its arguments, it is true or false of the argument as a whole” (Löbner, 2000: 239). Either mammals lay eggs or mammals don’t lay eggs. These different formulations are variations on a theme. In each case, the members of a set A either uniformly exhibit a property (e.g. egg-laying) or uniformly exhibit the opposed property (e.g. non-egg-laying); the possibility that there might be some a∈A in one camp and some b∈A in the opposite camp is excluded from consideration. The Law of Excluded Middle, in the form of the all-or-none, homogeneity, or indivisibility, applies in the absence of explicit quantification to both bare and definite plurals, and to mass nouns as well. Thus, it’s hard for me to deny your claim in (38a) by uttering (38b) without seeming to commit myself to a characterization of (all) meat, precisely as with the bare plural generics in (37). The

7  The monolithic behavior of sets denoted with bare or definite plurals results in the fact that these nominals scope over negation, not just in subject position as in (37) but even in object position. While I saw all the boys and I didn’t see all the boys are contradictories, I saw the boys and I didn’t see the boys are contraries (cf. Horn, 2015). Similarly, Leslie (2008, 39) points out that given our truth value judgments on cases like (i) and (ii), “we do not understand negations…as taking wide scope over the bare plural”.

(i) Fair coins come up heads.  (false) ( ii) Fair coins do not come up heads.  (false)

Ceteris Paribusiness: On the Power of Salient Exceptions

19

“middle” possibility that some meat might be healthy and some not healthy is tacitly excluded. 38. a. Meat is healthy to eat. b. (No,) meat isn’t healthy to eat. Some time before Löbner, von Fintel, and even Fodor, Aristotle formulated his own all-or-none—more technically, both-or-neither—for the case of “dialectical” or conjoined questions like that of A in the exchange in (39). 39. A: Are Coriscus and Callias at home? B. No. According to Aristotle (1994), A’s query “is exactly as though he had asked ‘Are Coriscus and Callias at home or not at home?’, supposing them to be both in or both out” (Sophistical Refutations 175b40–176a17; cf. also De Interpretatione 20b12). On this view, B’s negative response above is ill-formed if exactly one of them is home. In his commentary, Ackrill (1963: 145) notes that B’s negative answer above technically just commits him to an inclusive disjunction of negation propositions (= ‘Coriscus isn’t at home or Callias isn’t at home (or both)’), contra Aristotle’s view that A’s question “normally presupposes that they are both in or both out, and that the answer ‘no’ inevitably accepts this presupposition”. But note the difference between (40) and (40′), assuming that B knows that Finland but not Sweden is in the Eurozone. 40. A: Are both Finland and Sweden in the Eurozone.? B. No. 40′. A: Are Finland and Sweden in the Eurozone? B. #No. While Ackrill’s objection to Aristotle’s homogeneity presupposition applies to the exchange in (40), Aristotle’s misgivings about (39B) as a possible answer if Callias is home alone generalize to the case of (40′), where homogeneity (or excluded middle) is operative.8 Indeed, this neo-Aristotelian analysis of the difference between (40) and (40′) is essentially Fodor’s analysis of the difference between bare and quantified plurals: [T]he question “Do women enjoy washing dishes?” is not naturally answered by a straightforward ‘no’ simply on the grounds that some women do not, though one could give that answer to the question “Do all women enjoy washing dishes?” on those grounds.  (Fodor, 1970: 163)

In characterizing Aristotle’s view of (39) as invoking a presupposition, Ackrill (1963) anticipates the Fodor-Löbner view on the all-or-none or excluded middle for bare and definite plurals (see also Gajewski 2005: Sect. 3.4). But is homogeneity or excluded middle really a presupposition at all, or is it more like an implicature? One 8  Groenendijk and Roelofsen (2010) propose a promising treatment of conjunctive questions within inquisitive semantics, but they don’t distinguish the cases with and without both.

20

L. R. Horn

standard diagnostic is cancelability, and it appears that homogeneity can indeed be overridden in the appropriate context. Two attested cases are given in (41) (emphasis added). 41. Almost all the new media of that day [17th c. France] were working, in essence, for kinglouis.gov. Even later, full-fledged totalitarian societies didn’t burn books. They burned some books, while keeping the printing presses running off such quantities that by the mid-fifties Stalin was said to have more books in print than Agatha Christie.  (Adam Gopnik, New Yorker, Feb. 14 & 21, 2011, p. 125, on information control) Diane (laughs):  No, David. You would hate it. You hate people. David:     I don’t hate people. I hate…some people. (Exchange between Diane Lockhart, managing partner at Lockhart Gardner law firm, and David Lee, would-be co-managing partner, “The Good Wife”, CBS, 4 Mar. 2012)

While it is tempting to dismiss such examples as instances of metalinguistic negation (see von Fintel, 1997: 34–35 for such a move), it is worth noting that whatever the status of the negations in (41) may be, the familiar asymmetry holds:9 42. a. #Full-fledged totalitarian societies burned some books, but they didn’t burn books. b. #I hate some people but I don’t hate people. c. #Those are not all the questions, but those are the questions. Krifka (1996) argues on independent grounds that the homogeneity effect is non-­ presuppositional, deriving via pragmatic strengthening. He observes that while the universal reading is preferred in positive plural predications like (43), the existential is preferred in (43′). 43. a. The windows are made of security glass. b. ∀x[x ⊆ the windows → made of security glass (x)] [preferred reading] c. ∃x[x ⊆ the windows ∧ made of security glass (x)] 43'. a. The windows are not made of security glass. b. ¬∃x[x ⊆ the windows ∧ made of security glass (x)] [preferred reading] c. ¬∀x[x ⊆ the windows → made of security glass (x)] Drawing on the “strongest meaning hypothesis” formulated by Dalrymple, Kanazawa, Mchombo, and Peters (1994) for the interpretation of reciprocals, Krifka proposes this generalization: “In predications on sum individuals, the logically stronger interpretation is preferred” (Krifka, 1996: 12). More generally, “If grammar allows for a stronger or weaker interpretation of a structure, choose the one that results in the stronger interpretation of the sentence, if consistent with general  The homogeneity effect in definite plurals can be overridden in a similar way: But how do we understand the bias? How do we check the facts? How do we make some judgments? Those are the questions – but not all the questions. (Consortium for Data Literacy post on “Fake” News, https://tinyurl.com/swu88lf) Here again the usual asymmetry arises:

9

(i) #Those aren’t all the questions but those are the questions.

Ceteris Paribusiness: On the Power of Salient Exceptions

21

background assumptions” (Krifka, 1996: 39). Krifka suggests that this principle might be assimilated to R-based implicatures (Horn, 1984, 1989) that license a speaker to underspecify the force of her utterance while counting on the hearer to recover, in Krifka’s formulation, “the strongest possible interpretation that is consistent with the background knowledge.” Some recent accounts of the homogeneity and non-maximality problems for bare and definite plurals endorse related explanations based on the strongest meaning hypothesis (Breheny, 2005; Winter, 2001).10 These and other recent approaches share a general perspective in which the semantics of sentences containing bare and definite plurals is underspecified and in which pragmatic principles, context, and interlocutors’ goals play major roles in mediating possible meanings (Križ & Spector, 2017: 7; see also—besides sources already mentioned—Schwarzschild, 1994; Lasersohn, 1999; Brisson, 2003; Brogaard, 2007; Malamud, 2012; Križ, 2016). Thus, Lasersohn (1999) contrasts the truth conditions assigned to two sentences with identical predicates and plural definite subjects. In (44a), unlike (44b), the set is treated as monolithic, as we see from the distinct intuitions about the negative (“aren’t asleep”) versions, and exceptions are tolerated. 44. a. The townspeople {are asleep/aren’t asleep}. b. All the townspeople {are asleep/aren’t asleep}. 45. a. The townspeople are asleep. b. The subjects are asleep. But in fact, (45a) allows “only as many townspeople to be awake as can be, for practical purposes, ignored” (Lasersohn, 1999: 523); in the corresponding scenario in (45b), where it is crucial for the experiment that the metabolic state of each subject be determined, exceptions are ruled out. More generally, “Plural sentences allow only for exceptions that are irrelevant for the current purposes of the conversation” (Križ, 2016: 496). Contrast, for example, the two distinct sudden realizations by the speakers in (46a,b), both of whom are away from the house at speech time. (46a), uttered when the painters are expected, gets a universal reading while (46b), uttered in a context when a thunderstorm erupts, gets an existential reading. 46. a. Whew, the windows are open! b. Damn, the windows are open! (See Krifka, 1996; Malamud, 2012; Križ, 2016; Križ & Spector, 2017 for different ways to incorporate speaker/hearer goals and other aspects of contextual variation.) To take one more example, from Krifka (1996: 139): 47. The doors are open.

 For a critique of the strongest meaning hypothesis and an alternative account of the relevant data, see Bar-Asher Siegal (Siegal, 2020).

10

22

L. R. Horn

If the doors are arranged serially and all must be passed through to reach the safe, (47) is understood universally but if they are side by side and the safe can be reached by passing through any of them, it is understood existentially. Following Križ (2016), Champollion, Bumford, and Henderson (2019) offer this description of the role of context in determining truth conditions for definite plural statements.11 The doors are open communicates, roughly, that things are equivalent for current purposes to the way they would be if all the doors were open. Thus, The doors are open is judged definitely true if all the doors are open, and definitely false if all of them are closed, no matter how they arranged. If some but not all of them are open, judgements will depend on whether enough of the doors are open for the purposes of the exchange, and hearers will have to decide whether the state of affairs is more like one in which all the doors are open or more like one in which none of them are.  (Champollion et al., 2019: 6)

For Lasersohn, such judgments don’t directly represent truth conditions. On the basis of the contradictory flavor of (48), 48. Although the townspeople are asleep, some of them are awake. Lasersohn takes (49a,b) to be truth-conditionally identical, 49. a. The townspeople are asleep.  [= (45a)] b. All the townspeople are asleep. differing in that the former is an instance of loose speaking arising from “pragmatic slack”: if a couple of townspeople are awake, (49a) is “literally false”, but “close enough to the truth” that its falsity doesn’t matter. It is not clear, however, how such an account allows for the felicity of plural definites in (50a) or (50'a), while blocking their universal counterparts in (50b) and (50'b). 50. a. The townspeople are asleep, but not {all of them/every one of them}. b. #All the townspeople are asleep, but not {all of them/every one of them}. 50' a. The townspeople are asleep, but a couple of them are awake. b. #All the townspeople are asleep, but not {all of them/every one of them}. Crucially, as we would predict, the prior acknowledgment of outliers precludes any (unmodified) statement of the generalization they counterexemplify: 51. a. Not all the townspeople are asleep, but #(for the most part) the townspeople are asleep. b. A couple of the townspeople are awake, but #(basically) the townspeople are asleep. Thus, the same asymmetry we witnessed in the directionality of exceptions for generics (and conditionals) appears in the case of maximality effects for definite  Champollion et al. extend this account of definite plurals to the complex case of donkey anaphora, which exhibits its own vacillation between universal and existential readings and its own varying degrees of tolerance of exceptions. See also Krifka (1996) and Breheny (2005).

11

Ceteris Paribusiness: On the Power of Salient Exceptions

23

plurals. Two attested examples appear in (52a) and (53a); the constructed counterparts in (52b) and (53b) are ruled out. 52. a. “You locked up the knives”, he chants. “You locked up the knives. But did you lock up all the knives? Did you, did you, did you?”  —Lisa Gardner (2010), Live to Tell, p. 20 b. #You didn’t lock up all the knives, but did you lock up the knives? 53. a. The shutters in their bedroom, slightly twisted by age, succeeded in keeping out the light but not all of it.  —Alexander McCall Smith (2017), A Distant View of Everything, p. 30 b. #The shutters didn’t keep out all the light, but they did keep out the light. The same asymmetry holds for homogeneity effects, as in (54) and (55). Once again, the “raising of standards” in the sense of D. Lewis (1979: 172) is permitted, as in (54a) and (55a). Corresponding discourses requiring the lowering of standards are impossible. 54. a. (?)Uh-oh, the windows are open. But a couple of them are closed. b. #A couple of the windows are closed. But the windows are open. 55. a. I washed the dishes. But I left the messy lasagna dish to soak. b. #I left the messy lasagna dish to soak. But I washed the dishes. Once the existence of the closed windows in (54b) or the lasagna dish in (55b) is conceded, the assertion of the general claim in the succeeding sentence is, in Moss’s sense, epistemically irresponsible.

Universals, Exceptions, and Domain Adjustment We have seen that both generic bare plurals and definite plurals can be taken with more-or-less universal import, allowing maximality interpretations, but where, unlike in the case of true universals, maximality can sometimes be relaxed, whether this is accounted for pragmatically via pragmatic slack, as in Lasersohn (1999), or by relativizing truth-conditional computation to contextual variation. Universal statements admit explicit exceptions in the form of free exceptives as in (56), whether they precede or follow the general claim. (“Connected exceptives”— Everybody but X; No one but Y—are not relevant here; see von Fintel, 1993; Hoeksema, 1996; García, 2009.) 56. a. Except for a few losers, no one gets their navel pierced any more. b. No one gets their navel pierced any more, except for a few losers. c. Other than the tallest student, all the students passed. d. All the students passed, other than the tallest student.

24

L. R. Horn

In such cases, the exception is not asserted. In (57), where the exception is asserted, the universal statement is untenable; again, this holds in either direction (but see below). 57. a. #No one gets their navel pierced any more, but a few losers still do.12 b. #A few losers still get their navels pierced, but no one gets their navel pierced any more. c. The tallest student failed, but all the #(other) students passed. d. #All the students passed, but the tallest student failed. As Sally McConnell-Ginet reminds me, universals based on default or prototype domains (see earlier discussion) do permit subsequent exceptions, as in the celebrated passage from Claude Levi-Strauss in (58a), cited by Livia (2001: 87) among others. But here too, as with the asymmetries in our previous discussion, no universal generalization can be responsibly asserted after an exception is rendered salient: 58. a. All the villagers left the next day in canoes, leaving us alone with the women and children. b. #We were left alone the next day with the women and children, all the villagers having left in canoes. The facts seem plain enough, especially as regards the difference between overt unqualified universals like those in (50b) or (57) on the one hand and generic bare plurals (as in (30a)), definite plurals (as in (50a) or (55a)), or mass definites (as in (53a)) on the other. But in fact, under certain conditions maximality or universality can be relaxed even in the case of overt universals. As before, however, this is possible only when the exception follows rather than precedes the universal quantification. 59. a. Everyone loved the movie. But I hated it. b. I hated the movie. But everyone #(else) loved it. The acceptability of sequences like that in (59a) can be thought of as an extreme example of domain restriction: the second sentence makes it retroactively clear that the everyone in the first sentence must somehow exclude the speaker. But if the speaker begins the discourse by making their own status salient, a subsequent (unqualified) universal is impossible—epistemically irresponsible. The addition of an else here, like that of the for the most part in (51b), restores epistemic responsibility. Lewis’s asymmetry surfaces here once again. The recalculation of the quantificational domain necessitated by the subsequent appearance of a salient exception in (59a) raises the standards—everyone must not have actually been everyone

12

 Sally McConnell-Ginet adds a proviso (p.c., 8 Jan. 2018):

Things are complicated by the normative use of both explicit quantifiers and generics. I could be staring at the stud in my grandkid’s navel, which had not been there a day earlier, and exclaim, “How could you? No one gets their navel pierced anymore”. This seems rather like saying to a sobbing young boy, “Boys don’t cry”.

Ceteris Paribusiness: On the Power of Salient Exceptions

25

relevant—on second thought not everyone loved the movie after all. But (59b) would require lowering the standard: once the speaker’s negative judgment is made salient, no universality excluding it is possible. After the example in (59) was constructed to illustrate our asymmetry, the following passage fortuitously appeared in a New York Times Magazine piece referencing an observation in The Ringer, an online sports and culture blog: “It is a pop-culture law that everything generates a backlash,” The Ringer observed earlier this year, before pointing out that the ensuing cycles of opinion — from initial hype to stubborn critiques to indignant defenses — were now so familiar they could be predicted with almost scientific precision. It went on to do exactly that for “Three Billboards Outside Ebbing, Missouri,” a film about which you could eventually find some people arguing “everyone loves this movie, but I hate it”. (Tung, 2018, emphasis added)

As in (59), the generalization/exception sequence allows the speaker and hearer to coerce a reading of the universal where everyone is retroactively enriched to everyone else or everyone but me. Crucially, however, no such coercion is possible in an exception/generalization sequence, as (59b) shows.

Concluding Remarks It is easier to ask forgiveness than to ask permission. —dictum of uncertain origin13

We began this study by looking at two cases in which ceteris paribus or prototypebased generalizations become difficult or impossible to assert after salient exceptions or outliers to those generalizations have been identified. Salience is key, as reflected in the regularity observed by McConnell-Ginet in the pattern of putatively sex-neutral he: In contexts where femaleness has been made explicit or is especially salient, it is difficult to use he even where there is no reference to a specific individual. (McConnell-­ Ginet, 1988: 93)

We proceeded to identify several other circumstances in which the salience of material within a discourse makes subsequent generalizations inconsistent with that material unsayable or, in Moss’s terms, epistemically irresponsible, unless the speaker, based on the context and on world knowledge, is in a position to dismiss its relevance: [I]f a speaker cannot rule out a possibility made salient by some utterance, then it is irresponsible of her to assert a proposition incompatible with this possibility. (Moss, 2012: 568)

 For other versions of this adage (e.g. “It’s better to apologize than to get permission”), along with the results of an extensive but inconclusive hunt for its origins, see https://quoteinvestigator. com/2018/06/19/forgive/

13

26

L. R. Horn

We were able thereby to distinguish cases in which salient exceptions successfully defeat generalizations from retraction sequences in which apparent exceptions can be dismissed as too unlikely to arise, allowing the generalization to be salvaged. We reviewed in our third section, “Conditionals: Sobel sequences and other irreversibilia,” Lewis’s description of an asymmetry in the dynamics of conversation: “for some reason raising of standards goes more smoothly than lowering” (D. Lewis, 1979: 172). Throughout our survey, ranging from ordinary and reverse Sobel conditional sequences to the quasi-universality of generic bare plurals to the malleable properties of homogeneity and maximality in the case of definite plurals to the case of overt universals, we have focused on reflexes of this asymmetry, and attempted to solve the mystery of the “some reason” Lewis invokes. A ceteris paribus generalization—This match [assuming it’s dry] will light, Ravens [assuming they’re not albino] are black, I washed the [unoverly-sticky] dishes—can be responsibly asserted even if a subsequently acknowledged exception narrows the domain of that generalization retroactively. But if the salient exception comes first, the general rule it counterexemplifies is (in general) unassertable, modulo the possibility of retraction. The domain, once widened to admit outliers, cannot be easily contracted to exclude them. One more particularly vivid example. Consider (60a), due to Lycan (2001: 60), and its reordering in (60b): 60. a. I’ll eat anything on pizza. I’ll even eat pizza with a topping of squid or bull’s testicles, but not bull’s testicles with ground glass. b. I’ll even eat pizza with a topping of squid or bull’s testicles, but not bull’s testicles with ground glass. #But I’ll eat anything on pizza. Once that intriguing ground glass topping has been made salient, it’s hard to overtly ignore it—or, borrowing Lycan’s term, to unenvisage it—as a live (if deadly) option. We have thus attested a left-to-right asymmetry in the discourse sequencing and processing of generalization and exception: exceptions are easier to pardon in the rear-view mirror than to pre-empt in advance. In particular, homogeneity can be overridden retroactively but not pre-emptively—a special case of the edict that it’s easier to obtain forgiveness than permission. Acknowledgments  I am grateful to Barbara Abbott, Elitzur Bar-Asher Siegal, Betty Birner, Donka Farkas, Elena Herburger, Michela Ippolito, Paul Kay, Steven Kleinedler, Sarah-Jane Leslie, Karen Lewis, Floris Roelofsen, Gregory Ward, Arnold Zwicky, and especially Bill Lycan and Sally McConnell-Ginet for helpful suggestions, judgments, and objections, too many of which I’ve doubtless ignored.

Ceteris Paribusiness: On the Power of Salient Exceptions

27

 ppendix: Sobel sequences, small screen scenarios, A and Sunday sermons As discussed on Arnold Zwicky’s blogs (Zwicky, 2019a, b, c, d; the connection to Sobel sequences is noted in Zwicky, 2019d), the alternating pattern of good news/ bad news turns, each refining and overriding the previous one, is a theme of certain comedic routines, including some televised classics. One is the Archie Campbell – Roy Clark routine on the popular hillbilly Hee-Haw broadcasts in the early 1970s (https://tinyurl.com/uuw7jss). Clark, in the barber’s chair, plays straight man (comments in roman) for Campbell (in italics). “Hey I guess you heard about my terrible misfortune.” “No, what happened?” “Yeah, my great uncle died.” “Oh that’s bad!” “No that’s good!” “How come?” “Well, when he died, he left me $50,000 dollars. “Oh that’s good!” “No that’s bad!” “How come?” “When the Internal Revenue got through with it, all I had left was $25,000 dollars.” “Oh that’s bad.” “No that’s good.” “How come?” “Well, I bought me an airplane and learned to fly.” “Well, that’s good.” “No that’s bad. “How come?” “I was flying upside down the other day and I fell out of the thing.” “Well, that’s bad.” “No that’s good. “How come?” “When I looked down under me and there was a great big old haystack.” “Well, that’s good.” “No that’s bad. “How come?” “I got a little closer and I saw a pitchfork aimed right at me.” “Well, that’s bad.” “No that’s good. “How come?” “I missed the pitchfork.” “Well that’s good.” “No that’s bad.” “How come?” “I missed the haystack too.” [etc.]

This proceeds algorithmically for several more volleys, finally morphing into a mother-in-law joke. A more concise exchange between Homer Simpson and the inscrutable proprietor of Springfield’s friendly neighborhood House of Evil (Zwicky, 2019c) first aired on October 29, 1992 on The Simpsons’ “Treehouse of Horror III”

28

L. R. Horn

episode, S4E5: https://tinyurl.com/pwnlhw3. Proceeding from the ridiculous to the sublime, the minister at Washington City Church recruits the Archie Campbell routine as a metaphor for the life, death, and resurrection of the Savior: https://washingtoncitycob.org/2016/02/14/bad-news-good-news/. Rendering an excerpt from the House of Evil exchange into Sobel normal form may help make the connection explicit: (i) If this object carries a terrible curse, (ii) If this object carries a terrible curse but it comes with a free Frogurt, (iii) If this object carries a terrible curse and it comes with a free Frogurt but the Frogurt is also cursed, (iv) If this object carries a terrible curse and it comes with a free Frogurt and the Frogurt is also cursed but you get your choice of toppings, [etc.] It will be noted that while Sobel sequences themselves go on forever, the comic and transcendental versions are all finite, whether they end with the bad news or the good.

References Ackrill, J.  L. (ed. and trans., 1963). Aristotle: Categories and de interpretatione. Oxford, UK: Clarendon. Anderson, A.  R. (1951). A note of subjunctive and counterfactual conditionals. Analysis, 11, 35–38. https://doi.org/10.1093/analys/12.2.35 Aristotle. (1994). Sophistical refutations, trans. London: W.A.  Pickard-Cambridge. Internet Classics, http://classics.mit.edu/Aristotle/sophist_refut.mb.txt Black, M., & Coward, R. (1981). Linguistic, social, and sexual relations: A review of Dale Spender’s Man Made Language. In D.  Cameron (Ed.), Reprinted: The feminist critique of language (2nd ed., pp. 100–118). New York, NY: Routledge. Breheny, R. (2005). Exhaustivity, homogeneity and definiteness. In P. Dekker & M. Franke (Eds.), Proceedings of the Fifteenth Amsterdam Colloquium (pp.  59–65). Amsterdam, Netherlands: University of Amsterdam Press. Brisson, C. (2003). Plurals, all and the nonuniformity of collective predication. Linguistics and Philosophy, 26, 129–184. https://doi.org/10.1023/A:1022771705575 Brogaard, B. (2007). The but not all: A partitive account of plural definite descriptions. Mind & Language, 22, 402–426. http://dx.doi.org/10.1111/j.1468-0017.2007.00314.x Carlson, G. (1977). Reference to kinds in English. PhD dissertation. University of Massachusetts. Carlson, G., & Jeffrey Pelletier, F. (Eds.). (1995). The generic book. Chicago, IL: University of Chicago Press. Champollion, L., Bumford, D., & Henderson, R. (2019). Donkeys under discussion. Semantics and Pragmatics 12(1), 1–50. http://dx.doi.org/10.3765/sp.12.1 Dalrymple, M., Kanazawa, M., Mchombo, S., & Peters, S. (1994). What do reciprocals mean? In  M.  Harvey and L.  Santelmann (Eds.), Proceedings of the 4th Semantics and Linguistic Theory Conference, held May 6-8, 1994, at the University of Rochester (pp. 61–78). Ithaca, NY: Cornell University, Department of Modern Languages and Linguistics. Fodor, J. D. (1970). The linguistic description of opaque contexts. PhD dissertation. MIT. García Álvarez, I. (2009). Generality and exception: The semantics of exceptive constructions. PhD dissertation. Stanford University.

Ceteris Paribusiness: On the Power of Salient Exceptions

29

Gillies, A. (2007). Counterfactual scorekeeping. Linguistics & Philosophy, 30, 329–360. https:// doi.org/10.1007/s10988-007-9018-6 Groenendijk, J., & Roelofsen, F. (2010). Radical inquisitive semantics. Unpublished ms., University of Amsterdam. Available at https://projects.illc.uva.nl/inquisitivesemantics/papers/ unpublished Herburger, E., & Mauck, S. (2011). Strict conditionals, conditionally. Unpublished ms., Georgetown University. Hoeksema, J. (1996). The semantics of exception phrases. In J. van der Does & J. van Eijck (Eds.), Quantifiers, logic and language (pp. 145–177). Stanford, CA: CSLI Publications. Horn, L. (1984). Toward a new taxonomy for pragmatic inference: Q-based and R-based implicature. In D.  Schiffrin (Ed.), Meaning, form, and use in context (GURT ‘84) (pp.  11–42). Washington, DC: Georgetown University Press. Horn, L. (1989). A natural history of negation. Chicago: University of Chicago Press. Reissue edition, Stanford, CA: CSLI, 2001. Horn, L. (2015). On the contrary: Disjunctive syllogism and pragmatic strengthening. In A. Koslow & A.  Buchsbaum (Eds.), The road to universal logic (pp.  151–201). Heidelberg, Germany: Birkhäuser (Springer). Horn, L., & Kleinedler, S. (2000). Parasitic reference meets R-based narrowing: The case of he-­ man language. Paper presented at LSA meeting. Chicago. Huddleston, R., & Pullum, G. (2002). The Cambridge grammar of the English language. Cambridge, UK: Cambridge University Press. Ippolito, M. (2020). Varieties of Sobel sequences. Linguistics and Philosophy 43, 633–671. https:// doi.org/10.1007/s10988-019-09281-8 Karttunen, L., & Peters, S. (1979). Conventional implicature. In C.-K. Oh & D. A. Dinneen (Eds.), Syntax and semantics 11: Presupposition (pp. 1–56). New York, NY: Academic Press. Klecha, P. (2014). Two kinds of Sobel sequences: Precision in conditionals. Posted on lingbuzz. http://ling.auf.net/lingbuzz/002258 Krifka, M. (1996). Pragmatic strengthening in donkey sentences and plural predications. In T.  Galloway and J.  Spence (Eds.), Proceedings of SALT VI (pp.  136–153). New  York, NY: Linguistic Society of America. Križ, M., & Spector, B. (2017). Interpreting plural predication: Homogeneity and non-maximality. Posted on lingbuzz. https://ling.auf.net/lingbuzz/003458 Križ, M. (2016). Homogeneity, non-maximality, and all. Journal of Semantics, 33, 493–539. https://doi.org/10.1093/jos/ffv006 Lasersohn, P. (1999). Pragmatic halos. Language, 75, 522–551. https://doi.org/10.2307/417059 Leslie, S.-J., & Lerner, A. (2016). Generic generalizations. In E. Zalta (Ed.), The Stanford encyclopedia of philosophy. https://plato.stanford.edu/entries/generics/ Leslie, S.-J. (2008). Generics: Cognition and acquisition. Philosophical Review, 117, 1–47. https:// doi.org/10.1215/00318108-2007-023 Lewis, D. (1973). Counterfactuals. Cambridge, MA: Harvard University Press. Lewis, D. (1979). Scorekeeping in a language game. Journal of Philosophical Logic, 8, 339–359. https://doi.org/10.1007/BF00258436 Lewis, K. (2018). Counterfactual discourse in context. Noûs, 52, 479–720. https://doi.org/10.1111/ nous.12194 Liebesman, D. (2011). Simple generics. Noûs, 45, 402–442. https://doi. org/10.1111/j.1468-0068.2010.00774.x Livia, A. (2001). Pronoun envy: Literary uses of linguistic gender. Oxford, UK: Oxford University Press. Löbner, S. (1985). Definites. Journal of Semantics, 4, 279–326. https://doi.org/10.1093/jos/4.4.279 Löbner, S. (2000). Polarity and natural language: Predication, quantification, and negation in particularizing and characterizing sentences. Linguistics and Philosophy, 23, 213–308. https://doi. org/10.1023/A:1005571202592. Lycan, W. (2001). Real conditionals. Oxford, UK: Oxford Universtiy Press.

30

L. R. Horn

Malamud, S. (2012). The meaning of plural definites: A decision-theoretic approach. Semantics & Pragmatics, 5, 1–58. http://dx.doi.org/10.3765/sp.5.3 Martyna, W. (1980). Beyond the he/man approach: The case for non-sexist language. Signs, 5, 482–493. https://doi.org/10.1086/493733 McConnell-Ginet, S. (1979). Prototypes, pronouns and persons. Reprinted in McConnell-Ginet (2011), 185–206. McConnell-Ginet, S. (1988). Language and gender. In F.  Newmeyer (Ed.), Language: The Cambridge survey (Vol. IV, pp. 75–99). Cambridge, UK: Cambridge University Press. McConnell-Ginet, S. (2011). Gender, sexuality, and meaning. Oxford, UK: Oxford University Press. Miller, C., & Swift, K. (1977). Words and women (Updated edition) (p. 1991). New York, NY: Harper Collins. Moss, S. (2012). On the pragmatics of counterfactuals. Noûs, 46, 561–586. https://doi. org/10.1111/j.1468-0068.2010.00798.x  Nichols, C. (2017). Strict conditional accounts of counterfactuals. Linguistics and Philosophy, 40, 621–645. https://doi.org/10.1007/s10988-017-9209-8 Nickel, B. (2010). Ceteris paribus laws: Generics and natural kinds. Philosophers’ Imprint, 10, 1–25. Nickel, B. (2017). Generics. In B. Hale, C. Wright, and A. Miller (Eds.), A companion to the philosophy of language (Vol. I, 2nd ed., pp. 437–462). Chichester, UK: Wiley. Nunberg, G. (1992). Usage note on man. In American heritage dictionary of the English language (3rd ed., p. 1090). Boston, MA: Houghton Mifflin. Nute, D. (1980). Conversational scorekeeping and conditionals. Journal of Philosophical Logic, 9, 156–166. https://doi.org/10.1007/BF00247746 Pullum, G. (2004). Canada supreme court gets the grammar right. Language Log, 8(18), 04. http:// itre.cis.upenn.edu/~myl/languagelog/archives/001362.html Rosch, E. (1973). Natural categories. Cognitive Psychology, 4, 328–350. https://doi. org/10.1016/0010-0285(73)90017-0 Rosch, E. (1978). Cognitive reference points. In E. Rosch & B. B. Lloyd (Eds.), Cognition and categorization (pp. 27–48). Hillsdale, NJ: Lawrence Erlbaum. Saul, J. (2017). Feminist philosophy of language. In E. Zalta (Ed.), Stanford encyclopedia of philosophy. https://plato.stanford.edu/entries/feminism-language/ Schwarzschild, R. (1994). Plurals, presuppositions and the sources of distributivity. Natural Language Semantics, 2, 201–248. https://doi.org/10.1007/BF01256743 Siegal, B.-A., & Elitzur, A. (2020). The NP-strategy for expressing reciprocity: Typology, history, syntax and semantics. Amsterdam, Netherlands: John Benjamins. Stalnaker, R. (1968). A theory of conditionals. In N.  Rescher (Ed.), Studies in logical theory (pp. 98–112). London, UK: Blackwell. Stalnaker, R. (1974). Pragmatic presuppositions. In M. Munitz & P. Unger (Eds.), Semantics and philosophy (pp. 197–214). New York, NY: New York University Press. Tung, C. (2018). Online or in politics, ‘backlash’ is as predictable as weather. New York Times Magazine, 24 April 2018. https://www.nytimes.com/2018/04/24/magazine/online-or-in-politics-backlash-is-as-predictable-as-weather.html von Fintel, K. (1993). Exceptive constructions. Natural Language Semantics, 1, 123–148. https:// doi.org/10.1007/BF00372560 von Fintel, K. (1997). Bare plurals, bare conditionals, and only. Journal of Semantics, 14, 1–56. https://doi.org/10.1093/jos/14.1.1 von Fintel, K. (2001). Counterfactuals in a dynamic context. In M. Kenstowicz (Ed.), Ken Hale—A life in language (pp. 123–152). Cambridge, MA: MIT Press. von Fintel, K. (2008). What is presupposition accommodation, again? Philosophical Perspectives, 22, 137–170. https://doi.org/10.1111/j.1520-8583.2008.00144.x Warmbrod, K. (1981). Counterfactuals and substitution of equivalent antecedents. Journal of Philosophical Logic, 10, 267–289. https://doi.org/10.1007/BF00248853

Ceteris Paribusiness: On the Power of Salient Exceptions

31

Willer, M. (2017). Lessons from Sobel sequences. Semantics & Pragmatics, 10(4), 1–57. http:// dx.doi.org/10.3765/sp.10.4 Winter, Y. (2001). Plural predication and the strongest meaning hypothesis. Journal of Semantics, 18, 333–365. https://doi.org/10.1093/jos/18.4.333 Zwicky, A. (2019a). Oh that’s good. Blog entry, 22 July 2019. https://arnoldzwicky.org/2019/07/22/ oh-thats-good/ Zwicky, A. (2019b). Exception-triggered alternation. Blog entry, 25 August 2019. https://arnoldzwicky.org/2019/08/25/exception-triggered-alternation/ Zwicky, A. (2019c). Revisiting 31: That’s good/That’s bad. Blog entry, 26 August 2019. https:// arnoldzwicky.org/2019/08/26/revisiting-31-thats-good-thats-bad/ Zwicky, A. (2019d). Revisiting 32: Exception-triggered alternation. Blog entry, 27 August 2019. [L.  Horn guest blog.] https://arnoldzwicky.org/2019/08/27/ revisiting-32-exception-triggered-alternation/

I like you may actually implicate ‘I love you’: A reconsideration of some scalar implicatures Yan Huang

Introduction A scalar implicature is a Q-based conversational implicature that is derived from a set of salient contrastive alternatives usually linearly ordered in semantic or informational strength such as a Horn scale1. It is dependent on the non-use of the semantically or informationally stronger alternatives that could have been used in the same set. The derivation of a scalar implicature is due to Grice’s (1975, 1989) first sub-­maxim of Quantity, Horn’s (1984, 1989) or Levinson’s (1987, 2000) neoGricean pragmatic Q-principle. An example par excellence is that the use of the

Notes Earlier versions of this article were presented as a keynote speech at Pragmasophia II – the 2nd International Conference in Pragmatics and Philosophy, the 2nd China High-Level Forum on the Integration and Developments of (Functional) Linguistic Theories, and the 16th China International Pragmatics Conference. They were also given as an invited research lecture/seminar at the ‘General Linguistics Seminar’ series at the University of Oxford, and the ‘Cultures, Languages and Linguistics Seminar’ series at the University of Auckland. I would like to express my gratitude to the audiences on the five occasions, especially Jay Atlas, David Cram, Larry Horn, and Maria Jodlowiec for their insightful comments. My thanks also go to the two anonymous Springer referees for their general comments on this volume, and to Fabrizio Macagno and Alessandro Capone, the editors of this volume. Part of the material and argument contained especially in section “Chinese, Japanese, and English: Use of a Semantically Weak Scalar Expressions out of Face/ Politeness Considerations” of this essay have been presented in Huang (2020). A more detailed discussion and analysis of some of the issues examined in this article will appear in Huang (in preparation) Conversational implicature, contracted to be published by Oxford University Press. This article is dedicated to Professor Laurence R. Horn, my mentors at Yale. 1  The semantic or informational strength in a Horn scale can also be vertically ordered, as in where spaniel entails dog, and is thus semantically stronger than dog.

Y. Huang () University of Auckland, Auckland, New Zealand e-mail: [email protected] © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 F. Macagno, A. Capone (eds.), Inquiries in Philosophical Pragmatics, Perspectives in Pragmatics, Philosophy & Psychology 28, https://doi.org/10.1007/978-3-030-56696-8_3

33

34

Y. Huang

semantically weak existential quantifier some implicates ‘not the universal quantifier all’, as in (1). (I use +> to signify ‘implicate’.) (1) a. Some biologists think that human beings are born selfish. b. +> (The speaker does not believe that) all biologists think that human beings are born selfish. Or the stronger2 c. +> (The speaker believes that) not all biologists think that human beings are born selfish. But this is not always the case. In this article, I discuss three ‘marked’ cases of scalar implicature in a number of different languages: (i) the use of ‘I like you’ to implicate ‘I love you’ in Chinese, (ii) the use of a general noun such as onola ‘person’ to refer to the speaker’s husband or her boy’s father in Malagasy, and (iii) the use of a semantically weak scalar expression to imply the meaning of its stronger alternatives in a face-threatening speech situation in Chinese, Japanese, and English. Following Huang (2020), I call this type of marked scalar implicature, which has received very little attention in the literature, ‘non-canonical scalar implicature.’ I point out that the employment of non-canonical scalar implicatures is implicated in a classical way, with maximum theoretical parsimony, from Grice’s (1975, 1989) co-­operative principle and its attendant maxims of conversation. Finally, I provide a novel analysis within the same neo-Gricean pragmatic framework, combining both Horn’s (1984, 1989) and Levinson’s (1987, 2000) Q- and R/I-principles. In this account, two aspects of scalar implicature will be distinguished: epistemic and non-­ epistemic. For generating the non-epistemic aspects of scalar implicature ‘from weak to not stronger’, Horn scales and the Q-principle are retained; for engendering the epistemic aspects of scalar implicature, a set such as is taken as forming an Atlas-Levinson rather than a Horn scale and the computation of it is subject to the R/I-principle to strengthen the scalar implicature from ‘weak to stronger’. In what follows. I begin with an overview of standard scalar implicature in section “Standard Scalar Implicatures.” I then proceed to discuss three cases of non-­ canonical scalar implicatures and point out the motivations behind their use in section “Three Cases of Non-canonical Scalar Implicature.” At the end of this section, I point out that this type of non-canonical use of scalar implicature is implicated in a classical way, with maximum theoretical parsimony, from Grice’s (1975, 1989) co-operative principle and its attendant maxims of conversation. Finally, I provide a novel neo-Gricean pragmatic analysis of non-canonical scalar implicatures in section “A Novel Neo-Gricean Pragmatic Analysis.”

 In what follows, I will omit the semantically weaker scalar implicature, as in (1b).

2

I like you may actually implicate ‘I love you’

35

Standard Scalar Implicatures What Is a Scalar Implicature? As has already been mentioned, a scalar implicature is a Q-based conversational implicature that is derived from a set of salient contrastive alternatives normally linearly ordered in semantic or informational strength such as a Horn scale. It is dependent on the non-use of the semantically or informationally stronger alternatives that could have been used in the same set. In other words, the derivation of a scalar implicature rests on what is not but could have been said. Another point that is worth noting is that constructed thus, a scalar implicature, like other types of conversational implicature, represents a relation between a speaker and a proposition rather than a sentence produced by the speaker. It constitutes, therefore, a component of speaker meaning. Speakers scalar-implicate; addressees infer (e.g., Huang, 2014: 32–33). This speaker/implicating– addressee/inferencing relationship is reflected in the two expressions of scalar implicature in our examples.

Scalar Implicature-Triggering Scales Three types of scale can license a scalar implicature. Horn scales Following Huang (2020, in preparation), I define a prototypical Horn scale as follows (see also Horn, 1972, 2010; Gazdar, 1979; Atlas & Levinson, 1981; Matsumoto, 1995; Katzir, 2007; Huang e.g., 1991, 1994, 2000, 2007, 2014, 2015, 2017a, b). (2) A prototypical Horn scale For a set of linguistic alternatives to form a prototypical Horn scale, (a) S(x2) asymmetrically entails S(x1), where x2 > x1, (b) x1, x2, …, xn are ‘about’ the same semantic relation, (c) x1, x2, …, xn have the same monotonicity/polarity/scalarity, i.e., they must be in the same (upward or downward) scalar environment. (d) x1, x2, …, xn are of the same word class, (e) x1, x2, …, xn are equally lexicalized, and (f) x1, x2, …, xn are from the same register. A few examples of Horn scales in English are given in (3). (Throughout this article, the items in a Horn scale, rank order, or Hirschberg scale are ranked from weak to strong.)

36

Y. Huang

(3) a. 3 b. c.

Rank Orders Somewhat related to a Horn scale is another scale, dubbed a ‘rank order’ by Horn (1989). In a rank order, lexical items are intuitively felt to be informationally ranked, but unlike those in a Horn scale, the semantically strong linguistic expressions do not unilaterally entail the semantically weak ones. Rather, they asymmetrically entail the negation of them. (4) a. Military rank

b. Life stages rank

c. Mattress size rank

Hirschberg Scales Finally, there is what Levinson (2000) called a Hirschberg scale. While Horn scales and rank orders are entailing scales given by the lexicon without requiring any specific context, Hirschberg scales are non-entailing ones provided by some general assumptions about the world, context, and other pragmatic factors. Put in a slightly different way, whereas the former two scales are a semantic or lexical scale that can be based on the different structures of the lexicon such as taxonomies, metonymies, and helices, the latter is essentially a nonce, pragmatically-defined scale, that is, a contextually given ad hoc scale. Such a pragmatic scale can be based on any partially ordered sets (POSETs) involving relations like subset/set, subtype/type, part/ whole, specification/generalization, and attribute/entity in a contextually or discourse salient way (Hirschberg, 1985/1991). 3  (3a) forms a basic Horn scale. It can be expanded to form a more elaborated one including the two medium quantifiers many/much and most, as in (i). In a similar way, (3b) can be expanded into (ii). Next, (3) contains three positive Horn scales. Horn scales can also be negative. For example, the negative counterpart of (3a) is listed in (iii) (basic form) and (iv) (elaborated form). For further discussion, see Huang (2019b, 2020, in preparation).

( i) (ii) (iii) (iv)  

I like you may actually implicate ‘I love you’

37

(5) a. b. 4 c. (Context: Martin is cycling from Oxford to Manchester, England.)

Horn and Hirschberg scales are both related and distinct. They are related in the sense that Hirschberg scales can be regarded as a Horn scale in context; they are different in the sense that whereas scalar implicatures derivable form a Horn scale are in general generalized conversational implicatures (GCIs), those acting on a Hirschberg scale are particularized conversational implicatures (PCIs).

Computation The computation of a scalar implicature is due to Grice’s (1975, 1989) first sub-­ maxim of Quantity, Horn’s (1984, 1989) or Levinson’s (1987, 2000) neo-Gricean pragmatic Q-principle. (6) Grice’s (1975, 1989) first submaxim of Quantity Make your contribution as informative as is required (for the current purposes of the exchange). (7) Horn’s (1984, 1989) Q-principle (Addressee/hearer-based) Make your contribution sufficient; Say as much as you can (modulo the R-principle). (8) Levinson’s (1987, 2000) Q-principle (simplified) Speaker: Do not say less than is required (bearing the I-principle in mind). Addressee: What is not said is not the case. Let me now present three examples, one from each of the three scales. (9) A scalar implicature from a Horn scale a. Horn scale b. The tea is warm. c. +> (I/the speaker believe/believes that) the tea is not hot. (In addition, (1) above is also an example containing a scalar implicature.) (10) A scalar implicature from a rank order a. Rank order b. Hello, Rose. This is my girlfriend, Mary Lyons. c. +> (I/the speaker believe/believes) that this isn’t my/his wife.5

 This rank order is also vertically ordered.  Note that in the following, what is meta-linguistically negated is the potentially scalar-implicated upper-bounded reading. 4 5

38

Y. Huang

(11) A scalar implicature from a Hirschberg scale (Context: Martin is cycling from Oxford to Manchester, England.) John: How is Martin doing? Mary: He’s got to Birmingham. a. Hirschberg scale b. Mary: He’s got to Birmingham. c. +> (The speaker believes that) Martin hasn’t got to Manchester Since Horn (1972), one line of the standard Gricean/neo-Gricean analysis of the computation of scalar implicatures has been the one-step/process argument. Roughly, given a Horn scale, a rank order or a Hirschberg scale, by the Gricean and neo-Gricean pragmatic mechanism (6), (7) or (8), if a speaker asserts a lower-ranked or semantically/informationally/communicatively weak scalar expression (i.e., a leftwards expression in the linearly ordered set), he/she scalar-implicates that he/she is not in a position to assert any of the higher-ranked or semantically/informationally/communicatively stronger ones (i.e., the rightwards expressions) or sometimes the strongest one in the same set. Thus in (9), since the speaker used a semantically weak scalar item warm but not the semantically stronger alternative hot which he/ she could have used, he/she implied that he/she believed that ‘hot’ is false, and the addressee inferred the speaker’s scalar implicature to that effect. The same can be said of (10) and (11) (e.g., Horn, 1972; Huang, 1991, 1994, 2000, 2007, 2010a, b, 2014, 2017a, b, 2018a, 2019a, b, 2020; Levinson, 1983, 1987, 2000). Another, more sophisticated, neo-Gricean approach is the two-step/process argument, following Soames (1982). Again take (9) as an example. Roughly, the first step is that since the speaker could have used, but did not actually use hot, he/she implicated that he/she did not believe that ‘hot’ is true (what is called the ‘primary implicature’ by Sauerland, 2004). In the second step, (on the assumption that the speaker is likely to have an opinion or is opinionated about the semantically stronger scalar expression, so-called ‘opinionatedness’), he/she further engendered the stronger scalar implicature that he/she believed that ‘hot’ is false (what is called the ‘secondary implicature’ by Sauerland, 2004), and the addressee inferred them to that effect (see Hirschberg, 1985/1991; Horn, 1989; Sauerland, 2004; Katzir, 2007; see also Gazdar, 1979). (i) Charles: So John, how’s that your gorgeous girlfriend of yours? John: Ah, she’s no longer my girlfriend. Charles: Oh, dear. Still I wouldn’t get too gloomy about it. Rumour has it she never stopped bonking old Toby de Lisle in case you didn’t work it out, heh. John: She is now my wife. (pause) Charles: Excellent, well ah… excellent, congratulations.  (Four Weddings and a Funeral, 1994) a. b. Ah, she’s no longer my girlfriend… She is now my wife. (Compare: Ah, she’s my girlfriend. +> She is not my wife.)

I like you may actually implicate ‘I love you’

39

Three Cases of Non-canonical Scalar Implicature In the last section, I discussed standard scalar implicatures and their computation. In this section, I move to the use of non-canonical scalar implicatures in three cases in a few languages.

Chinese: From wo xihuan ni (I Like You) to wo ai ni (I Love You) One typical use of a non-canonical scalar implicature is that around half a century ago (that is, up to late 1980s/early 1990s, but see what is below), during courtship, a Chinese girl tended to express her love to her boyfriend/fiancé by saying wo xihuan ni (I like you) rather than wo ai ni (I love you). The same was true of a boy to a girl to a less extent. This way of expressing love between man and woman was especially in vogue during Mao’s brutal 10-year ‘Great Proletarian Cultural Revolution’ (1966–1976), when the use of ai in its romantic/sexual sense was extremely rare, if not non-existent. Consider now (12), which is drawn from Fanghua (Youth), a controversial Chinese film released in 2017. The film is a chronicale of the different fates of a group of young solider-dancers in a military dance troupe in the Chinese People’s Liberation Army (PLA) from Mao’s ‘Culture Revolution’ to the Sino-Vietnamese Border War in late 1970s (Youth (2017 film), Wikipedia). https://en.wikipedia.org/ wiki/Youth_(2017_film) The following excerpt is taken against the background of the eve of the border conflict between China and Vietnam in the 1979. In the excerpt, Liu is a male soldier-dancer and Lin is a female soldier-dancer. (12) (Chinese) 刘峰:我是不想离开这儿。 林丁丁:为什么啊 , 这有什么好的? 刘峰:这有你。 林丁丁突然停了下来 , 两个人沉默了一会儿。 刘峰:小林 , 其实这些年我一直在等你。 林丁丁:等我干什么啊? 刘峰:就等我们俩能像现在这样。 林丁丁:现在怎么了? 刘峰:我喜欢你 , 我以前怕影响你进步 , 所以我想等你入了党再跟你提 , 这不是昨天你预备期顺利通过了吗 , 我头一次遇见你 , 我听你唱一条 大河我就喜欢你。 林丁丁沉默不语 , 刘峰站起来靠近她 , 林丁丁闪躲一般站了起来。 林丁丁:说得我都不好意思了 , 我走了。 林丁丁离开 , 但刘峰手里拿着她的毛线 , 毛线拉了很长 , 她又折返回 来 , 刘峰上前几步 , 趁还她毛线握住了她的手。

40

Y. Huang

林丁丁想躲 , 但是被刘峰一把抱在了怀里。 刘峰:我听到那些歌 , 我觉得都是为我唱的 , 都是我想说的。 林丁丁想挣脱开:早知道不给你听了。 刘峰:我真的喜欢你 , 我真的喜欢你。 《芳 华》 Liu: I just don’t want to leave here. Lin: Why? What’s so great about this place? Liu: You.

(Suddenly, Lin stops talking. There is a brief silence between the two.) Liu: Little Lin, I’ve been waiting for you for all these years. Lin: For me to do what? Liu: For us to be like we are now. Lin: What’s now? Liu: I like you. I used to worry about affecting your future, so I decided to wait until you joined the (Communist) Party. Didn’t you pass probation in fine form yesterday? From the first time I saw you, and heard you sing The Big River, I’ve been liking/found of you.

(Lin becomes silent again. Liu stands up and moves closer to her. Lin also stands up but in an elusive way.) Lin: You’re making me embarrassed, I’ve got to go.

(Lin is about to leave, but her knitting wool thread is in Liu’s hand. She walks back. Taking a few steps forward, Liu holds her hands while returning her thread to her. Lin tries to evade him, but she is taken into his arms.) Liu: When I heard those songs, I thought the lyrics were for me, it was everything I wanted to say. Lin: If I knew that, I wouldn’t have sung them for you. Liu: I really like you, I really like you. (Youth)

In this excerpt, when Liu said to Lin ‘I (really) like you’, he did not implicate the standard scalar implicature ‘I don’t love you’, Instead, he engendered the non-­ canonical scalar implicature ‘I (really) love you’. In fact, the uttering of ‘I like you’ to non-canonically implicate ‘I love you’ is still widely used in contemporary China. Following is another example taken from a popular Chinese TV series entitled Meigonghe Da’an (Murders on the Mekong). It tells the story of a long manhunt and the subsequent bringing to justice of the criminals who massacred all the 13 Chinese crew members on the two Chinese cargo ships on a stretch of the Mekong River in the Golden Triangle region on the borders of Myanmar and Tailand on the morning of 5, October 2011 (Mekong River massacre, Wikipedia). https://en.wikipedia.org/wiki/Mekong_River_massacre (13) (Chinese) 高野:于惠, 我挺喜欢你的。你愿不愿意嫁给我? 于惠:愿意。 《湄公河大案》 Gao Ye: Yu Hui, I like you very much. Will you marry me? Yu Hui: I will.   (Murders on the Mekong)

I like you may actually implicate ‘I love you’

41

In this excerpt, Gao Ye, a male officer in an elite Chinses anti-narcotics police unit was proposing to Yu Hui, one of his female colleagues. Once again, he used ‘I like you very much’ to scalar-implicate ‘I love you’ in a non-canonical way. Two main factors seem to play a role here. In the first place, traditionally, a courting couple, especially girls in China were encouraged to express their romantic emotions/feelings in a roundabout or tactful way. Secondly and perhaps more importantly, for (12) but not for (13), during such a brutal political campaign as Mao’s ‘Cultural Revolution’, people were coerced to speak and behave in an ideologically/politically correct way. (According to the then communist ideology, a proletarian revolutionary should not have the petty bourgeois sentiments like that denoted by ai in its romantic sense.) If a semantically stronger scalar expression cannot be used, then there will be a (one-step) decrease in semantic strength, i.e., the next semantically strongest expression in the set is deployed. What examples like (12) and (13) indicate is this: there are at least some circumstances/contexts, under which the use of a semantically weak scalar item in a Horn scale may actually implicate its stronger alternative(s).

Malagasy: Use of a General Noun for Personal Reference There are other similar cases. One such case is that, as reported in Keenan (1976: 72), in talking to her son, a Malagasy mother would use a ‘general animate noun referring to some social category of which the referent is a member’ to refer to her husband or the boy’s father. (14) (Malagasy, Keenan, 1976) a. Mbola mator y ve ny olona? Is the person still asleep? b. +> E.g., is your father still asleep? Note that (14) is in sharp contrast to (15) in English. (15) Grice’s (1975, 1989: 37) a. X is meeting a woman this evening. b.+> The woman in question isn’t X’s wife, mother, sister, or perhaps even close platonic friend. While in Grice’s classic example (15), the use of the semantically general, thus weak noun woman gives rise to a standard scalar implicature, in the Malagasy example (14), once again, a semantically weak scalar expression olona ‘person’ is deployed to implicate, for example, the semantically stronger ‘your father’. Based on examples like (14), Keenan (1976) concluded that the classical Gricean pragmatic theory is culture-specific rather than universal. However, if we examine the Malagasy facts more closely, we find that the use of a general noun to refer to, for example one’s husband or one’s boy’s father is not just in keeping with Grice’s

42

Y. Huang

first sub-maxim of Quantity, Horn’s or Levinson’s Q-principle, but it actually requires the existence of this pragmatic principle for the usage to be interpreted (see also Brown & Levinson, 1978: 288–289, 1987; Levinson, 2000: 423). As Keenan (1976, 75–76) herself was aware, “[i]t would be misleading to conclude that the maxim ‘Be informative’ does not operate at all in a Malagasy community. We would not be justified in proposing the contrary maxim ‘Be uninformative’ as a local axiom”. In fact, Grice’s first sub-maxim of Quantity, Horn’s or Levinson’s Q-principle does generally hold for the Malagasy-speaking culture, as is attested by (16).6 (16) (Huang, 2014: 442) a. Misy tanora tia ny hira malaza. exist young like the song famous Some young people like famous songs. b. +> ‘(The speaker believes that) not all young people like famous songs. As pointed out by Huang (2018b), what Keenan showed, however, was that in the Malagasy society, Grice’s first sub-maxim of Quantity, Horn’s or Levinson’s Q-principle can be overridden by some sociolinguistic principle such as the one avoiding bringing tsiny (guilt) to the speaker or henatra (shame) to the speaker’s family. To quote Keenan (1976: 70): ‘A second and perhaps more significant motivation for revealing less information than would satisfy the addressee is the fear of committing oneself explicitly to some particular claim. Individuals regularly avoid making explicit statements about briefs and activities. They do not want to be responsible for the information communicated. For example, if someone asks, ‘Who broke the cup?’, most speakers would not like to be the one to specify the culprit. Such a statement may have unforeseen unpleasant consequences for him and his family, and he alone would have to shoulder the tsiny (guilt) for uttering such a claim (original emphasis).7

Other factors that overweigh the operation of Grice’s first sub-maxim of Quantity, Horn’s or Levinson’s Q-principle in the Malagasy-speaking community, mentioned by Keenan (1976: 69), include satisfying the principle “would be indiscrete, impolite, unethical, loss of face, etc.” In the case of personal reference, it is because of a particular Malagasy social taboo on avoiding identifying an individual in utterances 6  This was confirmed to me by Larry Horn (personal communication), who checked the fact with Keenan after her paper was published. 7  Note that as reported in Nwoye (1992: 315), in the traditional Igbo society, people would use (i) in preference to (ii) in their attempt to find the culprit of a theft.

(i) Qbu nwa onye zulu okuku? Whose child stole the chicken? (ii) Onye zulu okuku? Who stole the chicken? Clearly, question (i) is less offensive than question (ii), and as a result, it is easier to be answered than question (ii). The respondent would take less responsibility and potentially have less unpleasant consequences.

I like you may actually implicate ‘I love you’

43

that the mother did not deploy a more informative term such as ‘your father’ to refer to her husband or her boy’s father. Instead, she employed a general noun. By this way of referring, she succeeded in identifying him without bringing harm to him or shame to herself. If Grice’s first sub-maxim of Quantity, Horn’s or Levinson’s Q-principle did not work at some deeper level, her son would fail to recognize the intended referent.8 Summing up, once again, the non-canonical use of the semantically weak scalar item in (14) is not without a reason. In this case, it is due to the existence of some sociolinguistic principle such as the one which avoids guilt, especially the particular social taboo on exact identification in Malagasy that the mother employed a semantically weak general noun rather than a semantically stronger or more informative kinship term to refer to her husband or the boy’s father.9

 hinese, Japanese, and English: Use of a Semantically Weak C Scalar Expression out of Face/Politeness Considerations Finally, I turn briefly to non-canonical scalar implicatures that are found in a face-­ threatening context–a context in which a loss of face can potentially be induced for

8  More recently, Senft (2008) claimed that Grice’s maxim of Quality and his first and second submaxims of Manner are not adhered to by the Kilivila-speaking Trobriand Islanders of Papuan New Guinea. This is the case with both their ritualized communication and everyday conversation, especially with the use of the non-diatopical register/variety called biga sopa (the joking or lying speech, the indirect speech, and the speech which is not vouched for). While the details of Senft’s work need to be more carefully studied, his claim seems to be another apparent counterexample. Heated debates about the issue of universality in terms of the distinction between the etic versus emic approach/grid have been going on also relating to, for example, speech act theory, politeness/ impoliteness theory, and conversational structure. 9  In rejecting Keenan’s counterexamples as real ones, I am fully aware of what a recent Editorial in Nature (2015) calls the human ‘cognitive bias’, namely, ‘[t]he human brain’s habit of finding what it wants to find,’ which ‘is a key problem for research.’ As pointed out by the Editorial, “One enemy of robust science is our humanity–our appetite for being right, and our tendency to find patterns in noise, to see supporting evidence for what we already believe is true, and to ignore the facts that do not fit.” See also the other three relevant papers published in the same issue of Nature (vol. 526, no. 7572). Somewhat related is that as pointed out by Kuhn (1962/2012), the research methodology of observation, for example, is “strongly theory-laden”. This is because when a researcher does observations, he/she has already been significantly influenced by his/her previously held theoretical and methodological assumptions. The same can also be said of experimentation. See also Popper (1945).

44

Y. Huang

the addressee, the speaker or both–out of face/politeness concerns.10,11 As discussed in Huang (2020), these face-threatening speech situations include responses to compliments, disapprovals or criticisms, disagreements, refusals, and announcing or telling a piece of bad news. Needless to say, many other face-damaging contexts can be added. In what follows, let me just select three such cases. Compliment Responses (17) (Chinese) A: 你儿子的数学真好。 B: 一般 , 一般 。 A: Your son’s math is excellent. B: It’s OK, it’s OK. a. Horn scale b. Not +> (B believes that) her son’s math isn’t excellent  Face is originally a Chinese concept. There are a number of words in Chinese that can be used to express the notion: lian (face), mianzi (reputation, prestige, face), and lianmian (face, self-respect). According to some Chinese scholars (e.g., Ho, 1976), while lian refers to ‘the confidence of a society in the integrity of the ego’s moral character, the loss of which makes it impossible for him to function properly within the community’, by mianzi is meant an individual’s earned prestige, achieved through accomplishment or success. It seems that this concept of face was introduced to the West by the Chinese anthropologist Shien Chin Hu in the 1940s (see e.g., Huang, 2007:116, 2014: 143). A few of my students at the University of Auckland told me that in the Pacific cultures (such as Cook Islands, Maori, Niue, Samoan, and Tongan) face is in general considered to be one of the most sacred parts of the body, because it is part of the head. ‘Saving face’ has often saved a nation, launched a war against a tribe, and caused thousands of people to die. Technically speaking, face can be distinguished between first-order and second-order face. By first-order face is meant the everyday sense of the notion. In contrast, second-order face refers to a technical sense of the concept. The notion of face mentioned in the above two paragraphs of this note is that of first-order face. By contrast, the concept of face in classical Brown’s (2017) and Brown and Levinson’s (1978, 1987) ‘face-saving’ theory of politeness is that of second-order face (see e.g., Huang, 2014: 143). 11  In a similar fashion, politeness can also be divided into first-order and second-order politeness (Watts, Ide, & Ehlich, 2005). By first-order politeness is meant the common-sense notion of normative politeness, that is, a judgement about whether a particular behaviour is polite or not in keeping with the norms of a socio-cultural or speech community, made by lay members of that community. In contrast, second-order politeness refers to a scientific concept of politeness, that is, an abstract, theoretical construct defined within a theory of politeness. Furthermore, it should be pointed out that there is a dynamic trade-off between the two notions of politeness. For example, second-order politeness, which is informed by first-order politeness, is a concept that is more inclusive than first-order politeness. Consequently, first- and second-order politeness should be studied hand in hand rather than in isolation. Notice that the concept of politeness in classical Brown and Levinson’s ‘face-saving’ theory of politeness is that of second-order politeness (see e.g., Huang, 2014: 143). 10

I like you may actually implicate ‘I love you’

45

Or, stronger c. +> (There is a possibility that B believes that) her son’s math is excellent. ( 18) Japanese (Matsumoto, 1995) (An interviewer A is talking with the winner of a sumo match B.) A: Tsuppari ga subarashikatta desu ne. B: Maa maa desu. A: Your push was splendid, wasn’t it? B: It was so-so. a. Horn scale b. Not +> (B believes that) his push wasn’t splendid. Or, stronger c. +> (There is a possibility that B believes that) his push was splendid. B’s reply in both (17) in Chinese and (18) in Japanese is a compliment response. While there are cross-linguistic/cross-cultural parallel patterns of compliment,12 compliment responses vary from language to language and culture to culture. For instance, whereas a compliment normally generates acceptance/thanking in (the different varieties of) English, it engenders rejection and/or self-denigration in Chinese, Japanese, Korean and to a less extent Turkish, Persian, Polish, and Spanish. In fact, regarding compliment responses, some cross-linguistic/cross-cultural variations can be indicated along a three-point continuum put forward by Holmes (1986), namely, acceptance-deflection/evasion-rejection. Given this cline, Arabic speakers are most likely to accept a compliment, followed by speakers of South African, American, and New Zealand English. Next come Irish English and Thai. (I may add Malay here.) On the other hand, speakers of Chinese, Japanese, Korean and Turkish are most likely to reject a compliment, while speakers of certain non-English European languages like German and Spanish (I may add Romanian here) with the possible exception of French, may be positioned at the mid-point, d­ eflection/evasion, of this three-pronged continuum (Huang, 2014: 154–157; see also Chen, 2010). Returning to (17) in Chinese and (18) in Japanese, B’s compliment response did not give rise to a relevant standard scalar implicature. Instead, it generated a relevant, non-canonical scalar implicature due to face/politeness considerations.

 I am not saying that there are no cross-linguistic/cross-cultural differences in making compliments. Take compliment topics as a case in point. While in communist Poland, one would deliver a compliment on material possession more than any other categories, in communist Romania, this was avoided. The explanation for the former: it is due to scarcity of material goods speakers of Polish faced at the time (Herbert, 1991); that for the latter: it is because that individual wealth would have been frowned upon in what was supposed to be an egalitarian society (Dumitrescu, 2006). Another illustration. Children are a frequent topic of what is complimented on in Chinese, but it does not seem to be a significant topic in any other language/culture (Chen, 2010). The same can also be said of what linguistic expressions one can use to make a compliment. For instance, highly sexualised expressions like piropos can be used in Spanish as a compliment to certain people in certain contexts, the use of them is not acceptable in English. As observed by Dumitrescu (2006: 17): ‘if English has ‘piropos’ they would fall into the category of sexual harassment’.

12

46

Y. Huang

Disapprovals/criticisms (19) (Chinese) 你这样做有点不好。 It’s sort of bad for you to do that. (a) Horn scale 13 (b) 你这样做有点不好。 (c) +> 你这样做很不好。 (d) +> (There is a possibility that I/the speaker believe/believes that) it’s very bad for you/the addressee to do that. In uttering (19) the speaker was performing the speech act of disapproving and/ or even criticising, Speech acts of this kind are inherently face-threatening acts (FTAs), and as such they damage the addressee’s face. By using the semantically weak ‘sort of bad’ in (19) but not its semantically stronger counterpart, the speaker had succeeded in mitigating the face-threatening effect of these speech acts. Announcements of Bad News (20) (English, Bonnefon, Feeney, & Villejoubert, 2009) You are probably going to become deaf soon. (a) Horn scale (b) +> (There is a possibility that I/the speaker believe/believes that) you are certainly going to become deaf soon. In (20), the speaker was announcing a piece of bad news to the addressee. This type of speech act is also inherently face-attacking. That is why it is usually prefaced by a pre-sequence called pre-announcement in Conversation Analysis (CA), as is illustrated in the arrowed conversational turn in (21). ( 21) (Cited in Levinson, 1983: 351) → D: Didju hear the terrible news? R: No. What? D: Y’know your Grandpa Bill’s brother Dan? R: He died. D: Yeah. Note also that in this example, the speaker let the addressee rather than himself/ herself be the person who announced the bad news and himself/herself be the person who confirmed it. Going back to (20), the use of probably, which is semantically stronger than possibly and weaker than certainly did a good job in both conveying the bad news to the addressee and saving his or her face.

 Larry Horn (personal communication) may not take this as a Horn scale. But a similar one appears in Hirschberg (1985/1991).

13

I like you may actually implicate ‘I love you’

47

From the above discussion, we can conclude that the use of a semantically week scalar expression to generate a non-canonical scalar implicature in all the three cases is politically, socially and/or culturally motivated. Remember that Grice’s cooperative principle and its constituent maxims of conversation including his first sub-maxim of Quantity define an unmarked or socially neutral (and indeed asocial) presumptive framework for communication, the essential assumption being ‘no deviation from rational efficiency without a reason.’ Traditional cultural/social values, political/ideological correctness, social taboos, and face/politeness considerations are, however, just some of the principled reasons for such a deviation. Therefore, they are implicated in the classical way, with maximum theoretical parsimony, from Grice’s co-operative principle and its component maxims of conversation.14,15

A Novel Neo-Gricean Pragmatic Analysis Finally, in this section, following Huang (2020), I provide a new analysis of non-­ canonical scalar implicature within the neo-Gricean pragmatic framework. Let me start by assuming that there are two aspects of this type of scalar implicature. One, standard aspect is concerned with the now familiar epistemic status of a speaker. On the other hand, the second aspect is concerned with dimensions other than the speaker’s epistemic status. In the latter case, for some marked, non-epistemic reasons such as traditional cultural/social value, political correctness, social taboo, face/politeness concerns, the speaker would consider using a semantically weak scalar expression, although he/she is in an epistemic position to use a semantically stronger alternative. Consequently, the scalar implicature generated by the speaker is along the following lines in (22) (see also Geurts, 2009). (22) (The speaker believes that) it would be improper (such as impolite, indiscrete, unethical, immoral, and/or politically/ideologically incorrect) for him/her to use a semantically stronger alternative. In other words, in these non-canonical cases of scalar implicature, when the speaker uses a semantically weak scalar item, the scalar implicature produced by him/her is not one with regard to his/her epistemic status, but one indicating that he/  Leech (1983, 2007) argued that a politeness principle and a set of maxims of politeness be added to the Gricean programme and that they should be treated as co-ordinate in nature to Grice’s cooperative principle and its associated maxims of conversation. It goes without saying that maxims of this kind such as ‘Express your love in a roundabout way‘, ‘Be politically/ideologically correct’, and ‘Avoid exact personal identification’ can in principle be applied to the Chinese and Malagasy case as well. But see Huang (e.g., Huang, 2007, 2014, 2018b, 2020) for my arguments against such an expansionist approach, especially Leech’s (1983,  2007) proposal from a philosophical and methodological point of view. 15  Another potential research topic may be the non-canonical use of scalar implicature derived in a diplomatic context. 14

48

Y. Huang

she was not in a position to use a semantically stronger scalar expression for non-­ epistemic reasons (some of which are indicated above). This way, we can maintain Horn scales and the ‘from weak to not stronger’ nature of these non-canonical scalar implicatures. Next, for the epistemic aspect of these scalar implicatures, as soon as the addressee infers that the weak scalar expression is the most informative information the speaker can pick up in a Horn scale for some non-epistemic reasons, he/she will reinterpret it as epistemically implicating its next stronger or sometimes, the strongest alternative in the same set. In other words, for the epistemic aspect of this type of implicatures, ‘from weak to not stronger’ is changed to ‘from weak to stronger’. How can this be accounted for within the same neo-Gricean pragmatic mechanism? First, we treat, for example, as forming an Atlas-Levinson rather than Horn scale. Roughly speaking, a ‘prototypical’ Atlas-Levinson scale is to some extent the reverse of a Horn scale. A few ‘standard’ examples are given in (23)–(25) below (Atlas & Levinson, 1981; Levinson, 1987, 2000; see also Huang, 1991, 1994, 2000, 2007, 2014, 2017a for a representative list of other examples). (23) Conjunction buttressing

p and q +> p and then q (temporal sequence) +> p therefore q (causal connectedness) +> p in order to cause (teleology, intentionality) John pressed the spring and the drawer opened. +> John pressed the spring and then the drawer opened +> John pressed the spring and thereby caused the drawer to open +> John pressed the spring in order to make the drawer open (24) Conditional perfection  (iff = if and only if) If you help me my syntax essay, I’ll help you with your phonology essay. +> If and only if you help me with my syntax essay will I help you with your phonology essay. (25) Lexical narrowing

John doesn’t drink. +> John doesn’t drink alcohol As (23)–(25) indicate, in a stereotypical Atlas-Levinson scale, the semantically stronger expression entails the semantically weaker one, that is, for example, iff p, q entails if p, q in (24). To this list, we can add , as in (12) and (13) above, repeated here as (26) for ease of exposition. (26) a. b. I like you. c. +> I love you.

I like you may actually implicate ‘I love you’

49

Consequently, the computation of this new I-implicature is subject to Grice’s second sub-maxim of Quantity, Horn’s R- or Levinson’s I-principle. (27) Grice’s second submaxim of Quantity Do not make your contribution more informative than is required. (28) Horn’s R-principle (Speaker-based) Make your contribution necessary; Say no more than you must (modulo the Q-principle). (29) Levinson’s I-principle (simplified and modified) Speaker: Do not say more than is required (bearing the Q-principle in mind). Addressee: What is said is pragmatically strengthened. Given (27), (28) or (29), when a speaker uses a semantically weak linguistic expression, he/she epistemically implicates a semantically stronger proposition. By way of summary, in (26) form two scales: one Horn scale and one Atlas-Levinson scale. When the speaker uses the semantically weak like in the Horn scale, by Grice’s first sub-maxim of Quantity, Horn’s or Levinson’s Q-principle, he/she generates a scalar implicature in relation to the non-epistemic aspect of the implicature, namely (30); when he/she employs like in the Atlas-Levinson scale, given Grice’s second sub-maxim of Quantity, Horn’s R- or Levinson’s I-principle, he/she gives rise to an I-implicature with regard to his/her epistemic status, namely (31). (30) a. Horn scale b. like +> not love, … c. +> (The speaker believes that) it would be improper for him/her to use a semantically stronger alternative such as love. (31) a. Atlas-Levinson scale b. like +> love, … c. +> (There is a possibility that the speaker epistemically believes that) a semantically stronger meaning such as ‘love’ is the case. Therefore, a more accurate term for non-canonical scalar implicature may be ‘Q-scalar/I-implicatures’.

Conclusions My conclusions are: (i) Sometimes a scalar implicature, unlike its standard use, can be used markedly and non-canonically. (ii) There are normally some marked motivations behind the non-canonical use of scalar implicature. They may be cultural, social, political, etc.

50

Y. Huang

(iii) This non-canonical use of scalar implicature is implicated in a classical way, with maximum theoretical parsimony, from Grice’s co-operative principle and its component maxims of conversation. Non-canonical scalar implicatures can be given an elegant account of in neo-Gricean pragmatics. And, (iv) Within the neo-Gricean framework, non-canonical scalar implicatures are better renamed as ‘Q-scalar/I-implicatures’

References Atlas, J. D., & Levinson, S. C. (1981). It-clefts, informativeness and logical form: Radical pragmatics. In P. Cole (Ed.), Radical pragmatics (pp. 1–61). London, UK: Academic Press. Bonnefon, J.-F., Feeney, A., & Villejoubert, G. (2009). When some is actual all: Scalar inferences in face-threatening context. Cognition, 112, 249–258.  https://doi.org/10.1016/j. cognition.2009.05.005. Brown, P. (2017). Politeness and impoliteness. In Y. Huang (Ed.), The Oxford handbook of pragmatics (pp. 383–399). Oxford, UK: Oxford University Press. Brown, P., & Levinson, S. C. (1978). Politeness: Some universals in language usage. In E. N. Goody (Ed.), Questions and politeness (pp. 56–310). Cambridge, UK: Cambridge University Press. Brown, P., & Levinson, S. C. (1987). Politeness: Some universals in language usage. Cambridge, UK: Cambridge University Press. Chen, R. (2010). Compliment and compliment response research: A cross-cultural survey. In A. Trosborg (Ed.), Pragmatics across cultures and languages (pp. 79–101). Berlin, Germany: Mouton de Gruyter. Dumitrescu, D. (2006). Noroc!, Merci!, Qué Linso!, Sorry: some polite speech acts across cultures. Southwest Journal of Linguistics, 25, 1. Gazdar, G. (1979). Pragmatics: Implicature, presupposition and logical form. London, UK: Academic Press. Geurts, B. (2009). Scalar implicatures and local pragmatics. Mind and Language, 24, 51–79. https://doi.org/10.1111/j.1468-0017.2008.01353.x. Grice, H. P. (1975). Logic and conversation. In P. Cole & J. Morgan (Eds.), Syntax and semantics 3: Speech acts (pp. 41–58). London, UK: Academic Press. Grice, H. P. (1989). Studies in the way of words. Cambridge, MA: Harvard University Press. Herbert, R. (1991). The sociology of compliment work: An ethnocontrastive study of Polish and English compliments. Multilingua, 10, 381–402. https://doi.org/10.1515/mult.1991.10.4.381. Hirschberg, J. (1985/1991). A theory of scalar implicature. PhD dissertation. University of Pennsylvania. Published as (1991). A theory of scalar implicature. New York, NY: Garland. Ho, D. (1976). On the concept of face. The American Journal of Sociology, 81, 867–884. Holmes, J. (1986). Compliments and compliment responses in New Zealand English. Anthropological Linguistics, 28, 485–508. Horn, L. R. (1972). On the semantic properties of logical operators in English. Ph.D. Dissertation. Los Angeles, CA: University of California. Horn, L.  R. (1984). Toward a new taxonomy for pragmatic inference: Q-based and R-based implicature. In D. Schiffrin (Ed.), Meaning, form, and use in context: Linguistic applications (pp. 11–42). Washington, DC: Georgetown University Press. Horn, L. R. (1989). A natural history of negation. Chicago, IL: University of Chicago Press. Horn, L. R. (2010). From if to iff: Conditional perfection as pragmatic strengthening. Journal of Pragmatics, 32, 289–326. https://doi.org/10.1016/S0378-2166(99)00053-3. Huang, Y. (1991). A neo-Gricean pragmatic theory of anaphora. Journal of Linguistics, 27, 301–335. https://doi.org/10.1017/S0022226700012706.

I like you may actually implicate ‘I love you’

51

Huang, Y. (1994). The syntax and pragmatics of anaphora. Cambridge, UK: Cambridge University Press. Huang, Y. (2000). Anaphora: A cross-linguistic study. Oxford, UK: Oxford University Press. Huang, Y. (2007). Pragmatics. Oxford, UK: Oxford University Press. Huang, Y. (2010a). Implicature. In L.  Cummings (Ed.), The pragmatics Encyclopedia (pp. 234–238). London, UK: Routledge. Huang, Y. (2010b). Scalar implicature. In L.  Cummings (Ed.), The pragmatics encyclopedia (pp. 441–444). London, UK: Routledge. Huang, Y. (2014). Pragmatics (2nd ed.). Oxford, UK: Oxford University Press. Huang, Y. (2015). Neo-Gricean pragmatic theory of conversational implicature. In B.  Heine & H. Narrog (Eds.), The Oxford handbook of linguistic analysis (2nd ed., pp. 615–639). Oxford, UK: Oxford University Press. Huang, Y. (2017a). Neo-Gricean pragmatics. In Y. Huang (Ed.), The Oxford handbook of pragmatics (pp. 48–78). Oxford, UK: Oxford University Press. Huang, Y. (2017b). Implicature. In Y.  Huang (Ed.), The Oxford handbook of pragmatics (pp. 155–179). Oxford, UK: Oxford University Press. Huang, Y. (Ed.). (2017c). The Oxford handbook of pragmatics. Oxford, UK: Oxford University Press. Huang, Y. (2018a). Unarticulated constituents and neo-Gricean pragmatics. Language and Linguistics, 19, 1–31. https://doi.org/10.1075/lali.00001.hua. Huang, Y. (2018b). Research methodology in classical and neo-Gricean pragmatics. In A. H. Jucker, K. P. Schneider, & W. Bublitz (Eds.), Methods in pragmatics (pp. 155–183). Berlin, Germany: Mouton de Gruyter. Huang, Y. (2019a). Scales in neo-Gricean pragmatics. Unpublished MS.  The University of Auckland. Huang, Y. (2019b). Implicature. In J. S. Damico & M. J. Ball (Eds.), The Sage encyclopaedia of human communication sciences and disorders (pp. 881–882). Thousand Oaks, CA: Sage Publications. Huang, Y. (2020). Non-canonical scalar implicatures, face/politeness consideration, and neo-­ Gricean pragmatics. Waiyu Jiaoxue yu Yanjiu, 52(1), 25–39. Huang, Y. (In preparation). Conversational implicature. Contracted to be published by Oxford, UK: Oxford University Press. Katzir, R. (2007). Structurally-defined alternatives. Linguistics and Philosophy, 30, 669–690. https://doi.org/10.1007/s10988-008-9029-y. Keenan, E. O. (1976). The universality of conversational postulates. Language in Society, 5, 67–80. https://doi.org/10.1017/S0047404500006850. Kuhn, T. S. (1962/2012). The structure of scientific revolutions. (The 50th anniversary edition). Chicago, IL: University of Chicago Press. Leech, G. N. (1983). Principles of pragmatics. London, UK: Longman. Leech, G. N. (2007). Politeness: Is there an east-west divide? Journal of Politeness Research, 3, 167–206. https://doi.org/10.1515/pr.2007.009. Levinson, S. C. (1983). Pragmatics. Cambridge, UK: Cambridge University Press. Levinson, S. C. (1987). Pragmatics and the grammar of anaphora: A partial pragmatic reduction of Binding and Control phenomena. Journal of Linguistics, 23, 379–434. https://doi.org/10.1017/ S0022226700011324. Levinson, S. C. (2000). Presumptive meanings: The theory of generalized conversational implicature. Cambridge, MA: MIT Press. Matsumoto, Y. (1995). The conversational conditions on Horn scales. Linguistics and Philosophy, 18, 21–60. https://doi.org/10.1007/BF00984960. Nature. (2015). Editorial: Let’s think about cognitive bias. Nature, 526(7572), 163. Nwoye, O.  G. (1992). Linguistic politeness and socio-cultural variations of the notion of face. Journal of Pragmatics, 18, 309–328. https://doi.org/10.1016/0378-2166(92)90092-P. Popper, K. (1945). The open society and its enemies. London, UK: Routledge.

52

Y. Huang

Sauerland, U. (2004). Scalar implicatures in complex sentences. Linguistics and Philosophy, 27, 367–391. https://doi.org/10.1023/B:LING.0000023378.71748.db. Senft, G. (2008). The case: The Trobriand islanders vs. H.  P. Grice  – Kilivila and the Gricean maxims of quality and manner. Anthropos, 103, 139–147. Soames, S. (1982). How presuppositions are inherited: A solution to the projection problem. Linguistic Inquiry, 13, 483–545. Watts, R.  J., Ide, S., & Ehlich, K. (Eds.). (2005). Politeness in language studies in its history, theory and practice (2nd ed.). Berlin, Germany: Mouton de Gruyter.

Pragmatics and Grammar as Sources of Temporal Ordering in Discourse: The Case of And Kasia M. Jaszczolt and Roberto B. Sileo

Introduction: Anaphoric Tenses, Anaphoric Scenarios It goes without saying, and can be accepted without a need for empirical corroboration, that sentences such as (1) and (2) are interpreted in such a way that the second event is understood as following the first, and as such as dependent on the first in some way. ( 1) Mary broke her leg and went to hospital. (2) Mary had her supper and went to bed. The relation of dependence can be stronger or weaker, can consist in the causal link as in (1) or in merely setting the scenario as in (2), but the temporal sequence seems to be there by default. The order of the events can be manipulated by supplying a specific context but the fact remains that in the absence of such a specific context, speakers and addressees are governed by a strong sense of a default, temporal interpretation. (1a) and (1b) demonstrate that this contextual manipulation can be difficult when the standard scenario is well entrenched. (1a) Two things happened to Mary yesterday: she broke her leg and went to hospital. But not in this order: she went to hospital for some tests and broke her leg on a slippery hospital floor. ? (2a) Mary didn’t do much last night: she had her supper and went to bed. But not in this order: she had her supper in bed. ?

K. M. Jaszczolt (*) · R. B. Sileo Faculty of Modern and Medieval Languages and Linguistics, University of Cambridge, Cambridge, UK e-mail: [email protected] © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 F. Macagno, A. Capone (eds.), Inquiries in Philosophical Pragmatics, Perspectives in Pragmatics, Philosophy & Psychology 28, https://doi.org/10.1007/978-3-030-56696-8_4

53

54

K. M. Jaszczolt and R. B. Sileo

The semantic and pragmatic properties of and coordination have been discussed at length in the linguistic literature at least since Grice’s (1978) Modified Occam’s Razor compelled many pragmaticists to place the temporal meaning in the domain of conversational implicature. The discussion has gathered new momentum in the recent years in that treating semantic ambiguities as theoretically burdensome can no longer be taken for granted as a viable theoretical move in view of their increasing importance for various specific objectives associated with approaching meaning in computational linguistics. The Gricean argument that the ‘and then’ or ‘and therefore’ interpretation of and are to be sought in pragmatic inference rather than semantic content, or, alternatively, as in post-Gricean contextualism, that these aspects are provided via pragmatic inference but intrude into semantic representation, lose force somewhat when they are juxtaposed with the argument from the utility of specifying the available set of meanings (and, often, their frequencies as predicted from large corpora). When semantics is built ‘bottom up’, from individual interpretations to generalizations, as it is in the case of distributional semantics in computational linguistics, ambiguity ceases to be seen as a methodological burden. Instead, it becomes a trigger for listing the meanings that can be associated with a sentence. Seen against this background, various theories peg the temporal meaning of conjunction on various factors. Some emphasise the role of the aspectual class of the connected eventualities; others stress the role of grammar, and in particular the role of grammatical tenses. Both of these orientations ground these sources of temporality in the overall assumption of discourse coherence. On the other hand, post-­ Griceans remain committed to pragmatic inference as an explanans. But on a closer inspection, the differences between the orientations seem to be less rigid than some of the participants in the dispute allow them to be: when Gricean implicatures are seen as context-free presumptive interpretations, they are strong – almost as strong as the meanings provided by the grammar. We will say more about this evaluation in section “The Meaning of Sentential And: Pragmatics or Grammar?”. In short, the current state of the art offers a medley of superficially different but in fact very similar views on temporal anaphora where different aspects on which the ‘and then’ meaning can be pegged are emphasised or downplayed. In this paper we propose to put together the currently dominant views and compare and contrast them critically, in order to come up with the answer to some metatheoretic questions such as (i) whether it makes theoretical sense to attempt to choose between grammar- and pragmatics-based explanations of temporal ordering with and – that is, whether the objectives of the relevant approaches are the same, or at least compatible and can be recalibrated. In order to do so, we have to at least make some progress with answering such questions as (ii) what factors are decisive about the presence or absence of the temporal interpretation in a non-biased context, as well as (iii) what role exactly (iii.a) assumed scenarios as well as (iii.b) biasing contexts have to play. We end up by offering a tentative unified proposal of a generalized account of temporal ordering in discourse that would incorporate the factors emphasised by various extant and partially successful views and assign them their rightful place.

Pragmatics and Grammar as Sources of Temporal Ordering

55

The structure of the paper is as follows. First, in section “The Meaning of Sentential And: Pragmatics or Grammar?”, we briefly introduce the seminal approaches to propositional and coordination1 and highlight the examples and phenomena that they struggle with. We follow by assessing the current status of the debate and begin to build a case for a unified account. Section “And Coordination in the Absence of Simple Past: Vox Populi” presents a handful of case studies, following the methodology we defend here according to which, due to the nature of the phenomenon, what is required is an in-depth, essentially qualitative study of potential counterexamples across relevant scenarios. Section “And and the Degree of Eventhood” focuses on some new questions these case studies highlight, in particular a hypothesis concerning the role of the degree of eventhood in the interpretation of propositional conjunction and. Section “Conclusions and Future Prospects” concludes with a summary of our argument in favour of a unified account of temporal and coordination that, as we argue, covers a wider domain and at the same time fulfils the subject-specific objectives of its alternatives.

The Meaning of Sentential And: Pragmatics or Grammar? From Implicature to What Is Said (and Back to the Lexicon) Paul Grice’s (1975) maxim of Manner, and in particular its fourth sub-maxim ‘Be orderly’, summarises the following aspect of rational conversational behaviour: events can be expected to be narrated in the relevant order, the most common of which is the chronological order in which they occurred.2 Closer to the state of the art in Gricean pragmatics, the meanings of and that are richer than that of logical conjunction can be traced to (i) Larry Horn’s (1984, 1988) R principle, accounting for saying not more than one must3 and to (ii) Stephen Levinson’s (1987, 2000) I-principle of minimization of content. In the latter, the enriched meaning of and is regarded as a generalized conversational implicature (henceforth GCI), understood rather intra-theoretically within his theory of presumptive meanings.4 Following the arrival of radical pragmatics and the progressive inclusion of pragmatic meanings in the truth-conditional content initiated by what is now known as the Atlas-Kempson 1  We choose the term ‘propositional’ rather than ‘sentential’ because sentence structure is not always isomorphic with the coordination of events: ‘e1 and e2’ need not correspond to ‘s1 and s2’. Examples discussed in section “Examples: Five Case Studies” will be pertinent here. 2  See Horn (2018) on the extensive classification of types of ordering relations used in coordinated conjunction in discourse. 3  Taken in tandem with the Q principle, accounting for saying as much as one can. 4  The main distinctive feature of Levinson’s GCIs is their strength, in that they are facilitated by the language system itself, and their frequent sub-sentential origin. See Levinson (2000:1) and below. See also Carston (2002: 259, fn 1) on the differences between Grice’s and the neo-Gricean emphasis on GCIs.

56

K. M. Jaszczolt and R. B. Sileo

thesis,5 Robyn Carston (1988, 1993, 2002) opts for classifying such meanings as pragmatic enrichments belonging to what Relevance theorists call an explicature: the explicit, truth-conditional content of the utterance, pertaining to the output of the syntactic processing of the sentence and, frequently, also the developments of the logical form arrived at in the process of utterance interpretation. It is not the place to rehearse the debates on the semantics/pragmatics boundary6 but suffice it to say that proposals bourgeoned on all sides: (i) pragmatic, represented by neo-Griceans; (ii) contextualist semantic, represented by Relevance theorists (e.g. Carston, 1993, 2002; Sperber & Wilson, 1995) and Recanati’s early version of what came to be known as truth-conditional pragmatics (Recanati, e.g. 1989, 2010); as well as (iii) lexical-semantic that advocated rich semantic content and appeared as a response to Grice and his followers who equated the semantics of the English and with that of logical conjunction (Cohen, 1971; Bar-Lev & Palacas, 1980). Groups (i) and (ii) share the commitment to the pragmatic origin of the component of meaning that stands for temporal and causal ordering. But, interestingly, the characteristics of this commitment vary from what we can call ‘uncontaminated pragmatic’ to ‘pragmatic qua language system’ – the latter in a variety of forms. To explain: Grice derived the implicature of temporal and causal connections from general principles of human conversational behaviour tout court. Levinson, on the other hand, makes the relevant GCI language-system dependent: Utterance-type meanings are matters of preferred interpretations (…) which are carried by the structure of utterances, given the structure of the language, and not by virtue of the particular contexts of utterance.  Levinson (2000: 1)7

He also makes them local: it is the lexical item in its specific syntactic configuration that is responsible for the GCI. So, Levinson’s (2000) GCIs also derive these aspects of meaning from the structure: it is the lexical item and, as a component of the language system, that triggers the temporal or causal reading. It is possible to see this system-based solution as a step in the direction towards grammatical solutions. For example, the GCIs associated with (3) are listed in (3a). (3) John turned the switch and the motor started. (3a) p and then q, p caused q, John intended p to cause q, etc. (adapted from Levinson, 2000: 38). The list in (3a) is incomplete, so the proposal that GCIs are generated by the language system still leaves a lot of room for inference in context. But the very fact that these implicatures are local, attributed to the conjunct itself, and derived from the language system, places this solution on a sliding scale away from pure inferentialism. We return to this question while assessing the grammar-based approaches later on in this section. Broadly in a similar vein, Horn (2018) focuses on conventional patterns and as such also on standard structural units. In the recent work on coordination, he searches for patterns and conventions, recognising the fact that although types of  See Atlas (1977, 1979, 1989) and Kempson (1975, 1979, 1986).  For an extensive introduction see e. g. Jaszczolt (2012). 7  Our emphasis. 5 6

Pragmatics and Grammar as Sources of Temporal Ordering

57

ordering can vary, they follow certain patterns and adhere to certain conventional, or even natural, principles of ordering of ideas. So, the ‘etc.’ in Levinson’s (3a) indicates an open list but it is now a list to which systematicity has been added by Horn. Arguably, this systematicity is also in line with Carston’s account (1994: 697, 2002: 226) in which she mentions relying on stereotypical situations and frequently used scripts. In her essentially pragmatic account, Carston (1993, 1994) also points out that the structure of the sentence carries with it a strong assumption of temporal order: it is the fact that “[a]n and-conjunction is a single syntactic unit; that is, it constitutes one of the infinitely many sentences that the grammar generates” that triggers this enriched meaning in that “[t]he simplest assumption to make here, surely, is that an utterance unit is in some fairly direct correspondence with a grammatical unit” (Carston 1993: 41). On the other hand, merely juxtaposing two sentences does not carry the assumption of any such additional meaning. It looks like it is not the orderliness of clauses, focused on for example by Bar-Lev and Palacas (1980) in their proposal of semantic command, but rather the presence of and that can be credited with this extra meaning. Examples to corroborate the latter claim are ample and have been well rehearsed.8 Now, as we argue in what follows, a strict distinction between pragmatic and grammatical solutions is not achievable. In addition to displaying a sliding scale of reliance on the structure as the explanans, pragmatic solutions are compatible with syntactic solutions. Partee’s (1973) argument that tenses are structurally analogous to pronouns can be used in conjunction with many of the above views. If tenses pick out the time deictically or anaphorically, then they contain information about the relative ordering of the eventualities in the conjoined clauses. If temporal sequence can be attributed to the t-variable in the syntactic representation, then there is no longer any need to attribute it to the connective and itself: it can be added via pragmatic inference but at the same time it finds its representation in the logical form.9 The approach discussed in section “Finding It in the Grammar” below moves close to the original spirit of Partee’s proposal and attributes the temporal sequence meaning to grammar. It is this particular proposal that will trigger our case studies in section “And Coordination in the Absence of Simple Past: Vox Populi”. Finally, it is also feasible to attribute the rich meaning to the connective and, tout court. In other words, the lexical semantics can arguably comprise all of the standard relevant senses, to be subtracted, so to speak, from the interpretation on a particular occasion of the use of and. This is in a nutshell Cohen’s (1971) solution. We will have little to say about this lexicon-based approach in what follows and will focus instead on the grammar-pragmatics debate, also assessing the degree to which it is indeed a debate with different debating sides rather than, as we argue, a matter of the level of description and foregrounded objectives in what are essentially compatible explanations. 8  Semantic command states that the second conjunct is not prior to the first conjunct either chronologically or causally (adapted from Bar-Lev & Palacas, 1980: 141). For a discussion of counterexamples see Bar-Lev & Palacas (1980) and Carston (1993), i. a. 9  See Partee (1973: 57, n3) and Carston (1988: 161).

58

K. M. Jaszczolt and R. B. Sileo

Coherence-Based Views Maximisation of coherence of discourse is one of the facets of rationality of human communication that is emphasised and theoretically exploited in many approaches to discourse meaning. Segmented Discourse Representation Theory (SDRT, Asher and Lascarides, henceforth A&L, e.  g. 2003) focuses on it as the underlying principle behind rhetorical structure rules, where some of these rules pertain to the behaviour of sentential and in conjoined phrases such as (4).10 (4) Mary broke her leg. She was taken to hospital. Narration caters for the temporal ordering of events: “[i]f Narration (α, β) holds, and α and β describe the eventualities e1 and e2 respectively, then e1 occurs before e2.” (Lascarides and Asher  1993: 443). The reasoning summarised in this rule is defeasible: the rule captures a robust default. And can also come with the causal connection (Result) or signal two parallel eventualities (Parallel), so all these three rules have been grouped as the so-called ‘coordinators’. Gómez Txurruka (2003) derives their functions from the so-called Coordinated Discourse Topic. In the mechanism she proposes, there is no need, for example, to propose a temporal default because and coordination can only trigger defeasible meanings that fall under coordinators (of which Narration is one). What interests us here is the status of these rules vis-à-vis the grammar-­ pragmatics dispute concerning the origin of temporal, causal and other enriched meanings of and.11 In SDRT, this meaning originates in the middle, so to speak, in what A&L call glue logic. Glue logic caters for the rules of nonmonotonic reasoning, ‘if p then normally q’. The glue language is the language of reasoning that takes us from the information content to what A&L call ‘information packaging’. The fact that they distinguish the logic of information content from the logic of information packaging and identify glue logic, as placed, so to speak, ‘in the middle’, is significant. It allows them to model discourse meaning as different from pure content on the one hand, and from belief states on the other, where the latter are exploited, for example, in Gricean approaches. The difference is as follows. The concept of meaning adopted in Gricean approaches goes beyond the minimal, formalizable, purely lexicon- and structure-based nonmonotonic inference. In the interest of cognitive plausibility, it can even extend to any kind of going beyond the logical form of the sentence that in a given context leads to the main intended message, irrespective to  We are interested here in a narrower phenomenon of propositional and conjunction that is governed by more restrictions than sentence juxtaposition. For example, while the rule of Explanation applies to (i), it does not apply to (ii): the latter can, at best, convey an ordering of enumeration, unless the context strongly dictates a scenario- or goal-specific meaning. 10

(i) Mary broke her leg. She had a skiing accident. ? (ii) Mary broke her leg and she had a skiing accident. 11  We are using ‘enriched’ here in the a theoretic sense, bearing in mind the fact that some of the discussed approaches argue against pragmatic, inferential enrichment and in favour of languagebased conventions. ‘Enriched’ means here richer, or more specific meaning than that of the corresponding connective in propositional logic.

Pragmatics and Grammar as Sources of Temporal Ordering

59

its relation to that logical form (as in Default Semantics, Jaszczolt, 2005, 2010). In SDRT, on the other hand, there is a clear structural constraint on these enriched meanings. The logical form of the discourse captured in the discourse representation structures pertains to what they call ‘shallow understanding’ (A&L 2003: 247), that is those aspects of processing that come from aspects of the linguistic structure. Cognitive plausibility and ‘speaker meaning’ do not drive the delimitation of theoretical concepts. Now, if the aim of a theory of discourse is set out in the way that aims at representing this ‘shallow understanding’, then we have two choices: either to distinguish some form of a middle player such as a glue language, or to conflate these structure-­ based defeasible inferences with other aspects of meaning derivable from the grammar and, so to speak, build conventions into the products of grammar. The latter solution is explored in the following section.

Finding It in the Grammar The theoretical background discussed so far leads us to the view that is the main object of our critical scrutiny: the grammar-based account of the temporal reading of propositional and coordination. So, in what follows we not only present the view but intersperse the presentation with the critical analysis of various aspects of it as, and when, they appear in our discussion. In their much debated book Imagination and Convention: Distinguishing Grammar and Inference in Language, Lepore and Stone (henceforth L&S, 2015) place a lot of weight and explanatory value on the particular, single, isolated structure in (5). (5) Oil prices doubled and demand for consumer goods plunged.  (from L&S, 2015: 115) They propose that it is the past tense of the verbs in both clauses that triggers the succession meaning of and. Substituting Present Perfect forms eradicates, or at least substantially reduces, the possibility of this interpretation. They appeal to linguistic conventions to explain this temporal reading, suggesting that (i) grammar sets out more than one interpretation (resulting in the sentence’s ambiguity) and (ii) linguistic conventions, here embodied in the conventional function of the Simple Past, select the preferred reading. Grice’s sub-maxim ‘be orderly’ is deemed inadequate in that, first, it does not specify what kind of ordering is referred to and, second, it presupposes a pragmatic process of enrichment which, in their view, is redundant.12 Unsurprisingly, this grammar-based account triggers a plethora of questions. First, and most importantly, even if we concede that the temporal interpretation of (5) is arrived at by convention rather in the process of pragmatic inference, it is not clear why conventions ought to be pinned on linguistic conventions. L&S say that

 See also Horn (2016) for a discussion of differences with respect to the content of the truthconditional representation vis-à-vis the output of conventions with which we will not engage in here.

12

60

K. M. Jaszczolt and R. B. Sileo

the choice of a linguistic convention such as the deployment of a past tense triggers the sense of temporal succession. But, surely, conjuncts that use other grammatical tenses, or no tenses at all, can also coincide with conventional (in the broad sense of the term) temporal interpretations, that is interpretations in the case of which temporal succession cannot be attributed to the existing context. Likewise, conventions in the case of (5) are equally likely to come from the assumed scenario that involves either causation or, say, gradual progression of the economic crisis. Even if accumulating such examples does not result in a knock-out argument, it demonstrates that a methodologically preferable explanation, an explanation that covers a wider array of examples, could be found beyond grammar. In other words, instead of listing factors that may, in the case of different sub-sets of examples, be responsible for the temporal reading, it is methodologically preferable to search for an explanation that subsumes them all. From now on, we will adopt this methodological assumption and, having addressed the question of the relationship between the grammatical and pragmatic approaches, will proceed to exploiting it in our case studies. As we demonstrate in section “And Coordination in the Absence of Simple Past: Vox Populi”, a tenseless construction can lead to a strong temporal reading without any overt or salient assumed contextual biasing. Also, a strong presumption of causation in the absence of a past tense marker can shift the results uniformly to temporal succession. While these results are not incompatible with a grammar-based approach to anaphoric past-tense clauses, they open up the possibility of a more general explanation based on the conceptual category at large – the category of temporal succession that is sometimes conveyed by anaphoric past tenses and at other times by lexical, other structural, or outright pragmatic means. In other words, juxtaposed with the fact (that is not in need of testing) that, conversely, coordinated clauses with past tense verb forms yield themselves to various types of ‘orderliness’, including for example the order based on importance, we will begin to obtain a bigger picture that will then allow us to search for the true triggers of temporal, and ultimately any enriched,13 reading of and. So, if there are conventions, they are more likely to pertain simply to ‘generalised inferences’, so to speak, and as such fall under the Gricean program or under speech-act-theoretic generalizations.14 Before anticipating too far ahead, let us consider more detailed tenets of L&S’s proposal. Following Partee (1973),15 L&S claim that the Simple Past form of the verb is interpreted analogously to a pronoun, so, like an indexical. The anaphoric use of Simple Past triggers a linguistic convention of temporal succession. But if Simple Past form is an indexical term, then it would appear that the role of ‘linguistic convention’ as an explanans is redundant: we can simply pin the meaning on the grammar tout court, just as we can attribute reference assignment to the presence of slots in the logical form created by pronouns. Just as an indexical expression can

 See fn 11.  See Jaszczolt (2019) on the determinants of ‘being Gricean’ and Leezenberg (2006) on the compatibility of the Gricean program with convention-based accounts of meaning. 15  Discussed in section “From Implicature to What Is Said (and Back to the Lexicon)”. 13 14

Pragmatics and Grammar as Sources of Temporal Ordering

61

lead to different resolutions, so can the presence of coordinated past tense forms. So, this ‘double insurance’, so to speak, by grammar and by conventions, engenders redundancy. Chierchia’s (2004) proposal of the grammatical origin of scalar implicatures for example, arguably the closest in spirit to L&S’s own, does not resort to such moves. On Chierchia’s view, grammar produces a strong reading and a weak reading, and whatever preferences there may be among the derived readings are not the business of grammar. Needless to say, this type of grammar-based solution leaves much less scope for controversy. Next, somewhat analogous to Chierchia’s account, L&S’s proposal assumes the ambiguity view. More specifically, L&S (2015: 119) claim that grammar produces different alternatives and it is tense that enables the choice of the most likely one: With tense, then, as with speech act ambiguities, understanding involves choosing the reading that makes the most sense from the candidates delivered by the grammar. Understanding does not involve enriching those readings to transform them from ones that do not make sense to the ones that do.

But, as far as the employment of ambiguity is concerned, any comparison and contrast with the Gricean program would be talking at cross purposes – a fact that L&S fail to recognise, as the above quotation demonstrates. Ambiguity is a useful assumption indeed for those research programs that benefit from enumerating possible meanings. It is common practice in computational linguistics (for example in distributional semantics) to begin with collecting collocations from large corpora, only to end up with possible meanings that words and structures can adopt. This is a relatively ‘psychologism-free’ approach to meaning that is in stark contrast to many post-Gricean contextualist theories, as we discussed in section “Coherence-­ Based Views” in the context of A&L’s modelling of ‘shallow understanding’. So, if we are to pursue the question of the grammatical vis-à-vis pragmatic origin of temporal reading of and, it has to be pursued in isolation from the question of the underspecification/ambiguity debate triggered by Grice’s Modified Occam’s Razor and developed so abundantly in the post-Gricean semantic/pragmatics boundary disputes.16 In short, opting for the ambiguity approach has to be treated as a methodological choice, dictated by the desiderata of computational semantics. Criticising it on the grounds of failing the principle of economy (such as Grice’s Modified Occam’s Razor) would mean committing a fallacy. Equally, criticising it for not capturing speaker’s cognitive states à la Grice would be committing a fallacy. This, however, does not invalidate our earlier argument: the claim that it is the grammar that produces these alternatives remains a contentious assumption in that it makes grammar more powerful than most extant understandings would justify. Next, there also seems to be a misunderstanding of the concept of pragmatic enrichment. A GCI, or, on some post-Gricean accounts, an enrichment of the logical form that produces temporal, causal or other meaningfully ordered readings, does not necessarily involve a conscious inferential process of transforming meaningless 16

 For a comprehensive overview of the debate see e. g. Jaszczolt (2012).

62

K. M. Jaszczolt and R. B. Sileo

sentences into meaningful ones. All post-Gricean approaches alike acknowledge the role of conventions, shared information, default scenarios, and the like in short-­ circuiting the inferential process. The terminology and emphasis placed on these shortcuts vary from theory to theory but they are always there, be it as Levinson’s GCIs that stem from the language system, or Relevance-theoretic heuristics that make conscious inference unnecessary, or as cognitive or socio-cultural and world-­ knowledge defaults in Default Semantics. This is an important clarification in that it removes one aspect of the grammar-pragmatics dispute from consideration, relegating it to a terminological misunderstanding. All in all, our main argument is as follows. If tenses (i) are analogous to pronouns à la (Partee, 1973), (ii) produce slots in the logical form à la (Stanley & Szabó, 2000), and (iii) are represented as variables ranging over temporal intervals à la DRT (Kamp & Reyle, 1993), this does not yet invest them with the power to relate the intervals. The relations of succession, simultaneity or preference/importance can be equally strong for various sentence strings in various default scenarios. Conversely, if, as L&S claim, “temporal reference is not just a feature of grammatical tense; it is a ubiquitous part of interpretation” (p. 122), then there does not seem to be any theoretical advantage to, so to speak, ‘splitting the explanation’ between the past tense form when it is present and other triggers when it is not present – triggers that can produce exactly the same effect, and in addition produce it so often, out of context, and for a high proportion of speakers, as our examples in section “And Coordination in the Absence of Simple Past: Vox Populi” demonstrate. In this context, one also wonders if L&S would not opt for temporal symmetry and extend their hypothesis to future tense in order to capture the sequential reading in (6). (6) I will write a novel and make a lot of money. Clearly, the temporal and causal reading is the default one here but it is much more plausible that this is so due to the default scenario it triggers, namely that the sales of a popular novel can bring substantial royalties. And yet, following L&S’s line of argumentation, a symmetrical grammatical solution would be preferred by methodological parsimony. Even if it is the case that, as they claim, Simple Past in English by default “draws attention to the interval immediately afterwards” (p. 124), it is very likely that this direction of attention can only accidentally, and superficially, so to speak, be attributed to the grammatical tense; the true explanation lies underneath, on a more general level of invoked scenarios  – scenarios that are expressed with, or without, the help of the grammatical tense. In short, maximising coherence, invoking the ‘normally’ rule in defeasible inferencing may exploit the grammar, but likewise it may exploit the lexicon and all that the grammar, lexicon, default scenarios, or pragmatic inference bring to mind. Focusing on grammar is like wearing blinkers: it offers a tree when we want to see the whole forest. There are likely to be various pertinent explanantia involved but they are not governed by linguistic rules.17 So, broadly analogous to SDRT, we have here a move away from Gricean explanations of utterance interpretation where general, maxim- or principle-governed 17

 See also Stojnić, Stone, & Lepore (2017) on this topic.

Pragmatics and Grammar as Sources of Temporal Ordering

63

reasoning underlies the derivation of the enriched meaning of and. We have an assumption that the representation of discourse should not stretch to speakers’ cognitive states but rest with, to use A&L’s term, ‘shallow understanding’ whose benefits are diaphanous when one’s goals fall within the domain of computational linguistics. But there is a substantial difference. While conventions accounted for by ‘glue logic’ in SDRT are well supported by common conversational practice, the common-sense layman’s explanation of which, arguably, provides them with external justification, putting the conventions in the grammar seems to be an intra-­ theoretic move, justified only by L&S’s intra-theoretic objectives. If those suffice, then we can rest the case and embrace the differences of goals. But the metatheoretic question of compatibility, and the all-important follow-up possibility of kicking the explanation, so to speak, to the level at which it embraces the entire conceptual phenomenon, ought to be rehearsed first.

And Enrichment: Finding the Culprit The next move is to take a step back from the debate and clearly identify the points of convergence and divergence. There are different ways of approaching the attribution of information to the grammar. First, attribution can mean that the semantic representation that corresponds to the syntactic structure is adapted, reconceptualised, so as to cater for the aspects of meaning in question, analogous to already widely acknowledged functional categories in generative grammar. Next, it can mean that certain overt categories have the status of indexical expressions – here the attribution of the temporal reading of and to the Simple Past in English is a good example. As we have seen, there is a reasonably compelling methodological argument available for building the specific meanings of and into the grammar. The relevant procedure is to agree, by methodological assumption, that the indicators of temporal (and possibly also causal, and other order-based) meanings of and are produced as part of the output of syntactic processing on a par with the indicators produced by indexical expressions where the latter function as slots, partially semantically constrained, to be filled by the context. The necessary move here is to make indexicals sufficiently pliable to comprise the phenomenon of and conjunction. The question we have asked here is: What exactly would the status of such a solution be? What would it mean for the explanatory power of a theory of meaning to say that grammar produces a slot to be filled with information from the text itself (co-text, as in the case of, say, the past tense of the conjoined clauses), or from the context, if it is necessary to go beyond information from the sentence alone? The answer is, it would not mean much in itself, apart from signalling a certain theoretical commitment. For example, there are tense/time mismatches where the present tense is used for past events as in (7). Would we then want to say that while the

64

K. M. Jaszczolt and R. B. Sileo

temporal meaning of and (and the ordering in general18) is derived from grammar in (7a), it is derived from context via pragmatic inference in (7) where the past tense is absent? Arguably, it would be methodologically preferable to have a uniform, more economical explanation, derived from the common scenario for (7) and (7a). (7) This is what happened to me yesterday. I arrive at the station, I buy the ticket and go to the platform. Then this voice behind me says… (7a) This is what happened to me yesterday. I arrived at the station, I bought the ticket and went to the platform. Then this voice behind me said… Furthermore, tenseless constructions also acquire temporal reading out of context. Does this reading come from pragmatics? And if so, why not start with the assumption that it comes from pragmatics in tensed and in tenseless constructions alike? And, perhaps a fortiori, also in tensed and tenseless languages alike? Clearly, the onus of proof is on the defenders of a fragmented account. Let us now assume, for the sake of argument, that L&S are correct in pinning the temporal succession reading on the tense in clear cases with Simple Past tense such as (5). The t variable is used quite standardly in formal semantics, for example in DRT, where it is assumed that tense introduces the indexical component and as such is in need of temporal reference assignment. But what happens to the t variable when the sentence contains a temporal adverbial such as ‘last week’? Partee (1973), on whose proposal L&S base their solution, explicitly argues that the t element is then redundant. So, for the sake of increasing the scope of application of the solution, and thus in agreement with one of the basic methodological principles by which theories are selected, wouldn’t we prefer to pin the reading directly on and itself? Or, since the temporal sequence is also conveyed by simple juxtaposition as in (7) in the case of tense-time mismatches, and in the (5a) variant of (5), perhaps the temporal anaphoric reading ought to be pinned on social conventions, background knowledge, mutual manifestness of assumptions, socio-cultural default, to bring together relevant concepts from a selection or relevant approaches? (5a) Oil prices doubled, demand for consumer goods plunged, the whole economy went belly-up. Examples we discuss in section “And Coordination in the Absence of Simple Past: Vox Populi” further demonstrate that context as well as default assumptions are extremely powerful tools in fixing temporal reference – much more powerful than grammar-based approaches take them to be. But L&S also invoke another, albeit not independent, strand of explanation, appealing to the rule of Attention (e. g. Grosz, Joshi, & Weinstein, 1995) that states that there is a ranking of candidates, called the attentional state of discourse, that is reflected in the structure. So, it appears that the attribution of the temporal reading to the grammatical tense is just a sub-case of a more general rule that orders the

18  Note that the temporal succession is derived equally easily for the listed events where and is absent. So, the order of narration in itself triggers their temporal reading.

Pragmatics and Grammar as Sources of Temporal Ordering

65

conjuncts. On another view they entertain, the rule of Coherence19 (e.  g. Hobbs, 1990; Kehler, 2002) caters for the interpretive relations between sentences as a higher-level principle (see also A&L, 2003). As such, it can also affect the attention (prominence). In the case of ambiguities (which, to repeat, mean available choices of interpretation of a sentence rather than hearer’s uncertainty as to how to proceed with the interpretation), various interpretive choices can be attributed to the generalizations captured by such rules. In short, in addition to the ‘double insurance’ referred to in section “Finding It in the Grammar”, we also have an appeal to the generalizations of Coherence or Attention, but not in the sense of the ‘middle player’ assumed in SDRT but rather as an external support for the power of grammar. This turns the picture inside out: what would normally allow us to search for a solution outside grammar is now employed in its aid. While the rules pertaining to information structure can be approached from many different theoretical orientations, and arguably, an appeal to them can also be made in the psychological, such as Gricean, approaches to discourse meaning, the difference lies in where exactly they are placed on the syntax/semantics/pragmatics scale. For L&S, their contribution is semantic.20 And it is semantic in a particular sense of ‘semantics’ in that the latter acts as an evaluator, so to speak, of the interpretations provided by the grammar. On post-Gricean contextualist approaches, these rules could also be seen as semantic but in a very different sense of ‘semantics’, afforded by the contextualist (as opposed to relativist) assumption that one interpretation is pursued in one given context and any aspect of meaning that contributes to the proposition derived from the uttered sentence is part of its semantics. This path can sometimes produce representations that are relatively close to the logical form of the uttered sentence in that they allow its developments and some concept modification, and at other times representations that ‘override’, so to speak, the output of the grammar of the sentence in order to capture the primary meaning when the latter is indirectly conveyed, no matter how remote such representations may be from the logical form of the uttered sentence. This open-endedness of representations, in the sense of the lack of morphosyntactic restrictions, is a view proposed in Default Semantics (Jaszczolt, 2005, 2010) that accommodates all these possibilities. What is a pragmatic contribution to the semantic representation in post-Gricean contextualist accounts, on L&S’s proposal is pragmatic qua derived from the options offered by the grammar. So, while discussing L&S’s stance, we will talk of the semantic qua grammatical contribution of the rules of information structure (or rhetorical structure rules) in that, qua semantic, they are the output of grammar. As such, they are linguistic rules, being at the same time conceptual, pragmatic, discourse rules. This comparison shows how L&S’s perspective means, in fact, ‘turning the relations inside out’: what is on the outside of the grammar becomes its inner lining.

 Needless to say, Coherence and Attention can yield different predictions. See Kehler (2002) for some comparison and contrast. 20  On the semantic status of these rules see especially Stojnić et al. (2017). 19

66

K. M. Jaszczolt and R. B. Sileo

It seems that the best way to approach L&S’s double level of explanation is to see it as we have just proposed to see it: as an explanation coming from the structuring of information in discourse (see here A&L, 2003 on ‘information packaging’) but at the same time as an explanation that percolates to different devices actually used in an utterance. In the case of propositional and,21 sometimes it will be the grammar, as in the case Pluperfect and Simple Past combination, at other times background knowledge concerning the mutual relevance of two eventualities, and at yet others some other ‘coherence facilitators’. But even such a ‘percolation view’ would not yet give us a license to adopt L&S account of the Simple Past + Simple Past combinations above the Gricean, rational-orderliness-based generalization. All in all, to repeat, we haven’t presented a knock-out argument against the grammatical origin of the various readings of and. Neither did we attempt one, since, as we explained here through a methodological argument, and further illustrate through some examples in the following section, it is the very fact that non-­ past-­tense constructions readily yield to temporal anaphoric interpretation that has to be explored, and not the question as to what theory to adopt for the past-tense ones. With this tenet in mind, it suffices to present methodological counterarguments as the one above, followed by a discussion of an example of such a non-past construction widely regarded as having a temporal-anaphoric reading. So, if an illustration in the form of examples is called for, a single case-study, that is a pertinent construction subjected to native speakers’ judgements, will suffice to substantially weaken the justification for L&S’s stance. Putting the enriched meaning of and in the grammar and thereby building walls around what is only an instance of a broader phenomenon comes with substantial methodological drawbacks indeed. Therefore, our next step is not to offer experimental evidence against the default temporal reading of past-tense and coordination, neither is it to offer evidence for a temporal coordination of non-past-tense and conjunction. As we explain in section “Metafalsification as a Supplementary Method of Inquiry”, since both are common facts of everyday parlance, neither test is methodologically required. Moreover, neither would be feasible in the first place in addressing this particular question.

And Coordination in the Absence of Simple Past: Vox Populi Metafalsification as a Supplementary Method of Inquiry We have a good reason for not offering an experimental study of and coordination of clauses in past tense. We have equally good reasons for not offering an experimental study of the complement set, that is and coordination of clauses without past-tense marking. An experiment will not fix the issue. Rather, we opted here for a theoretical argument that demonstrates that placing temporal anaphora in the 21

 See fn 1.

Pragmatics and Grammar as Sources of Temporal Ordering

67

grammar is a theoretical choice, but a choice that comes with theoretical and methodological disadvantages. As discussed in the previous section, in addition to the inflated understanding of grammar that such an approach necessitates, there is also the problem of carving off parts of what is conceptually a broader phenomenon; assigning temporal reading to and coordination is by no means limited to clauses with past-tense markers. And if it is not, then temporal anaphora in and coordination requires a different, uniform analysis. What we propose to do now is take a handful of pertinent examples and subject them to careful scrutiny – see how they behave, according to native speakers’ judgements, free of context and in biasing and non-­ biasing contexts. For our purposes it would suffice to show that, in the case of one example of and coordination without past tense marking, a substantial number of native speakers comes up with the temporal interpretation even in the absence of a biasing context. In other words, if, say, an example such as (7) triggers readings with temporal anaphora for a substantial number of speakers, the case for a uniform approach is already justified: the example will show that the temporal interpretation of and can be triggered not by grammar but by something else altogether: assumed scenario, default interpretation of contiguous expressions of events, and so forth. The exact provenance of the temporal anaphoric reading in such cases is a separate question that we will touch upon in section “And and the Degree of Eventhood”. For now, we will discuss the behaviour of a handful of pertinent examples. In other words, we opt for a method that is close to falsification but is not falsification as such. Since, to repeat, the proponents of grammar-based approaches would not deny that such temporal readings can be triggered outside of their domain of interest, we can at best call it metafalsification: providing an example of how strong temporal anaphora can be outside that domain of interest. Metafalsification is understood as part of a metatheoretic inquiry  – an inquiry that addresses foundational questions about theories of meaning. A full-blown experimental study of the domain and the complement domain is not only incompatible with our method but, quite independently, not feasible. While one can control for the grammatical structures and possibly a handful of other variables such as presumed causation or aspectual class of the eventualities, there will always be various aspects of discourse, and of the situation of discourse, that can sway the interpretation one way or another and that cannot be predicted in pilot study and therefore controlled for  – at least not without compromising naturalness of the examples and scenarios. We concluded that the provenance of temporal ordering in discourse does not yield itself to promising quantitative studies. But we believe it yields itself to close observations of individual cases. By asking a substantial number of consultants for their intuitions about the situations to which the examples belong, and, in addition, manipulating the context conditions in which such examples are presented, we obtain information from individual examples that does not suffer from the above drawbacks and at the same time serves the purpose of metafalsification. We tested five such sentences and obtained what we call five case studies (CSs). ‘Case’ is understood for this purpose as a linguistic construction exemplifying a grammatical condition – here the absence of the Simple Past marking on the verbs that are joint by and coordination. To repeat, since we are interested in the

68

K. M. Jaszczolt and R. B. Sileo

conceptual category, coordination is understood as coordination of propositions that is not always the same as the overt coordination of natural language clauses. This will allow us to obtain quantitative data per case study, that is information about the native speakers’ judgements about each example, as part of the overall qualitative study. ‘Case study’ is also understood here as a research strategy: an empirical investigation of a phenomenon using its one instantiation, founded on prior theoretical argumentation, and optionally involving some quantitative evidence. As such, case study is a combination of qualitative and, within it, (minimally) quantitative research.22 It is employed here in the sense justified for pragmatic and metapragmatic analyses by Haugh (2018) where he argues for the role of in-depth analyses of individual transcripts extracted from corpora of conversation.23 We looked into how British English speakers interpret and, in and out of context, in the absence of grammatical conventions usually associated with temporally successive interpretations such as the Simple Past tense. As predicted, the picture seems to be that and in English can trigger temporal interpretations regardless of the grammatical conventions that L&S single out as the pertinent explanans – interpretations that can be manipulated by context to variable degrees. Next, in the following section, we present our case studies and suggest their significance. Finally, in section “And and the Degree of Eventhood”, we offer informed speculation that there may be interesting patterns of possible correlation of temporal interpretations with the ‘degree of eventhood’.

Examples: Five Case Studies We looked at the possibility of temporal (T) and non-temporal (NT) interpretations in five examples of and coordination without consecutive past-tense markings. In order to include a range of grammatical features not customarily associated with T interpretations of and, our testable items included a sentence with both conjuncts in the Present Perfect from (CS1), two gerundial structures (CS2 and CS3), and two bare infinitival phrases (CS4 and CS5). In order to control for the identity of agents in both eventualities and at the same time utilise naturally sounding constructions,

 ‘Case’ in ‘case study’ is also often understood as an event. If we were to construe it so, a ‘case study’ (qua a study of the individual behaviour of a construction) would mean one of the, 90 in total, outcomes of testing a construction in that each sentence was placed in 3 different context conditions and each resulting scenario was then presented to 30 consultants who reported the scenarios they associated with them. Nothing hinges on the granularity of the concept of a ‘case’ but we will not follow this meaning here. 23  Note that Haugh employs a different use of the term ‘metapragmatics’. For different uses for this label see Caffi (2006, 2016) and for a discussion Jaszczolt (2018). 22

Pragmatics and Grammar as Sources of Temporal Ordering

69

our examples were not necessarily standard p and q structures but rather instances of coordination of eventualities that can merely be unpacked as such.24 CS1: CS2: CS3: CS4: CS5:

John’s done some shopping and he’s had a haircut. They saw Patricia selling a laptop and buying a guitar. Luke should avoid eating and swimming. They let the detective have a holiday and move to a new office. Jill must go to the library and write an essay.

Sentences used as CS1–CS5 were extracted from a questionnaire that was designed for a broader task to test the grammatical vis-à-vis pragmatic provenance of different types of constructions. The sentences were then placed in three different context conditions, resulting in three separate surveys (Q1–Q3). Three groups of 30 native British English speakers participated in Q1 (mean age = 20.03, 13F/17M), in Q2 (mean age  =  20.03, 21F/9M) and in Q3 (mean age  =  23.10, 21F/9M).25 Participants were recruited through advertisements placed at various institutions within the University of Cambridge and in university-related forums and social media platforms.26 They were required to confirm their interest to take part via email; they were also notified of the eligibility criteria that they were required to meet. Apart from being native British English speakers, participants had to be over 18 years old, they could not be linguists, and they had to confirm residence in the United Kingdom for the previous 3 years. Finally, as required by the University of Cambridge, participants signed a participation consent form and were offered £6 sterling as compensation. Questionnaires were emailed to them together with the consent and payment forms. Participants completed the questionnaires in their own time. They were requested to depict in full detail the scenario that they first thought about. Table 1 in the Appendix shows the items as they were provided in Q1-Q3. Since no generalizations over the CSs are intended, it suffices to say for the purpose of this study that the questionnaires contained 20 items each.27 From now on we shall focus only on our five CSs. Each sentence was presented to 30 native speakers of English in a context-free condition (Q1), as well as, to further two different groups of 30 consultants, contextualised in such a way as to trigger predicted T (Q2) and NT (Q3) interpretations. We counted T and NT interpretations in each of the three context conditions. It was assumed, in agreement with the metafalsification method introduced in the previous section, that if such sentences in the Present Perfect, with tenseless gerundial forms or with bare infinitival phrases lead to T understandings out of context and in an unbiased context in a fair number of cases, this will considerably weaken the methodological grounds for L&S’s  See fn 1.  Participants’ ages ranged from 18 to 28, 18 to 23, and 18 to 35 in each questionnaire. The design of our study was preliminarily tested by six volunteers (two for each questionnaire) and no material amendments needed to be made. 26  We are grateful to Elaine Schmidt for helping us to find participants through the Cambridge Linguistics Research Study System. 27  The ordering of the items was automatically randomised by the RAND Excel function. 24 25

70

K. M. Jaszczolt and R. B. Sileo

grammar-­based explanation. As we argued in section “The Meaning of Sentential And: Pragmatics or Grammar?”, rather than provide a knock-out argument, or indeed knock-out evidence (which in any case would be easy to explain away), we opt for an ‘inclusion argument’: if and-coordinated clauses in Simple Past behave similarly to and-coordinated clauses with no Simple Past tense marking, then the onus of proof is on those who resist a more general explanation. If the similarity expected to support the ‘inclusion argument’ were to be the case, participants’ responses could be expected to favour a fair proportion of T interpretations in context-free Q1, and possibly also be further swayed to T and NT interpretations, respectively, in Q2 and Q3. In other words, if the and conjunction triggers T interpretations regardless of the grammatical conventions that language users encounter, our metafalsification has been accomplished. Our predictions were supported in CS1-CS5, although they revealed interesting case-by-case differences that opened up a possibility of a new direction of inquiry. The case-by-case differences also further corroborated the correctness of our decision not to opt for a fully-fledged experimental study in that relevant pragmatic variables pertaining to the assumed or otherwise constructed scenarios cannot all be predicted and controlled in advance. To repeat, we are not suggesting that L&S would contest our predictions or object to the limitations of the case studies. Our strategy is to test whether the temporal reading of and coordination is a phenomenon that is sufficiently wide-spread to justify a more ‘inclusive’ explanation. As such, they can only contest the utility of a unified, an essentially pragmatic qua conceptual analysis on the grounds of the differences in the object and purpose of such analyses. For each CS responses were classified following a set of pre-designated procedures. First, if participants clearly indicated two temporally successive events, we considered such responses to be T interpretations. A straightforward T response is illustrated in (8). (8) Patricia’s friends were in town, and saw Patricia in a pawn shop. She was at the counter with a laptop. Later, her friends saw her again, this time in a music shop, stood at the counter with a guitar. (CS2, Q1) Second, we regarded responses as T when it was possible to infer temporally successive understandings from participants’ generic or enumerative descriptions, as illustrated in (9) and (10). (9) Luke vomited in the pool. (CS3, Q1) (10) John woke up hung-over and decided it was one of those days when he can’t revise so he should get some other stuff done. After showering he went out to do shopping and get his hair done then felt more refreshed. He went home, ate subway and had a nap. (CS1, Q1) Third, we classified responses as NT when (i) two eventualities28 were mentioned without any indication of temporal succession, as in (11); (ii) two  To avoid terminological controversies, we employ here the neutral term ‘eventuality’ (after Bach, 1981; for degrees of eventhood see also e.g. Kamp & Reyle, 1993). See also section “And and the

28

Pragmatics and Grammar as Sources of Temporal Ordering

71

eventualities were understood as simultaneous, as exemplified in (12); (iii) only one eventuality was referred to, as in (13); (iv) no eventuality was mentioned, as in (14); and finally when (v) the second eventuality preceded the first one in the relevant response, as in (15). (11) Luke had a nasty head injury with several painful stiches. At A&E the nurse warned his mother that he should avoid eating and swimming in order to let the stiches heal fully. (CS3, Q1) (12) Luke was at the beach with friends and family. Luke was wading through the water holding a hot dog. He was enjoying the beautiful beach, the sunshine and the sea. However, when he took another bite out of his half eaten hot dog he realised it was wet and salty. He realised he should not eat and swim. (CS3, Q1) (13) John’s friends are talking about him. John never usually gets haircuts, so they are surprised. (CS1, Q2) (14) A wife is describing her husband’s activities that day to a friend on the phone. (CS1, Q1) (15) John has gone into town to get some jobs he’s needed doing done; his haircut and some shopping. (CS1, Q1) Finally, we regarded responses that considered two eventualities as a single one as non-classifiable (N/C) responses. (16) is an example of an N/C response in that ‘having a holiday’ is identified with ‘moving to a different office’. (16) The detective has had a really hard case and is feeling shaken. It doesn’t help that they didn’t catch the killer. She wants to get out of the town, so they relocate her. (CS4, Q1) Table 2 in the Appendix puts together the numbers and percentages of participants’ T, NT and N/C responses in Q1-Q3 for each CS. The patterns displayed by each of CS1-CS5 can be more clearly discerned from Fig. 1 below. We shall now discuss speakers’ responses for each CS in turn, commenting on the different degree to which they trigger temporal interpretations in biased and non-biased contexts. CS1 is an example of and coordination with two simple sentences in the Present Perfect, both expressing events. As such, it is directly analogous to the example (5b) L&S used to support their explanation based on grammatical conventions. While (5), repeated below, conveys temporal anaphora, the past-tense-free equivalent, they say, does not. (5) Oil prices doubled and demand for consumer goods plunged. (5b) Oil prices have doubled and demand for consumer goods has plunged. Our CS1 differs from it in two obvious aspects: it retains the same referent in the subject position and it does not, arguably, convey even a hint of causation in a context-­free scenario. The latter should make it, if at all, even less susceptible to any temporal-anaphoric interpretations. However, our three groups of 30 respondents

Degree of Eventhood”.

72

K. M. Jaszczolt and R. B. Sileo

100 90 80 70 60 50 40 30 20 10 0 Q1

Q2 CS1

Q3

Q1

Q2

Q3

Q1

Q2

CS2

Q3

CS3 T

NT

Q1

Q2 CS4

Q3

Q1

Q2 CS5

Q3

N/C

Fig. 1  Percentages of T, NT and N/C responses for CS1–CS5 in Q1–Q3

were quite ready to assign such a temporal-anaphoric reading: in a context-free condition of Q1, 43.3% of respondents supplied a scenario that pertained to a temporal-­anaphoric interpretation, 76.7% were ready to endorse a temporally-­ biased context of Q2 as a temporally-anaphoric scenario, and, most interestingly, in a context without an overt bias, the temporal reading of and represented 90% of answers, clearly triggered by an automatically arising, typical or otherwise salient filling in of the given scenario. Given that this type of construction was invoked by L&S as the sole, salient example in favour of isolating the past-tense forms in order to give them a separate, grammatical-convention-based explanation, we already have enough evidence to support our argument for a pragmatic, concept-based theory of temporal and. In other words, CS1 suffices for our metafalsification task. However, we were also independently interested in the behaviour of other non-past-tense constructions, so we throw in four more examples. CS2 contains two coordinated gerundial forms, both pertaining to resultative processes (telic situations represented through progressive aspect).29 Here 46.7% of respondents thought of a temporally anaphoric interpretation out of context: an interpretation that first came to their minds due to the scenario they thought of first.30 A temporally biased context, involving the causal  They are embedded under a past-tense verb form pertaining to the report on the situation and as such not contributing to the potential temporality of and. CS4 is an analogous case save with bare infinitives. 30  For the purpose at hand it was not necessary to time the responses: we were only interested in the quality of the situations that these examples triggered. 29

Pragmatics and Grammar as Sources of Temporal Ordering

73

link between selling one item and using that money to buy the other, predictably increased the temporally anaphoric interpretations to 76.7%. A context of Q3 on which a non-temporally-anaphoric scenario seems likely resulted in 20% of temporally anaphoric interpretations. CS2 thus testifies to the power of contextual manipulation, as well as the power of salient scenarios. At this juncture it becomes apparent why we found it prudent to opt for the method of metafalsification and as such for case studies. Context manipulation is one thing, and supplying a default or otherwise salient scenario is quite another. Adding the vagaries of the potentially relevant aspect of context to the vagaries of the potentially relevant aspects of scenarios, not to mention the strength of the salience of certain scenarios caused for example by presumed causation, renders any quantitative experimental studies dubious indeed. CS3 is a case of gerundial and coordination of two activities (non-resultative processes) but this time without any hint of the past-time reference higher up in the structure: it is an example of deontic modality and as such is future-oriented. Context-free Q1 triggered temporally anaphoric scenarios in 46.7% of respondents, temporally biased Q2 in all 30 (100%), and non-temporally biased Q3 in 26.7% of our consultants. The unanimous convergence on the temporal conjunction in Q2 is predictable in that advice to avoid swimming is causally linked to the contextual clue that Luke is ‘very full’. At this point we need a disclaimer that, in agreement with our overall judgement that all potentially pertinent pragmatic variables cannot be controlled for in this study, we made no attempt to even try to keep the contextual bias constant across the case studies; rather, when a potential context created a possibility of a salient, standard, anaphoric interpretation, it was selected for Q2. As such, in Q2, we, so to speak, were after ‘reaping what we sowed’: we guided the readers to an interpretation, checking how successful this manipulation in fact was. Next, in CS4, and conjoins two bare infinitives in a statement of a deontic modal report pertaining to eventualities of mixed aspectual classes: a process that can, arguably, be interpreted either as resultative or non-resultative, and an event. 33.3% of responses point to scenarios with temporal anaphora in Q1, only 16.7% in the attempted contextual modification aiming at a sequential interpretation, and 26.7% in the non-biased context of Q3. We hypothesise that choosing a work-related situation of stress-avoidance may have triggered a strong scenario on which both eventualities were seen as collective rewards that the subject had been granted. We return to this example in more detail in section “And and the Degree of Eventhood”. Finally, CS5 triggered an overwhelming convergence on one unique interpretation across the three conditions in Q1–Q3 due to the fact that the eventualities were uniformly taken as a case of situation embedding. We have here two and-conjoined bare infinitives embedded under a deontic modal (and as such, as in CS3, two future-­ oriented eventualities, on this occasion events). A strongly salient, familiar scenario of writing essays while being in the library triggered what we classified as a uniformly temporal-anaphoric interpretation, produced without a hint of past-time reference: the subject must go to the library and then/in order to write an essay there.

74

K. M. Jaszczolt and R. B. Sileo

Again, we return to this example while discussing degrees of eventhood in section “And and the Degree of Eventhood”. The responses show, first, that the temporal interpretation of and arises easily in past-tense-marker-free constructions in context-free conditions. It can also be manipulated by various factors pertaining to the provided contexts that then trigger salient scenarios, various aspects of which further contribute to the strength of this sense of the succession of the eventualities. A strongly biasing standard scenario can shift all the votes to temporal anaphora as in CS5, or substantially lower such an interpretation as in CS4. The results are also consistent with our initial methodological assumption that a great number of variables can affect the interpretation of and coordination. They are therefore consistent with our metatheoretic proposal that a pragmatic account that would give this multitude of partial triggers of temporal anaphora their due place is preferable to an account in which grammatical tense selects a sub-set of cases of this general phenomenon and dictates a specific explanation as a sole identified trigger. It might be argued that coordination of past-tense forms should also be put to the test through case studies. However, as our handful of cases amply exemplifies, it is the scenario that mostly affects the interpretation. L&S’ example (5) is not immune to inference from such contextual assumptions and as such is just one more potential case study where a great number of potential triggers are at play – arguably, most of all, the powerful scenario of an economic slow-down that the sentence describes.31 In the absence of stronger arguments from grammar, pragmatics reigns.

And and the Degree of Eventhood In spite of, and in some ways thanks to, the substantial cross-case variation, the findings from the study exceeded our expectations. Not only did we manage to (i) assess the strength of the temporal reading of sentential and that cannot be attributed to the (absent) Simple Past tense and (ii) assess the role of biasing contexts on the meaning of and, but the results allowed us also to (iii) glean some factors that potentially affect the strength of the temporal reading. This, by necessity somewhat speculative, section presents our preliminary thoughts on this question. First, the overwhelming convergence on one unique interpretation of CS5 finds a simple explanation when we consider ‘eventhood’. While on the surface the sentence has the pattern e1 and e2 where e1 and e2 stand for events, and it can potentially be interpreted as such (that is, that Jill must do two things: go to the library (to do some reading, borrow books, etc.) and write an essay (on the basis of this reading, using the borrowed books, or as an independent task, say, on the basis of some other lecture notes etc.), in practice our consultants chose the interpretation triggered by the default scenario whereby students write their essays in the library. Once the ‘one  Among others, agency is one obvious trigger. While in CS1 it is, so to speak, ‘undiluted’, CS2 imposes a perspective of the observer (‘They saw…’), and CS3–CS5 add a deontic frame.

31

Pragmatics and Grammar as Sources of Temporal Ordering

75

location’ scenario is activated, the ‘one scenario’ interpretation follows, on which the events are not only temporally but also causally related: Jill must go to the library in order to write an essay there. The main communicated message is likely to be even simpler: Jill must write an essay in the library (so, she must go there). What appeared to be a coordinated structure e1 and e2 now seems to carry a message e1 in order to e2, or even, arguably, be about one multi-part event e. So, eventhood explains the result. The role of eventhood as an explanans does not end there. Let us now look at CS1–CS3. In CS1, we have two coordinated events: John’s doing shopping and having a haircut. There is no default temporal or causal relation between them: one does not have to complete one’s shopping before going to the hairdresser’s (except on a very specific scenario where one ranks shopping as more urgent or otherwise preferred), neither does one have to be well coiffured in order to be served in shops. Temporal biasing achieved in Q2 and Q3 alike is somewhat of a puzzle but perhaps any attempt, by any other than very explicitly leading context, to destabilise the standard (balanced) interpretations in Q1 achieves the Narration result triggered by the strong eventhood of both e1 and e2. This, however, remains at present in the realm of informed speculation. Eventhood is a degree concept. As Kamp and Reyle point out, “the division between event-describing and state-describing sentences appears to be gradual rather than sharp”. Events entail change: an obtaining condition (a state) becomes terminated. At the same time, Kamp and Reyle note that events and states are conceptual constructs: “…a condition is something conceptual, something that has as much to do with the way in which we choose to see reality as with reality itself”. In other words, the situation can often be interpreted as a state or as an event, depending on the focus of attention, its role in the discussed string of eventualities, illocutionary purpose, and so forth. For the purpose of our gradation approach, it will suffice to assume a broad conceptual rule that “[e]vents involve some kind of change” (Kamp & Reyle, 1993: 507). Next, in CS2, we also have two events but it takes less imagination to see them as causally related: perhaps Patricia had to obtain money from the sale of her laptop in order to buy a guitar. It seems plausible to speculate that the overwhelming success in achieving the non-temporal interpretation in the non-temporally biased scenario of Q3 can be attributed to this fact: while Q1 and Q2 did not reveal differences in behaviour between CS1 and CS2, Q3 directed the subjects to an interpretation that was unexpected. CS3 exemplifies a coordination of two eventualities. These are processes that, on the gradation view adopted here, could also be called ‘weak’, atelic events.32 But the structure is not necessarily that of e1 and e2. We have here an ambiguity between sentential and (Luke should avoid eating and he should avoid swimming too) and coordination within the verbal category (Luke should avoid eating and swimming,

 The literature on delimiting events and on the classification of aspectual properties is vast. We will not enter here into terminological debates, neither will we rehearse the debates on the nature of events. For a pertinent introduction see e. g. Pianesi & Varzi (2000). See also fn 28.

32

76

K. M. Jaszczolt and R. B. Sileo

meaning eating ‘followed by’ or alternatively ‘at the same time as’ swimming, say, chewing food while in water). So, three basic interpretations are possible. Processes (‘weak’, atelic events) behave differently from (telic) events in that the lack of telos carries with it a weaker expectation of a sequence. However, they still easily enter into coordination relations – unlike pure states, to wit (17). (17)

?

Luke jumped into the pool and it was shallow.

But in CS3 we have coordination within deontic modality: Luke should not do two things, in whatever relation (if any) to each other. The overwhelming success in biasing the interpretation to the temporal one in Q2, and the overwhelming majority of the non-temporal readings in Q3 can clearly be attributed to biasing contexts. CS4 is an interesting example when assessed in the light of aspectual properties. While having a holiday is likely to be conceptualised as a process in this sentence, or perhaps an event with some weak telic meaning (as, say, taking action to feel rested and recharged) and moving to a new office as an event, the focus on deontic modality (‘they let the detective e1 and e2’) appears to weaken the degree of eventhood of both activities: they let him be on holiday and they let him be in a new office. Or, alternatively, it even reverses the degree of eventhood: they let him go on holiday (an event with a clear telos) and they let him work in a new office (a process). Moreover, some participants even interpreted the sentence to mean that the move to the new office constituted a holiday, so they conflated the two eventualities.33 This interpretation was counted as non-classifiable in that temporal ordering (unlike in the case of CS5) cannot be predicated of it. The degree of eventhood, thus, again, appears to be a plausible explanans for the different behaviour of CS4 in Q1 as compared with CS1-CS3. It is also likely to be the reason for the entrenchment with respect to the attempted temporal interpretation in Q2 that keeps it at odds with CS1–CS3. We are fully aware that the interpretations of the results by attempting to correlate them with the degree of eventhood are still at the early stages of hypothesising, and they have only emerged post hoc from our handful of case studies that were not designed to test for eventhood in the first place. But the case studies show that there is a research question here ready to pursue. To recall, the questionnaires were designed with the aim to test different types of structures in the context of the grammar/pragmatics dispute. So, any indications of correlations count as a bonus: the study of the association with eventhood is now ready for a more thorough empirical scrutiny.34  Four participants in Q1, one in Q2, and two in Q3 (2.7%, 0.7%, and 1.3% of responses respectively) conflated the two eventualities. 34  One possible design would be as follows. One could start with (i) clear ‘L&S cases’: e1 and e2, both verbs in Simple Past, high degree of eventhood, clear temporal reading in a default scenario, and proceed to (ii) e1 and e2, both verbs in Simple Past, high degree of eventhood, ambiguous 33

Pragmatics and Grammar as Sources of Temporal Ordering

77

Conclusions and Future Prospects To conclude, it is clear from our theoretical discussion of the interrelations between the proposed theories of sentential and coordination, as well as from our case studies, that the proposal of attributing the temporal reading to grammatical tense does not capture the phenomenon in its entirety. Neither does it capture it on a sufficiently high level of generalization. Sentences without, or with very diversified, indicators of temporality, also give rise to temporal succession interpretation in unbiased scenarios in a fair number of cases and this fact ought to be thrown into any adequate explanation. And so should the contextual trigger of such readings. To repeat, our aim was not to present counterevidence to L&S grammar-based account; our aim was merely to point out its narrow coverage. As we explained earlier, what we haven’t done here is the apparently most obvious study one would conduct in order to test L&S hypothesis: we have not tested examples with Simple Past verb forms with different degree of stipulated context-free temporal reading. This is so because, first, in our view there are no truly context-free interpretations: the addressee will always attempt to supply a scenario, be it a default, common one or one that can be traced to however scarce background factors. Second (and relatedly), L&S have a clear response for such examples that comes from rules governing the structure of information in discourse. Neither did we attempt to compare the behaviour of tensed and untensed verb forms. Such a comparison would not have equipped us with any useful arguments in that, as above, the distribution of responses can be attributed to other than grammatical factors (although it has a role to play in the study of degrees of eventhood proposed). What we have done is put L&S’s thesis in a broader perspective of (i) other, including higher-level, potential explanations of the behaviour of sentential and, as well as (ii) other types of constructions, where the reading of and cannot be attributed to the Simple Past verb forms. Both achieve the result of weakening the power of L&S’ grammar-based, narrowly focused generalization to the point where its explanatory power is of little use, and its predictive power, considering the myriad potential triggers of temporal anaphora, arguably negligible.35

temporal/atemporal reading in a default scenario; then (iii) e1 and e2, both verbs in Simple Past, high degree of eventhood, no temporal reading in a default scenario. Next, one can manipulate the degree of eventhood in each of the three categories. 35  Research for this paper was partly supported by the Cambridge Humanities Research Grant Rethinking Being Gricean: New Challenges for Metapragmatics https://www.mml.cam.ac.uk/ rethinking-being-gricean.

78

K. M. Jaszczolt and R. B. Sileo

Appendix Table 1  Tested scenarios in Q1–Q3 Q1 CS1 John’s done some shopping and he’s had a haircut. CS2 They saw Patricia selling a laptop and buying a guitar.

CS3 Luke should avoid eating and swimming.

CS4 They let the detective have a holiday and move to a new office.

CS5 Jill must go to the library and write an essay.

Q2 [John likes to do his shopping first thing in the morning.] John’s done some shopping and he’s had a haircut.

Q3 [John likes to go to the shopping centre. He enjoys the huge variety of shops there.] John’s done some shopping and he’s had a haircut. [Patricia got £200 when she sold her [Patricia met Mel, who needed to buy a laptop from her. Patricia laptop to Mel. Patricia was happy she had some money to spend on a knew that Mel was selling a guitar.] guitar.] They saw Patricia selling a They saw Patricia selling a laptop laptop and buying a guitar. and buying a guitar. [Luke has been swimming for [Luke has had a big lunch. He is 5 hours. He is hungry but he very full.] needs his hands free to swim.] Luke should avoid eating and Luke should avoid eating and swimming. swimming. [The detective was resting on the [The detective was stressed and beach. He avoided all the could not work anymore. He was changes at the office.] also unhappy about his working They let the detective have a conditions.] They let the detective have a holiday holiday and move to a new office. and move to a new office. [Jill gets distracted when she [Jill has not done any research for her Psychology assignment yet. She tries to do her homework at home. She never gets anything is really behind with her reading.] Jill must go to the library and write done there.] Jill must go to the library and an essay. write an essay.

Pragmatics and Grammar as Sources of Temporal Ordering

79

Table 2  Numbers and percentages of T, NT and N/C responses for CS1–CS5 in Q1–Q3 Responses Q1 T NT Total T & NTa N/C Total Q2 T NT Total T & NT N/C Total Q3 T NT Total T & NTa N/C Total

CS1

%

CS2

%

CS3

%

CS4

%

CS5

%

13 17 30 0 30

43.3 56.7 100.0 0.0 100.0

14 16 30 0 30

46.7 53.3 100.0 0.0 100.0

14 16 30 0 30

46.7 53.3 100.0 0.0 100.0

10 16 26 4 30

33.3 53.3 86.7 13.3 100.0

30 0 30 0 30

100.0 0.0 100.0 0.0 100.0

23 7 30 0 30

76.7 23.3 100.0 0.0 100.0

23 7 30 0 30

76.7 23.3 100.0 0.0 100.0

30 0 30 0 30

100.0 0.0 100.0 0.0 100.0

5 24 29 1 30

16.7 80.0 96.7 3.3 100.0

30 0 30 0 30

100.0 0.0 100.0 0.0 100.0

27 3 30 0 30

90.0 10.0 100.0 0.0 100.0

6 24 30 0 30

20.0 80.0 100.0 0.0 100.0

8 22 30 0 30

26.7 73.3 100.0 0.0 100.0

8 20 28 2 30

26.7 66.7 93.3 6.7 100.0

30 0 30 0 30

100.0 0.0 100.0 0.0 100.0

Percentages corresponding to T and NT responses do not always add up to Total T & NT percentages due to rounding

a

References Asher, N., & Lascarides, A. (2003). Logics of conversation. Cambridge, UK: Cambridge University Press. Atlas, J.  D. (1977). Negation, ambiguity, and presupposition. Linguistics and Philosophy, 1, 321–336. https://doi.org/10.1007/BF00353452. Atlas, J. D. (1979). How linguistics matters to philosophy: Presupposition, truth, and meaning. In C.-K. Oh & D. A. Dinneen (Eds.), Syntax and semantics (Vol. 11, pp. 265–281). New York, NY: Academic Press. Atlas, J.  D. (1989). Philosophy without ambiguity: A logico-linguistic essay. Oxford, UK: Clarendon Press. Bach, E. (1981). On time, tense, and aspect: An essay in English metaphysics. In P. Cole (Ed.), Radical pragmatics (pp. 63–81). New York, NY: Academic Press. Bar-Lev, Z., & Palacas, A. (1980). Semantic command over pragmatic priority. Lingua, 51, 137–146. https://doi.org/10.1016/0024-3841(80)90004-2. Caffi, C. (2006). Metapragmatics. In K.  Brown (Ed.), Encyclopedia of language and linguistics (2nd ed., pp.  82–88). Oxford, UK: Elsevier. First published in 1994 in: Asher, R.  E. (Ed.),  Encyclopedia of language and linguistics (pp. 2461–2466). Oxford, UK: Pergamon Press. Caffi, C. (2016). Revisiting metapragmatics: “What are we talking about?”. In K. Allan, A. Capone, & I.  Kecskes (Eds.), Pragmemes and theories of language use (pp.  799–821). Dordrecht, Netherlands: Springer. Carston, R. (1988). Implicature, explicature, and truth-theoretic semantics. In R.  M. Kempson (Ed.), Mental representations: The interface between language and reality (pp.  155–181). Cambridge, UK: Cambridge University Press.

80

K. M. Jaszczolt and R. B. Sileo

Carston, R. (1993). Conjunction, explanation and relevance. Lingua, 90, 27–48. https://doi. org/10.1016/0024-3841(93)90059-6. Carston, R. (1994). Conjunction and pragmatic effects. In R. E. Asher (Ed.), The encyclopedia of language and linguistics (pp. 692–698). Oxford, UK: Pergamon Press. Carston, R. (2002). Thoughts and utterances: The pragmatics of explicit communication. Oxford, UK: Blackwell. Chierchia, G. (2004). Scalar implicatures, polarity phenomena, and the syntax/pragmatics interface. In A. Belletti (Ed.), Structures and beyond: The cartography of syntactic structures (Vol. 3, pp. 39–103). Oxford, UK: Oxford University Press. Cohen, L. J. (1971). Some remarks on Grice’s views about the logical particles of natural language. In Y. Bar-Hillel (Ed.), Pragmatics of natural languages (pp. 50–68). Dordrecht, Netherlands: D. Reidel. Gómez Txurruka, I. (2003). The natural language conjunction and. Linguistics and Philosophy, 26, 255–285. https://doi.org/10.1023/A:1024117423963. Grice, H. P. (1975). Logic and conversation. In P. Cole & J. L. Morgan (Eds.), Syntax and semantics (Vol. 3). New York, NY: Academic Press. Reprinted in: Grice H. P. (1989) Studies in the way of words (pp. 22–40). Cambridge, MA: Harvard University Press. Grice, H. P. (1978). Further notes on logic and conversation. In P. Cole (Ed.), Syntax and semantics (Vol. 9). New York: Academic. Reprinted in: Grice H. P. (1989) Studies in the way of words (pp. 41–57). Cambridge, MA: Harvard University Press. Grosz, B. J., Joshi, A. K., & Weinstein, S. (1995). Centering: A framework for modelling the local coherence of discourse. Computational Linguistics, 21, 203–225. Haugh, M. (2018). Corpus-based metapragmatics. In A. H. Jucker, K. P. Schneider, & W. Bublitz (Eds.), Methods in pragmatics (pp. 619–643). Berlin, Germany: Mouton de Gruyter. Hobbs, J. R. (1990). Literature and cognition. Stanford, CA: CSLI. Horn, L. R. (1984). Toward a new taxonomy for pragmatic inference: Q-based and R-based implicature. In D. Schiffrin (Ed.), Georgetown University round table on languages and linguistics 1984 (pp. 11–42). Washington, DC: Georgetown University Press. Horn, L. R. (1988). Pragmatic theory. In F. J. Newmeyer (Ed.), Linguistics: The Cambridge survey (Vol. 1, pp. 113–145). Cambridge, UK: Cambridge University Press. Horn, L.  R. (2016). Conventional wisdom reconsidered. Inquiry, 59, 145–162. https://doi. org/10.1080/0020174X.2015.1125133. Horn, L. R. (2018). First things first: The pragmatics of “natural order”. Presentation delivered at Pragmasophia 2, Lisbon, 19 September 2018. Jaszczolt, K. M. (2005). Default semantics: Foundations of a compositional theory of acts of communication. Oxford, UK: Oxford University Press. Jaszczolt, K. M. (2010). Default semantics. In B. Heine & H. Narrog (Eds.), The Oxford handbook of linguistic analysis (pp. 215–246). Oxford, UK: Oxford University Press. Jaszczolt, K.  M. (2012). Semantics/pragmatics boundary disputes. In C.  Maienborn, K. von Heusinger, & P.  Portner (Eds.), Semantics: An international handbook of natural language meaning (Vol. 3, pp. 2333–2360). Berlin, Germany: Mouton de Gruyter. Jaszczolt, K.  M. (2018). Pragmatics and philosophy: In search of a paradigm. Intercultural Pragmatics,15, 131–159. https://doi.org/10.1515/ip-2018-0002. Jaszczolt, K. M. (2019). Rethinking being Gricean: New challenges for metapragmatics. Journal of Pragmatics, 145, 15–24. https://doi.org/10.1016/j.pragma.2019.01.024. Kamp, H., & Reyle, U. (1993). From discourse to logic: Introduction to modeltheoretic semantics of natural language, formal logic and discourse representation theory. Dordrecht, Netherlands: Kluwer. Kehler, A. (2002). Coherence, reference, and the theory of grammar. Stanford, CA: CSLI. Kempson, R.  M. (1975). Presupposition and the delimitation of semantics. Cambridge, UK: Cambridge University Press. Kempson, R. M. (1979). Presupposition, opacity, and ambiguity. In C.-K. Oh & D. A. Dinneen (Eds.), Syntax and semantics (Vol. 11, pp. 283–297). New York, NY: Academic Press.

Pragmatics and Grammar as Sources of Temporal Ordering

81

Kempson, R. M. (1986). Ambiguity and the semantics-pragmatics distinction. In C. Travis (Ed.), Meaning and interpretation (pp. 77–103). Oxford, UK: B. Blackwell. Lascarides, A., & Asher, N. (1993). Temporal interpretation, discourse relations and commonsense entailment. Linguistics and Philosophy, 16, 437–493. https://doi.org/10.1007/BF00986208. Leezenberg, M. (2006). Gricean and Confucian pragmatics: A contrastive analysis. Journal of Foreign Languages, 6, 2–21. Lepore, E., & Stone, M. (2015). Imagination and convention: Distinguishing grammar and inference in language. Oxford, UK: Oxford University Press. Levinson, S.  C. (1987). Minimization and conversational inference. In J.  Verschueren & M.  Bertuccelli-Papi (Eds.), The pragmatic perspective. Selected papers from the 1985 International Pragmatics Conference (pp. 61–129). Amsterdam, Netherlands: John Benjamins. Levinson, S. C. (2000). Presumptive meanings: The theory of generalized conversational implicature. Cambridge, MA: MIT Press. Partee, B. H. (1973). Some structural analogies between tenses and pronouns in English. Journal of Philosophy, 70, 601–609. Reprinted in: Partee, B.  H. (2004). Compositionality in formal semantics: Selected papers by Barbara H. Partee (pp. 50–58). Oxford, UK: Blackwell. Pianesi, F., & Varzi, A. C. (2000). Events and event talk: An introduction. In J. Higginbotham, F.  Pianesi, & A.  C. Varzi (Eds.), Speaking of events (pp.  3–47). Oxford, UK: Oxford University Press. Recanati, F. (1989). The pragmatics of what is said. Mind and Language 4(4): 295–329. Reprinted in:  S.  Davis (Ed.), (1991) Pragmatics: A reader (pp.  97–120). Oxford, UK: Oxford University Press. Recanati, F. (2010). Truth-conditional pragmatics. Oxford: Clarendon Press. Sperber, D., & Wilson, D. (1995). Relevance: Communication and cognition (2nd ed.). Oxford, UK: Blackwell. First published in 1986. Stanley, J., & Szabó, Z.  G. (2000). On quantifier domain restriction. Mind and Language, 15, 219–261. https://doi.org/10.1111/1468-0017.00130. Stojnić, U., Stone, M., & Lepore, E. (2017). Discourse and logical form: Pronouns, attention and coherence. Linguistics and Philosophy, 40, 519–547. https://doi.org/10.1007/ s10988-017-9207-x.

Inferential Patterns of Emotive Meaning Fabrizio Macagno and Maria Grazia Rossi

Introduction […] people often do not consider the entire meaning of words. That is, words often signify more than they appear to, and when people try to explain their meaning, they do not represent the entire impression made in the mind. This is so because for an uttered or written sound to signify is nothing other than to prompt an idea connected to this sound in the mind by striking our ears or eyes. Now, frequently, in addition to the main idea which is considered its proper meaning, a word may prompt several other ideas – which may be called incidental ideas – without our realizing it, although the mind receives their impressions. For example, if we say to someone, “You lied about it,” and we consider only the principal meaning of this expression, this is the same as saying: “You know that the contrary of what you say is true.” But in common use these words carry an additional idea of contempt and outrage. They make us think that the person who says them does not care whether they injure us, and this makes the words insulting and offensive. (Arnauld & Nicole, Logic, I, 14).

This passage from the Logic of Port Royal addresses a very actual philosophical problem, which has crucial implications not only on how language is conceived, but also more importantly on how language is used and modified strategically. This ancient account of “incidental ideas” and the related issues of derogatory words and redefinitions become strikingly modern when we look at the 2016 Presidential elections. By rejecting the “politically correct” image of a presidential candidate, Trump adopted a type of rhetoric characterized by the frequent use of

F. Macagno (*) · M. G. Rossi Instituto de Filosofia da Nova (IFILNOVA), Facultade de Ciências Sociais e Humanas, Universidade Nova de Lisboa, Lisbon, Portugal © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 F. Macagno, A. Capone (eds.), Inquiries in Philosophical Pragmatics, Perspectives in Pragmatics, Philosophy & Psychology 28, https://doi.org/10.1007/978-3-030-56696-8_5

83

84

F. Macagno and M. G. Rossi

what has been reported by the press as “charged expressions” or “loaded language,”1 which in the philosophical literature have been commonly referred to as “ethical” or emotive words (Stevenson, 1937, 1944). Words such as “losers,2” “bimbo,3” “stupid,” “scum,” “fools” are only the most stereotypical examples of terms that are not simply used to describe reality by modifying the cognitive response of the interlocutor (i.e., “informing” him), but more importantly to affect the interlocutor’s attitudes towards a state of affairs and suggest a course of action. Ethical (or emotive) words are a powerful rhetorical instrument used for “framing” an issue (Druckman, 2002; Entman, 1993; Sniderman & Theriault, 2004) and influencing the interlocutors’ decisions. However, their nature and their effects constitute also crucial philosophical problems, involving philosophical positions, such as nominalism (Richard, 2008), contextualism, externalism, or inferentialism (Hom, 2010), some of which have been debated over centuries. Such issues are recently attracting increasing attention from both a philosophical and linguistic perspective, even though most of the studies have focused on a specific type of ethical words, slurs (Blakemore, 2015; Croom, 2011, 2014; Hom, 2008, 2010; Richard, 2008). This paper investigates a specific dimension of loaded language, namely the defeasibility (or rather cancellability) of the expressive force it triggers. Many studies have advanced various theories, often conflicting with each other, on how the expressive force is or can be originated – whether externally through social conventions (Anderson & Lepore, 2013), semantically from semantic content (Hom, 2008, 2010) or presupposition triggers (Schlenker, 2007, p. 137), or inferentially as part of the broader meaning (rules of use) of a word (Whiting, 2013; Williamson, 2009). Such works, however, tend to dismiss the other dimension of the semantic/pragmatic interface, namely how such an expressive force is triggered, developed, cancelled, or modified in different contexts. Our goal is to focus on the cancellable nature of the “expressive force” of emotive words (Capone, 2014, pp. 303–304) and investigate it as a defeasible inference used for performing presumptively a specific speech act.

1  See for instance: Eli Stokols (2016). Trump’s loaded words fuel campaign freefall. Politico, 8 September 16. Retrieved from: http://www.politico.com/story/2016/08/donald-trump-campaignstatements-226840 (accessed on 01 September 2020); John Wildermuth (2016). Trump’s loaded language a reminder of words’ power. San Francisco Chronicle, 9 August 2016. Retrieved from: http://www.sfchronicle.com/politics/article/Trump-s-loaded-language-a-reminder-ofwords-9133067.php (accessed on 01 September 2020); A Surprising Thing Happens When Presidential Candidates Use Emotional Language. The Huffington Post, 21 March 2016. Retrieved from: http://www.huffingtonpost.com/entry/donald-trump-emotional-inflammatory-language_ us_56e84b60e4b0860f99da8d3e?utm_hp_ref=science (accessed on 01 September 2020). 2  Trump Lashes Out At McCain: ‘I Like People Who Weren’t Captured’. NPR, 8 July 2015. Retrieved from: http://www.npr.org/sections/thetwo-way/2015/07/18/424169549/trump-lashesout-at-mccain-i-like-people-who-werent-captured (accessed on 01 September 2020). 3  Donald Trump retweets Megyn Kelly ‘bimbo’ jab as women react to insults. The Guardian, 7 August 2015. Retrieved from: https://www.theguardian.com/us-news/2015/aug/07/donald-trumpmegyn-kelly-bimbo-women-republican-debate (accessed on 01 September 2020).

Inferential Patterns of Emotive Meaning

85

Characteristics of Emotive Meaning The notion of “incidental ideas” mentioned in the aforementioned excerpt from Port Royal Logic is a vague metaphor for a dimension of meaning that has been investigated in the recent years under the label of “expressive meaning” (Croom, 2014; Hedger, 2012, p.  76), “expressive force,” or “derogatory content” or “derogatory properties” of slurs and pejoratives (Hom, 2008, 2010), and was analyzed in the emotivistic tradition as “emotive meaning” (Stevenson, 1937). In order to address the relationship between ethical terms (and slurs in particular) and emotive meaning, it is useful to focus on the defeasible nature of the latter, its complex structure, and how it can be represented.

The Defeasibility of Emotive Meaning In the literature, the essential characteristic of slurs and emotive words in general has been acknowledged to be their expressive force (Croom, 2014, p. 230; Hom, 2008, 2010, p. 168), namely their potential to express the speaker’s (negative) psychological attitudes towards the state of affairs referred to (the target) (see also Bianchi, 2014b). This “force” is in itself a complex notion, as it is a characteristic of specific words that explains a stable use of the utterances in which they occur. The most important features of expressive force can be summarized as follows (Hom, 2010, pp. 164–170): 1. Force variability. The derogatory force of distinct pejoratives varies in strength, type and intensity of the emotion expressed. 2. Historical variability. The force of pejorative terms varies over time, and is sensitive to the relevant social facts. 3. Content-dichotomy puzzle. Pejoratives can be used in an orthodox and non-­ orthodox way. Orthodox occurrences are characterized by the following properties, which are contrary to the non-orthodox ones: (a) Non-displaceable. They appear to derogate even when embedded in different contexts such as in the antecedent of a conditional, under negation, inside of a question, in an attitude report, etc. (b) Agent centered. They appear to be inseparable from the attitudes of their speaker. (c) No-truth conditional contribution. Slurring uses of derogatory words do not have truth-conditional content, whereas non-slurring uses do have truth-­ conditional content. These three aspects define various dimension of cancellability (defeasibility) of the expressive force. The expressive force varies according not only to the different emotive words, but also to the broader conversational setting, defined in terms of the culture of the members of a speech community (Kecskes, 2013,

86

F. Macagno and M. G. Rossi

p. 141), and the presumable intention of the speaker (orthodox vs. non-orthodox uses). In particular, the historical variability captures the strict interdependence between emotive words and the culture of the relevant speech community, which can vary in time depending on the shared evaluations of the states of affairs or entities referred to (Capone, 2016, p.  177). Moreover, the expressive force is absent when emotive words (and in particular pejoratives) are used for specific non-­derogatory purposes, which include appropriated uses (the use of slurs by the targeted members or groups for non-derogatory purposes) (Bianchi, 2014a) and “meta-linguistic” uses, namely when the speaker is using such epithets for “objecting to discriminatory discourse,” reporting the racist contents carried by epithets (Bianchi, 2014b, p. 477). Such characteristics seem to indicate that (in at least several cases, which we will illustrate below) the “expressive force” can be regarded as a type of intended and automatic inference, which is drawn in lack of contrary evidence and until contrary evidence is provided (Capone, 2011; Jaszczolt, 2005, p.  46). Unless the speaker expresses his or her strong intention by providing specific clues, the inference usually triggered by emotive words is hardly cancellable (Capone, 2013). Considering the usual association between the concept of conversational implicature and the conscious processing of a contextual inference (Carston, 2002, 2004; Sperber & Wilson, 2002), we will focus only the logical dimension of such inferences, which we will refer to as “intended inferences” to avoid terminological confusions. In the next sections, the problem of their automaticity (Capone, 2010) will be discussed.

The Twofold Nature of Expressive Force As mentioned above, the notion of expressive force is complex, as it is in itself defined by referring to the (prototypical) uses of emotive words, namely to express an evaluation of their targets (to display an attitude) (Hom, 2010, p. 171), thus conveying an emotion towards it (Capone, 2014). In case of pejoratives, this evaluation is negative, and is presumptively used for insulting, degrading, or performing other derogatory acts (Bianchi, 2014b; Capone, 2016). This twofold nature was clearly captured by Stevenson’s emotivistic approach. In his works on ethical utterances, Stevenson (Stevenson, 1937, 1938a, 1938b, 1944) developed an account of ethical words based on a causal account of meaning (Dancy, 2013, p. 736). On this view, the meaning of a term was understood as an unchanged dispositional property (Stevenson, 1944, p. 46), a tendency to produce certain responses that is the result of the habits of the speakers. According to Stevenson, words can be used for two distinct and often interrelated purposes (Richard, 2012, p. 443): descriptively, to “record, clarify, and communicate beliefs,” or dynamically, to “give vent to feelings, create moods, or to incite people to actions or attitudes” (Stevenson, 1937, p.  21). These two tendencies are referred to as “descriptive” and “emotive” meaning.

Inferential Patterns of Emotive Meaning

87

Stevenson noticed how some words (including slurs and ethical terms such as “peace,” “democracy,” or “terrorism”) are connected with (or prototypically used to produce) emotions, or rather action-oriented, immediate responses. Some of them consist in venting emotions, others in expressing an attitude that can affect the interlocutor’s evaluation of the represented state of affairs. As Stevenson (Stevenson, 1937, pp. 18–19), put it “instead of merely describing people’s interests, they change and intensify them. They recommend an interest in an object, rather than state that the interest already exists,” as they have the tendency to lead the hearer towards a decision by affecting his or her system of interests (Stevenson, 1938a, 1938b, 1944, p. 54). Words characterized by an emotive meaning can be used for persuading the interlocutor or the audience to carry out a specific action. On Stevenson’s view, ethical or emotive terms are characterized by the combination of their descriptive and emotive meaning (Stevenson, 1944, p. 206; 210). Since their potentiality consists in expressing an attitude towards a state of affairs, which can direct or affect the interlocutor’s interests towards it, they provide at the same time a representation and an evaluation, which is presupposed by the emotion expressed. For this reason, the speaker can manipulate or redirect the interlocutor’s attitude (i.e. evaluation or interests) towards a state of affairs (in cognitive approaches to emotions, a “target,” see De Sousa, 1987, p. 20) by redefining the “descriptive meaning” or the “emotive meaning” of the ethical word used to refer to it (Stevenson, 1938b, p. 332). In the first case, the ethical term is “persuasively defined,” in the second case, it is “quasi-defined.” The complex structure of emotive words, characterized by a “tendency” to be used for affecting the interlocutors’ attitudes towards a state of affairs and the inherently defeasible nature of the emotive meaning leads to the problem of representing their “expressive force.” In the next sections, an inferentialist account of emotive words will be presented, in which the “tendency” or “potential” characterizing emotive words and the prototypical dynamic (or, in case of pejoratives, derogatory) uses will be explained as forms of argument. More specifically, the implicitly conveyed meaning will be investigated and represented through argumentation schemes or argument patterns, namely the informal reasoning that can be used to account for a specific intended inference (Leech, 1983, p. 30). This approach will be shown to explain the complex nature of the emotive meaning.

Emotive Meaning as Defeasible Inferences Stevenson’s approach to emotive meaning has the clear advantage to account for the defeasible nature of the expressive force and the prototypical association between the uses of emotive words and specific speech acts (Stevenson, 1944, p.  210). However, by investigating emotive meaning in terms of the acts presumptively performed by using emotive words, emotivism fails to explain two crucial aspects thereof, namely 1) the inferences from descriptive uses to dynamic ones (and vice versa) (Geach, 1965, p. 463), and 2) the inferences from unasserted occurrences of

88

F. Macagno and M. G. Rossi

ethical terms (Hom, 2010, p.  171). More precisely, Stevenson’s theory cannot explain the dynamic use of the following conclusion: If doing a thing is bad, getting your little brother to do it is bad. Tormenting the cat is bad. Ergo, getting your little brother to torment the cat is bad. Supposing that the major premise is asserted by the father to his older son, the father’s condemnation of the son’s behavior (dynamic use of “bad”) cannot be inferred from an assertion (descriptive use of “bad”). Similarly, the embedding of an emotive term under negation or in a question (“are they scum?”) triggers inferences that are not related to the dynamic purpose of affecting the interlocutor’s attitudes. Finally, the distinction between the two types of “potentialities” of a word can explain why the emotive meaning can vary in time (habits change) but not how it can be modified on purpose. In order to address these shortcomings, an inferentialist approach will be proposed, in which the relationship between emotive words and slurring effects are mediated by evaluative inferences.

An Argumentative Account of Ethical Words The “descriptive” and “emotive” meaning of ethical words, i.e. the tendency to elicit cognitive or emotive responses, can be interpreted from a logical4 perspective as different types of intended and defeasible inferences guided by different types of presumptions guiding the inferential process. Such inferences are normally described as conversational implicatures, and analyzed according to the set of presumptions codified by Grice in his maxims. However, the mechanism underlying the reconstruction of an intended inference can be represented using more specific inferential patterns, and relying on specific presumptions that can be ordered in levels of acceptability. As Grice underscored (Grice, 1975, p. 50), together with the conversational maxims, presumptions of different type are involved in reconstructing an implicature, such as the conventional meaning of the words used, contextual information, background knowledge, and the shared knowledge. These different types of information can be represented through different levels of presumptions, leading to default (and thus defeasible) conclusions that can be stronger or weaker depending on the weight of the presumption within a specific context and the available evidence rebutting it (Atlas & Levinson, 1981; Bach & Harnish, 1979, Chap. 1; Dascal & Wróblewski, 1988; Macagno, 2012a; Patterson, 2005). The levels of presumptions can be represented in Fig. 1 below (Macagno, 2017, p. 287).

4  The term “logical” shall be considered as referring to the natural language logic investigated in the dialectical tradition and in argumentation theory.

Inferential Patterns of Emotive Meaning

89

Levels of presumptions

0. Pragmatic

Use-Act; Type of dialogue-type of move (ex. Assertive sentences are usually used to inform the hearer; In eristic dialogues interlocutors are expected to vent emotions).

1. Linguistic

Definitions, syntactic structures (ex. ‘Man’ is usually used to mean a ‘rational animal’).

2. Factual, encyclopedic

Facts, events, stereotypes (ex. People usually know that France is not a monarchy now).

3. Values, interests

The interlocutor’s interests/values... (ex. Professor x is usually very critical and writes no recommendation letters; x is usually against the freedom of press...).

Fig. 1  Levels of presumptions

These presumptions can be classified according to their content and ordered and assessed based on their defeasibility conditions (Clark, 1996, Chap. 4; Clark & Brennan, 1991). Presumptions closer to the conversational situation in which the utterance is performed (namely more related to the interlocutors’ specific goals and knowledge) are likely to be less subject to default than more generic presumptions. Thus, specific presumptions resulting from direct evidence concerning the interlocutor’s behavior (specific conversational goals, values, interests, behavior of the interlocutor) are less likely to be subject to default than generic pragmatic (based on form-purpose presumptions), encyclopedic or linguistic information, as less abstract from the specific setting in which the utterance is made. These presumptions can constitute the premises of more specific inferential patterns (Macagno, 2012a; Macagno & Walton, 2013), represented using patterns that are more specific than formal rules of inference – called loci or topoi in the dialectical tradition (Macagno, Walton, & Tindale, 2014) – and described as argumentation schemes (Walton, Reed, & Macagno, 2008). In case of emotive words, the descriptive meaning can be regarded as an inference or a set of inferences attributing a predicate to an entity (a subject) based on specific characteristics and definitional premises (the process of “naming” reality). The emotive meaning can be interpreted as an inference (or as set of inferences) leading from a description of a state of affairs to a value judgment thereon (evaluation), and in some cases to a further inference to the proposal of a commitment to course of action (decision-making inferences) (Leech, 1983, p. 30).

90

F. Macagno and M. G. Rossi

The inferences that can be drawn from the use of an ethical term are thus constrained by a limited number of inferential patterns, leading to a limited number of conclusions. In the following subsection, we will show in detail the inferential structure of the complex inferences drawn from emotive words.

Inferences and Descriptive Meaning Based on the presumptive structure described above, the descriptive meaning can be represented as an inferential passage aimed at classifying a state of affairs. This inference is grounded on classificatory premises, which can be very different in kind. A classification (the attribution of a property to a subject/entity based on some properties characterizing it) can be grounded on definitions, descriptions, or stereotypes that are presumed for distinct reasons to be shared by the interlocutors. While definitional characteristics can be presumed to be hardly defeasible (stronger contextual evidence is needed to defeat them), elements of collocative meaning (such as “non-criterial properties that we have learnt to expect a referent to possess,” including physical, social, typical properties within specific contexts, see Leech, 1981, p. 12) lead to more defeasible conclusions. In both cases, the inference can be represented using the following scheme from classification (Walton et  al., 2008, p. 319). Argumentation scheme 1: Argument from classification Premise 1

If some particular thing a can be classified as falling under verbal category C, then a has property P (in virtue of such a classification). Premise 2 a can be classified as falling under verbal category C. Conclusion a has property P.

The reasonable and correct application of this scheme can be assessed dialogically by taking into account its defeasibility conditions, represented by the following critical questions: CQ1: What evidence is there that a definitely has property C, as opposed to evidence indicating room for doubt about whether it should be so classified? CQ2: Is the classification premise the definition of P? Or is it merely a description thereof? Or is it a stereotype reporting only on a shared and unproven relationship between the two properties? This scheme underlies both the use of an ethical word and the possible inferences that can be drawn from it (Macagno & Walton, 2014, Chap. 3). In particular, the use of an emotive word (or slur) can be justified based on its definition or generally accepted description. An individual can be named as “politically correct” because his or her “avoiding language and practices which could offend political sensibili-

Inferential Patterns of Emotive Meaning

91

ties.” A group of people can be labeled metaphorically as “scum” based on alleged low actions. The classification scheme is one of the grounds of evaluative inferences, namely reasoning steps leading to a specific type of predication, consisting in the attribution of an evaluative predicate (the species of good and evil, see Aristotle, Topics, 123b9) based on specific presumptions (Aristotle, Topics, 116a13-116a21), such as “which is more lasting or secure is more desirable than that which is less so.” These types of evaluative premises are culture-dependent reasons for classifying something as desirable or not (Perelman & Olbrechts-Tyteca, 1951), and – just like other types of classifications – can be grounded on the definitions of evaluative predicates (what is good, desirable, dangerous, etc. see Vendler, 1963; von Wright, 1963c) or stereotypes, and thus be more or less defeasible. For example, “politically correct” can be judged as desirable and good as it is the privation of something evil (offending, which provokes pain). Similarly, “scum” is generally evaluated negatively, as it is commonly described as something ill or worthless. Emotive words rely on prototypical associations between a classification and an evaluation, which lead defaultively to a conclusion that is not based on reasons different from the act of labeling the entity. This type of short-circuited inference skips the justification of the evaluation, which otherwise would have to take into account factors such as the relationship between actions and habits, or the link between consequences (actions) and values. For example, a typical evaluative pattern can be the following argument from consequences to evaluation (adapted from Walton et al., 2008, p. 332). Argumentation scheme 2: Argument from consequences to evaluation. Premise 1 Consequence premise Evaluation premise Conclusion

If agent A brings about (doesn’t bring about) B, then C will occur. C is a good (bad) outcome (from the point of view of A’s goals). That whose production is good is itself also good, and vice versa; that whose destruction is bad is itself also good, and vice versa (Boethius, De Topicis Differentiis, 1190A 7-1190B 1). Therefore, B and A are good (bad).

This scheme presupposes an evaluation of the consequences of actions, and the judgment on the formal or effective cause of the consequence (the action or the agent) is limited to the circumstance (not generalizable to similar actions or to all agent’s actions). Emotive words skip this inferential evaluative mechanism by providing a simpler, less critical evaluation based on stereotypies or commonly accepted and presumptive premises (Walton, 1995). In the extreme case of slurs, the evaluative properties can be explained as part of the properties used for describing the term itself, either as definitional characteristics (such as “bimbo,” distinguished by the feature of “being foolish”) or properties typically and culturally associated with the aspect of the referents considered (such as in case of racial epithets).

92

F. Macagno and M. G. Rossi

The Dynamic Uses of Ethical Words Value judgments can be the premises of further action-oriented inferences, which can account for the “dynamic uses” of ethical words, namely “inciting people to action” or venting emotions (Stevenson, 1937, pp.  21–22). In both cases, the dynamic uses are aimed at altering people’s attitude (affection), either by encouraging the hearer to commit himself to a course of action or behavior, or arousing sympathy (Ben-Ze’ev, 2000, Chap. 11). In particular, the action-oriented effect can be represented in terms of inferences, which can lead to a commitment (an attitude of the individual towards a judgment or a state of affairs, see Hamblin, 1970; Walton & Krabbe, 1995) or in more complex patterns, leading to further inferences. The passage from a value judgment to a specific commitment can be represented in the following scheme of argument from values (Walton et al., 2008, p. 321). Argumentation scheme 3: Argument from Values Premise 1

The state of affairs x is positive/negative as judged by Agent A according to Value V (value judgment). Premise 2 The fact that x is positive/negative affects the interpretation and therefore the evaluation of goal G of agent A (if x is good, it supports commitment to G). Conclusion The evaluation of x according to value V is a reason for retaining/retracting commitment to G.

The object of the commitment is generically represented as a Goal, which can refer to a generic end towards which an action can be directed (approval or disapproval) or a specific one (supporting a candidate or voting against him). For example, the use of the term “politically correct” for describing one’s behavior can be normally considered as a reason for approving of her, or at least of her behavior. On the contrary, describing a person as “scum” provides a reason for evaluating him or her negatively as worthless, leading to disapproval. This generic commitment can become more specific by drawing the possible actions that can pursue the goal of the agent. Practical reasoning and reasoning from consequences (von Wright, 1963a, 1963b, 1972) can be represented as schemes of inference connecting a desired situation – or rather a “declaration of intention,” a commitment to bringing about a state of affairs – with conditions. In the first case, the agent can reason by selecting (committing to) the productive or necessary means for bringing about the desired state of affairs (the best way of disapproving of Clinton is not to vote her). In the other type of reasoning, called argument from consequences, the agent only considers one relationship between a desirable or undesirable state of affairs to the action that is the necessary or productive cause thereof. This inference can be represented as follows (Walton et  al., 2008, pp. 332–333).

Inferential Patterns of Emotive Meaning

93

Argumentation scheme 4: Argument from consequences. Premise 1 Premise 2 Conclusion

If action Q is brought about, good (bad) consequences will plausibly occur. Good (bad) consequences are (not) desirable (should (not) occur). Therefore, Q should (not) be brought about.

For example, this type of reasoning underlies the passage from the negative evaluation of an individual labeled as “politically correct” or “scum” and his or her possible future actions (politically correct people do not offend others/are good at diplomacy; if someone is scum, he is worthless) to a specific action (I should support him or her; I should despise him or her). This inferential framework can be used for investigating some famous uses of emotive terms in Trump’s campaign, pointing out the strategies used.

Using Emotive Words to Craft Emotions The most prototypical example of use of emotive words to steer the audience’s emotions and evaluation of a state of affairs consists in the use of slurs. For example, we consider the following excerpt from Trump’s campaign rally in Hilton Head Island, South Carolina,5 in which he describes the journalists: Case 1: Scum and other slurs6 They’re scum. They’re horrible people. They are so illegitimate. … Some of the people in the press are honorable. But you’ve got 50% who are terrible people. […] I would never kill them, but I do hate them. And some of them are such lying, disgusting people. It’s true.

In particular, “scum” is a lexicalized metaphor, a slur whose definition includes an evaluation of the subject matter (very bad or immoral person) (Hom, 2008). Other adjectives (“horrible,” “terrible,” “disgusting”) are purely evaluative, namely provide an assessment of the name modified, which is described referring to the emotion that is encouraged (fear, unpleasantness, disgust). Finally, words such as “illegitimate” or “lying” can be considered as properly “ethical words,” i.e. words that describe a state of affairs without involving in their definition any evaluative component (“not accepted by the law as rightful;” “not telling the truth”).

  Transcript retrieved from http://www.kansascity.com/news/local/news-columns-blogs/thebuzz/article55604115.html (Accessed on 01 September 2020). 6  Rick Hampson (2016). Donald Trump’s attacks on the news media: A not-so-short history. USA Today 10 March 2016. Retrieved from: http://www.usatoday.com/story/news/politics/onpolitics/2016/03/10/donald-trump-versus-the-media/81602878/ (Accessed on 01 September 2020). 5

94

F. Macagno and M. G. Rossi

The distinct types of emotive words lead to inferences different in effectiveness, defeasibility, and force, depending on the presumptions relied upon. “Scum,” horrible,” “terrible,” “disgusting” are used by taking for granted the classificatory criteria underlying their assertion (linguistic presumptions), leading to the inferential structure represented in Table 1 below. In contrast with the aforementioned emotive words, the adjectives “illegitimate” and “lying” carry a value judgment indirectly (Bentham, 1824, pp. 213–200; Walton, 1999, pp. 130–137). The evaluative inference triggered by them is distinct from the one on which the attribution of the predicate is based. In other words, the linguistic presumption underlying the conventional meaning is distinct from the evaluative one (presumption 3) and can be represented in Table 2 below. Table 1  Describing the inferential structure of “scum” Reasoning from classification  1. A low and worthless person is scum (Pres. 1).  2. The speaker has reasons to believe that (or it is common knowledge that) journalists have committed actions that makes them low and worthless people.  3. Journalists are scum. Evaluative classification  4. A scum is a low and worthless person (Pres. 1).  5. (from 3 and 4) Journalists are low and worthless people. Reasoning from values  6. (from 5) According to the value of dignity and honor, journalists are considered as extremely low.  7. People characterized by low dignity and honor are inferior and can be dangerous, justifying avoidance and neglect (Ben-Ze’ev, 2000, p. 390) (Pres. 2).  8. (from 3, 6, and 7) The classification of journalists as scum is a reason for avoiding and neglecting them. Table 2  Describing the inferential structure of “lying” Reasoning from classification  1. Who is not telling the truth intentionally to deceive is a lying person (Pres. 1).  2. If the speaker classifies someone as “lying”, he has reasons to believe that he has a habit of not telling the truth intentionally to deceive.  3. Journalists are lying people. Evaluative classification  4. Lying people are unethical and dangerous (Pres. 3).  5. (from 3 and 4) Journalists are unethical and dangerous. Reasoning from values  6. (from 5) According to the value of truthfulness, journalists are considered as extremely low.  7. People characterized by low truthfulness are inferior and can be dangerous, justifying avoidance and neglect, i.e. contempt (Ben-Ze’ev, 2000, p. 390) (Pres. 2).  8. (from 3, 6, and 7) The classification of journalists as lying people is a reason for avoiding and neglecting them (despising them).

Inferential Patterns of Emotive Meaning

95

In this analysis, the defeasibility conditions of the first classificatory inference appear more clearly. Trump uses the epithet “lying” without providing any reason therefore. Thus, the burden of providing reasons for the c­ lassification is not avoided, but taken for granted as fulfilled and shifted onto the interlocutors (Macagno & Walton, 2014, Chap. 5). In Case 1 above, we notice that the emotive meaning can be associated with a term in different fashions. Apart from slurs, some of which at least carry an evaluative component in their definition, and evaluative adjectives, Trump uses metaphors and “ethical” words. Metaphors in particular, carry with them the common ground commonly associated with the vehicle (Clark, 1996; Kovecses, 2015, pp. 179–180; Ritchie, 2006, pp. 88–96), namely the frame or the script (Samet & Schank, 1984) that is activated, or the context in which it has been previously or prototypically used (Clark, 1996, pp.  38–41; Kovecses, 2015, pp.  180–181; Ritchie, 2006, pp. 190–191). In this sense, such previous contexts make specific accidental properties (evaluative) or inferences salient, i.e. accessible because of their conventionality, frequency, familiarity, or prototypicality (Giora, 2003, 2008). For this reason, a metaphor can lead to associations and inferences that can trigger emotions, as pointed out in Port Royal Logic (Arnauld & Nicole, Logic, Chapter I, 13). A clear example of the use of metaphors for arousing emotions can be found in the aforementioned speech given by Trump in South Carolina: Case 2: Behemoth I order thousands of televisions, they’re all from South Korea. So we have 28,000 people on the border separating South Korea from this maniac in North Korea, we get nothing. What do — we get nothing. They’re making a fortune. It’s an economic behemoth. […] A lot of you don’t know we protect Germany. Germany! Mercedes Benz, how many people have a Mercedes Benz? We protect Germany. It’s an economic behemoth.

Here, “behemoth” is used not only for referring to Germany or South Korea using the image of the biblical mighty beast.7 More importantly, the common ground associated with the vehicle due to the previous narrations (monster provoking chaos; monster destroying and eating the world) leads to negative assessments of the targets, leading to immediate emotional responses (danger, ergo fear). The use of this metaphor involves both a classificatory reasoning, based on the premises that Trump shortly provides (they are making a fortune; thousands of TV; etc.) and the common description of “behemoth” (Pres. 1), and a distinct evaluative one, based on the commonly shared actions it performs (Pres. 2 and 3). This reasoning can follow the pattern of the Argumentation scheme 2: (1) A behemoth is a monster eating the world, and (2) since this is a terrible consequence, (3) a behemoth is also a terrible and hateful being.

7  The common ground associated with “behemoth” is usually an entity bigger than it should be (Robert Stevens, 2018. Opinion: Don’t build this behemoth of a project in Villa Hills, USA Today May 8, 2018, Retrieved from https://eu.cincinnati.com/story/opinion/2018/05/08/opinion-dontbuild-behemoth-project-villa-hills/566292002/ (Accessed on 30 May 2020).

96

F. Macagno and M. G. Rossi

A similar strategy is used for eliciting the emotive response of anger (Ben-Ze’ev, 2000, p.  380) by depicting a state of affairs as an undeserved offence. Trump describes the United States as a “dumping ground” in which the world is depositing their garbage, i.e. the migrants: Case 3: Dumping ground And we’re like a dumping ground for the world. We’re a dumping ground. They want to take these migrants  — the migrants, you know, and I feel terrible about the migration, caused by Hillary Clinton and Barack Obama. They’re the ones that caused it. They go into Libya. They knocked the hell out of Gadhafi. OK, so Gadhafi — they backed rebels who end up killing the ambassador and the other young people.

In addition to implicitly labelling the migrants as “garbage,” provoking the evaluation of inferiority which is the basis of contempt (Ben-Ze’ev, 2000, p. 390), Trump presents the problem of migration as an offence of littering the property of the Americans. The emotive words are then connected with the offenders (Obama and Clinton) who can thus become the objects of anger. In addition to metaphors, also similes can trigger similar emotive effects. In the following case, Trump emotionally describes his Republican competitor, Marco Rubio, as “weak like a baby.” The word “baby” has no negative emotive meaning; however, Trump uses to reinforce the evaluation of the opponent as weak, relying on the concept of helplessness usually associated to a baby. This simile, however, presents Rubio as a helpless baby, which is in striking contrast with the position he is aiming at, resulting in contempt and ridicule:8 Case 4: Baby Nice person, weak on illegal immigration … like, weak like a baby. Like a baby. Not a good poker player, because every time he’s under pressure he just starts to profusely sweat. If he was playing poker with me, I’d say ‘Ah!’ The water would start pouring off his body.

The inferential structure of metaphors has the great advantage of providing clear evaluative premises associated with them (a dumping ground is a despicable place), but unclear definitional characteristics, as the vehicle is in itself implicitly redefined (Macagno & Walton, 2014). For example, “dumping ground” can be understood as meaning a place where unwanted beings or entities are collected, but this definition is not necessarily shared, nor is it stated, allowing Trump to use the term without justifying it. In this sense, the weakness of the linguistic presumption is exploited by the speaker to limit the possibilities of being criticized. These cases are all characterized by evaluative premises that can justify the triggering of an emotion, such as contempt (Case 1 and 4), fear (Case 2), or rage (Case 3). When ethical words trigger emotional responses, they contribute to judgment and belief creation. However, such emotional judgments and beliefs are hasty and 8  Margaret Hartmann (2016). Trump Explains Why Rubio Is ‘Weak Like a Baby,’ and Other Hits From His 95-Minute Rant. The New York Magazine, 13 November 2015. Retrieved from http:// nymag.com/daily/intelligencer/2015/11/best-of-trumps-95-minute-rant-compilation.html (Accessed on 01 September 2020).

Inferential Patterns of Emotive Meaning

97

biased, leading to automatic conclusions of right and wrong (Damasio, 1994; Greene & Haidt, 2002; Keltner & Lerner, 2010, p. 331), which in turn are at the basis of a sudden action tendency (depending on the intensity of the emotion), an automatic, time-tested responses, requiring low processing efforts (Loewenstein & Lerner, 2003, p. 628). Within moral psychology, this relation between emotion and moral judgment has been explored, inter alia, by Jonathan Haidt (2001; Haidt & Hersh, 2001; Wheatley & Haidt, 2005). According to Haidt, when we assess moral issues, we almost never behave like a judge who seeks the truth by objectively assessing arguments and evidence. Instead, we act as lawyers: we already have a preferential viewpoint, a conclusion that we intend to support by means of evidence. Thus, we proceed in search of evidence that might persuade the others of the truth of our opinion (Haidt, 2001; Rossi, 2012). The rhetorical effectiveness of the political examples discussed above clearly illustrates this description. In these cases, ethical words trigger emotional responses, which are, from an argumentative point of view, shortcuts of the inferential reconstruction of micro-arguments described in the previous section. Ethical words thus provide or suggest an evaluation that activates a pattern of automatic (heuristic) reasoning (Chen & Chaiken, 1999; Kahneman, 2003, 2011; Petty & Cacioppo, 1986), which has devastating effects on the critical (systematic) assessment to the described or referred state of affairs (Blanchette & Richards, 2004; Loewenstein & Lerner, 2003).

Manipulating the Definition of Emotive Words The “wedding” of descriptive and emotive meaning in ethical (emotive) words leads to a dangerous possibility, namely the use of an emotive word to refer to a state of affairs that is normally assessed neutrally, or even in a way different from the emotive meaning of the term used (Arnauld & Nicole, Logic, Chapter I, 13). This possibility is rooted in the redefinition of the emotive word, i.e. in the strategy that Stevenson named persuasive definition (Macagno & Walton, 2008, 2010; Stevenson, 1938b). Persuasive definitions are explicit or implicit redefinitions of ethical words, aimed at redirecting the emotive meaning on a different state of affairs, which would not be normally predicated of such a word. For example, by redefining “culture” as originality, the speaker can change “interests by changing names,” redirecting the interlocutor’s evaluation of their referents and his related emotions (Macagno & Walton, 2014; Stevenson, 1938b, p. 332). Trump uses this strategy very clearly in his aforementioned speech. In particular, he does not redefine an emotive word; instead, he uses a term (“deal”) commonly used in business to mean a transaction, an arrangement for mutual advantage, to refer to diplomatic negotiations. However, as in the case of metaphors mentioned above, this term carries an additional emotive meaning resulting from its previous and common contexts of use. A deal is related to money, business, private interests,

98

F. Macagno and M. G. Rossi

and in particular to private economic advantages, which are incompatible with the public interests defended in diplomatic negotiations. The effect is the following: Case 5: Deal These are people that have deceived us. They’ve lied to us. They’re a terrorist state. But — and I used to say it’s the worst deal that I’ve ever seen negotiated. And by the way, just to finish prisoners. So they’d come back; we get our prisoners. But then when I hear the other day that now this deal is done. It’s all done. […] Now, we’ve already taken off the sanctions. They’re already rich as hell. What — what’s going on there? That’s why I say, I mean, some people say it’s worst than stupidity. There’s something going on that we don’t know about. I mean, honest. And you almost think  — I’m not saying that, and I’m not a conspiracy person. As you said, “We are; we’re saying it.”

Trump is using the term “deal” to refer to a practice (negotiations between nations) that falls outside the ordinary definition of this term (agreement for mutual advantage). In this case, we notice that the strategy pursued is based on a conflict of presumptions. The generic linguistic presumption governing the use of “deal” is contrasted with the evidence of its repeated use by Trump to refer to political negotiations (Pres. 0), which leads to a different, specific one. However, the factual presumption that in deals the agents are pursuing their own private interests is maintained and indeed reinforced by the pragmatic presumption that the text is intended to denounce the dealmakers (Pres. 0). In this sense, it is not replaced by the factual ones characterizing political discussions. The inferential structure can be reconstructed in the following Table 3. Through this persuasive definition, Trump acts based on a linguistic presumption that is introduced contextually through various uses of this word and the similarity with the commonly shared one, modifying the common ground (Kecskes & Zhang, 2013) and thus avoiding the burden of proving the reasonableness of his classification.

Table 3  Describing the inferential structure of “deal” Reasoning from classification  1. A deal is an agreement for mutual advantage (Pres. 1).  2. Democrats have made deals (stated).  3. Trump uses “deal” in a sense different from “agreement for mutual advantage” as “pursuing one’s own interests also damaging the others.” (from the conclusion of 1 and 2 and the textual evidence). Evaluative classification  4. Who makes a deal, he or she pursues his own interests also damaging the others (Pres. 2; evidence from the text).  5. (from 3 and 4) Democrats pursued their own interests also damaging the others. Reasoning from values  6. (from 5) According to the value of fairness, democrats are considered as extremely low.  7. People characterized by low fairness can be dangerous and unjust, justifying rage (Pres. 2).  8. (from 3, 6, and 7) The classification of democrats as dealmakers is a reason for being upset for and disgusted by their behavior.

Inferential Patterns of Emotive Meaning

99

Crafting Emotive Meaning The use of emotive words exploits the evaluative inferences that can be conceived as based on linguistic, factual, or evaluative presumptions. Persuasive definitions are grounded on a more complex structure, consisting in altering the linguistic presumptions characterizing the shared use of a word. A more complex strategy, called quasi-definition, consists in the modification of the evaluative or factual p­ resumptions that are associated with the use of a word. As persuasive definitions, quasi definition are also characterized by a conflict of presumptions. Evaluative or factual presumptions can be associated with the use of a term by placing it in a context in which it can be interpreted only as a derogatory or praising expression (a pragmatic presumption), and needs to be justified by introducing the presumption that the referent is in fact negative or has done something negatively evaluated. A clear example can be drawn from Trump’s campaign. One of his most successful rhetorical moves was the modification of the emotive meaning of “politically correct.” This phrase is not commonly perceived as necessarily derogatory, even though the history of its use reveals how it commonly suggested a view characterized by Stalinist orthodoxy.9 However, Trump used it in contexts in which it could be only interpreted as meaning an insincere, uncommitted, and even coward person, referring to his political opponents. Politically correctness was identified by Trump as a problem during the first Republican debate in August, when he answered to the attack by the moderator on comments that he had made disparaging women.10 Case 6: Political correctness as the problem (waste of time) I think the big problem this country has is being politically correct […] I’ve been challenged by so many people, and I don’t frankly have time for total political correctness. And to be honest with you, this country doesn’t have time either. This country is in big trouble. We don’t win anymore. We lose to China. We lose to Mexico both in trade and at the border. We lose to everybody.

Politically correctness is here characterized as a problem, and contrasted with serious occupations (I don’t have time) and more importantly with addressing national interests and solving big troubles (this country doesn’t have time either). The pragmatic presumption is that Trump is criticizing politicians for their political correctness, and for this reason a new presumption needs to be introduced connecting political correctness with despicable properties. In this sense, Trump implicitly quasi-defines it as a practice that is usually carried out as a non-serious waste of  Philip Bump (2015). How ‘politically correct’ moved from Commies to culture and back into politics. The Washington Post, 17 December 2015. Retrieved from https://www.washingtonpost. com/news/the-fix/wp/2015/12/17/the-interesting-evolution-of-political-correctness/?utm_ term=.38cf66c867ed. (Accessed on 01 September 2020). 10  Karen Tumulty and Jenna Johnson (2016). Why Trump may be winning the war on ‘political correctness.’ The Washington Post, 4 January 2016. Retrieved from https://www.washingtonpost. com/politics/why-trump-may-be-winning-the-war-on-political-correctness/2016/01/04/098cf832afda-11e5-b711-1998289ffcea_story.html?utm_term=.0f77f4556e35. (Accessed on 01 September 2020). 9

100

F. Macagno and M. G. Rossi

time, a distraction from facing real problems, which characterises a group to which he, the non-politically correct one, does not belong at all. Trump uses the pragmatic presumption (summarized in the conclusion, 9) for taking for granted that political correctness is presumed by everyone to lead to a dangerous waste of time (presumption 2). In this fashion, he modifies the evaluation of the term within this specific context. The quasi-definition of political correctness is achieved by placing the phrase in other contexts, in which the purpose is not attacking a policy or arousing indignation against some politicians, but rather ridiculing the opponent, triggering emotions of contempt. In describing how Rubio and Bush praise each other publicly, mimicking their expressions of mutual esteem, he claimed the following: Case 7: Political correctness as hypocrisy Rubio and Bush “hate each other,” Trump said, blasting Rubio as “overly ambitious, too young, and I have better hair than he does, right?” “But I am so tired of this politically correct crap.”11 So many “politically correct” fools in our country. We have to all get back to work and stop wasting time and energy on nonsense!12

“Politically correct” was implicitly quasi-defined by suggesting the same inferences as above based on the presumed premise that it leads to (and justifies) useless waste of time. However, here Trump presupposes another implicit premise, namely that political correctness is used to hide real beliefs. The most effective speech quasi-defining “politically correct” as “a serious problem” was delivered after a Muslim gunman killed 49 people at a gay nightclub in Orlando.13 Trump identifies the problem that led America to incurring dramatic consequences with being “politically correct:” Case 8: Politically correct as the problem (danger and foolishness) The Obama Administration, with the support of Hillary Clinton and others, has also damaged our security by restraining our intelligence-gathering and failing to support law enforcement. They have put political correctness above common sense, above your safety, and above all else. I refuse to be politically correct.

This example can be reconstructed through an analysis similar to the inferences from consequence to an evaluation and the reasoning from values illustrated in Table  4. Trump here explicitly presents “political correctness” as a principle, or objective, opposite to or at least alternative to common sense and the Americans’

 Nick Gass (2015). Trump: I’m so tired of this politically correct crap. Politico 23 September 2015. Retrieved from http://www.politico.com/story/2015/09/donald-trump-politically-correctcrap-213988 (Accessed on 01 September 2020). 12  https://twitter.com/realdonaldtrump/status/629992743788523520 13  Donald J.  Trump Addresses Terrorism, Immigration, and National Security. Retrieved from https://www.donaldjtrump.com/press-releases/donald-j.-trump-addresses-terrorism-immigrationand-national-security (Accessed on 01 September 2020). 11

Inferential Patterns of Emotive Meaning

101

Table 4  Describing the inferential structure of “politically correct” Reasoning from classification  1. Political correctness is the avoidance of forms of expression or action that are perceived to exclude, marginalize, or insult groups of people (Pres. 1).  2. Politicians usually avoid insulting or excluding groups of people (Pres. 2).  3. Politicians are politically correct. Consequence to evaluation  4. Political correctness causes dangerous waste of time (Pres. 2, justifying Pres. 0; evidence from the text).  5. Dangerous waste of time is extremely negative, as it is sign of lack of will and indifference for the nation (Pres. 3).  6. (from 4 and 5) Political correctness is extremely negative as it is sign of lack of will and indifference for the nation. Reasoning from values  7. (from 3 and 6) According to the values of duty and will, politically correct politicians are considered as extremely low.  8. Poorly determined politicians indifferent to their own nation can be dangerous and inferior, justifying contempt (Pres. 2).  9. (from 3, 7, and 8) Politically correct politicians should be despised.

safety, and is illustrated with the Obama’s Administration decision to restrain “intelligence-­gathering” and fail “to support law enforcement.” In a fourth instance, “political correctness” is presented as the cause of the inability to respond to the attacks: Case 9: Politically correct as a hindrance (inability to act) We need to respond to this attack on America as one united people – with force, purpose and determination. But the current politically correct response cripples our ability to talk and think and act clearly. […] If we do not get tough and smart real fast, we are not going to have a country anymore. Because our leaders are weak, I said this was going to happen – and it is only going to get worse. I am trying to save lives and prevent the next terrorist attack. We can’t afford to be politically correct anymore.

Trump here relies on the inference from consequences to evaluation, presenting political correctness as negative because resulting in inability act and react. The presumption that Trump introduces as shared is that politically correct people are only interested in pursuing their own interests and avoiding getting involved in problematic issues, even to the detriment of the whole country. All the aforementioned negative characteristics of the new emotional meaning of “politically correct” were enhanced by associating them with the idea of injustice or deceit, which triggers emotions such as rage or indignation14:

 Moira Weigel (2016). Political correctness: how the right invented a phantom enemy. The Guardian, 30 November 2016. Retrieved from: https://www.theguardian.com/us-news/2016/ nov/30/political-correctness-how-the-right-invented-phantom-enemy-donald-trump (Accessed on 3 February 2020).

14

102

F. Macagno and M. G. Rossi

Case 10: Politically correct as a deceit That is the choice I put before the American people: a mainstream immigration policy designed to benefit America, or Hillary Clinton’s radical immigration policy designed to benefit politically-correct special interests.

Here, special interests, i.e. the interests (usually economic) of a specific group influencing the government and opposed to the interests of the whole population, are associated with the already quasi-defined concept of politically correct. Trump’s choice to refer to the “special interests” by qualifying them as “politically-correct” can be reconstructed by taking for granted the presumption that political correctness is used as an instrument for justifying to the public a policy detrimental to the American people. Trump considers this premise as commonly shared, which leads to the evaluative inference that political correctness is negative because deceitful. From these example, we notice two crucial features of emotive meaning. First, emotive meaning can be altered by means of re-contextualization, namely placing the term or phrase in a context different from the one in which it was used before or it is commonly used. By re-contextualizing the “politically correct,” Trump links it with a set of negative characteristics (evaluative and factual presumptions), and then uses it to attack his opponents and silence the challenges to his aggressive, racist, or womanizer behaviour. Second, the modification of emotive meaning is achieved implicitly. Trump never states that politically correct people are hypocrites (or bad for other reasons). This evaluative judgment is only implied. The hearer draws the evaluative inference necessary for making the phrase relevant to the context. For example, without this intended inference, it would be impossible to explain why in Case 7 the speaker uses it as an adjective specifying two concepts clearly and commonly negatively evaluated (“crap” and “fools”) and uses the corresponding sentences as a premise for concluding that American needs to “stop wasting time and energy on nonsense.” Without the bridging inference (Clark, 1977, pp.  247–248; Haviland & Clark, 1974) that “politically correct” is a problem (or more specifically useless hypocrisy), Trump’s argument would not make much sense. This inferential aspect of emotive meaning can account also for the fact that the aforementioned recontextualizations are effective in specific contexts. Before a different public (not supporting Trump or not disapproving the existing administration), the evaluative difference would have been difficult to accept, and the sequence of discourse hardly considered as reasonable.

Crafting Heuristics. Connotation and Emotive Meaning In sections “Using Emotive Words to Craft Emotions”, “Manipulating the Definition of Emotive Words”, and “Crafting Emotive Meaning”, we showed how ethical words can be used for eliciting an “affective” (using Stevenson’s terminology) or rather emotional response, and how the “tendency” to be used for this purpose can be modified. In section “Crafting Emotive Meaning” we pointed out how in his speeches,

103

Inferential Patterns of Emotive Meaning

Trump alters the “emotive meaning” of the otherwise neutral or positively evaluated concept of “politically correct.” He does not use the word improperly, namely he does not redefine it (as he in contrast did with “deal”). However, he changes the prototypical context in which it is used, using the phrase to convey negative evaluative inferences. The repeated use of this strategy changes the emotive meaning of the phrase, and makes it as stably associated with negative evaluative inferences. The relative stability of the “emotive meaning” of ethical words in different contexts appears to be an automatic or automatized process parallel to conversational implicatures (Capone, 2011; Macagno, 2017; Nunberg, 2017). Apart from question-­ begging epithets – characterized by evaluative semantic properties – the majority of emotive words analyzed above are characterized by inferences drawn also from the assessment of the state of affairs referred to. The process of “emotivization” (or associative engineering, see Leech, 1981, p.  46) of a word described in section “Crafting Emotive Meaning” suggests how evaluative inferences can become gradually automatized. This process can be investigated in terms of connotation, or rather, a specific account of connotation. The connotative meaning of the majority of the aforementioned emotive words can be considered as resulting from the uses of the term, acquiring an additional meaning triggered not directly by its relationship with its definitional features (and the referent), but as a sign used within a linguistic system or a context. Connotation can be represented as a distinct, but not independent, level of meaning of a sign, using the terminology of Hjelmslev (1969; Kerbrat-Orecchioni, 1977; Molino, 1971) (Fig. 2). The sources of the connotative meaning can be different, including ideological, axiological (value judgment) and emotive characteristics (Kerbrat-Orecchioni, 1977, p. 167) that can convey value judgments that can trigger emotional responses (Cato et al., 2004). These non-definitional features can be the result of the use of a term in a specific utterance to pursue a specific purpose, such as the aforementioned uses of the phrase “politically correct” by Trump. However, this contextual meaning – which we have investigated in terms of inferences – can become integrated at a cultural level (Garza-Cuarón, 1991, pp. 213–214). The context can become culturally associated with the use of a specific term, which acquires a more stable connotative meaning (Kerbrat-Orecchioni, 1977, p. 119). In a sense, the lexical items “encapsulate” prior contexts of experience: “they carry context (prior context), encoding the history of their prior use (prior context) in a speech community”

Expression (sign) Expression

Content

Fig. 2  Representing connotation

Content

Level of Connotation Level of Denotation

104

F. Macagno and M. G. Rossi

(Kecskes, 2008, 2013, p.  129; 131; Kecskes & Zhang, 2009). As Kecskes put it (Kecskes, 2013, p. 133): Lexical items encode the history of their use, which basically creates a record of prior contexts. They trigger frames and cultural models that the interlocutor has experienced before. […] Linguistic units encapsulate the history of their use, i.e., the situations in which they have been used. What happens in communication is that prior context encoded in the ­utterances interplays with the actual situational context, and this interplay results in what we call “meaning.”

In particular, Kecskes points out a dimension of meaning (which involves “word-­ specific elements” and “culture-specific conceptual properties,” see Cruse, 1992; Kecskes, 2003:40–43) distinct from the semantic (definitional, or better conceptual) one. We can draw a correspondence between the first dimension (including both lexicalized and cultural properties of a lexical item) and the “emotive meaning,” and between the second (conceptual) dimension and the “descriptive meaning” (Leech, 1981, p. 12). They jointly constitute what Kecskes refers to as “coresense,” namely “a summary of the most familiar, regular, typical, and (generally, but not always) frequent uses of a word. It reflects the history of use of the word and is the common core information that was called public context above, usually shared by members of a speech community” (Kecskes, 2013, p. 141). This account of connotative or rather contextual (parole) nature of emotive meaning leads to the problem of explaining the mechanisms underlying how it is generated and stabilized in terms of stereotypical inferential patterns (Capone, 2011; Macagno, 2017). A possible explanation can be found in Ducrot’s notion of topos. According to Ducrot, an utterance can be described as a bundle of topoi, namely argumentative connections representing instructions such as “uttering x, the conclusion y is supported” (Anscombre & Ducrot, 1983; Ducrot, 1979). As a consequence, words can be described not starting from previous knowledge of reality (their “descriptive meaning”) but considering their discursive (argumentative) potential (Ducrot, 1984, 1993). Such topoi, or argumentative contents, are considered by Ducrot as presuppositions, namely (in his theory) the content of illocutionary acts of presupposition (Anscombre & Ducrot, 1983, p. 49). The theory of topoi can explain the heuristic judgments that ethical words trigger. For example, we consider some of the cases illustrated in previous section. By referring to journalists as “liars” in Case 1, Trump triggers automatically the judgment that journalists are despicable and should not be trusted, based on the heuristic or topos that, “if someone is a liar, he is untrustworthy and contemptible person.” Similarly, other words commonly used in a pejorative sense, such as “behemoth” or “dumping ground,” can be analyzed in terms of culturally stabilized argumentative inferences. Such “pejoratives” are hardly considered as insults; however, the use in specific contexts to elicit negative conclusions has stabilized the inferences that a hearer can heuristically draw from them (to be dangerously aggressive; to be besmeared by x). The analysis of emotive meaning in terms of connotation and topoi can also explain the process of emotivization referred to as “quasi-definition.” As shown in the section above, Trump implicitly quasi-defines “politically correct” by present-

Inferential Patterns of Emotive Meaning

105

ing this description as the reason of distinct but related negative value judgments. In Case 6, “politically correct” is presented as conflicting with “solving the problems” (if I am politically correct, I cannot/do not have the time to solve more serious problems) taking for granted that political correctness is not serious. In Case 7, “politically correct” is used taking for granted the premise that “people use political correctness to hide their lack of courage.” In Case 9, the implicit premise is that “political correctness is the cause of inability to act,” while in Case 10 Trump takes for granted that, “political correctness is a way for justifying to the public a policy detrimental to the American people.” The most effective context in which this phrase is used for introducing topoi to negative value judgments is Case 8. In a highly dramatic context (a shooting), Trump took for granted that “political correctness” led to security failures. Clearly, all the aforementioned reconstructions are possible interpretations; what matters is that these tacit premises are similar in kind, and their repeated use as commonly shared contributes to create common ground topoi associated with this phrase. The evaluative inferences become thus automatized due to their repeated use. From a logical perspective, the process of automatization of evaluative inferences can be described as the passage from the “quality” of one or more specific instances (for example, of a “politically correct person”) to the “characteristic” of a kind (the “politically correct”) (Dewey, 1938, p. 292). In this sense, emotive force can be regarded as a continuum between the inference drawn from a specific word use to its lexicalization as a property (a feature drawn from the definition of a term, such as “receiving unwanted items” from “dumping ground”) or definitional trait (“being worthless” from “scum”). This continuum can also account for the different “cancellability” conditions of emotive meaning depending on the “ethical word” considered.

Conclusion This paper addresses the argumentative structure of the emotive meaning of “ethical” or “emotive” words, investigating how it can be represented, how it is related to emotions, and how it is generated or manipulated. Ethical words are analyzed starting from Stevenson’s approach to meaning, which distinguishes the descriptive (definitional) meaning from the emotive one. These two dimensions can be represented in terms of the intended inferences conveyed by the use of a word. In particular, ethical words are used to trigger classificatory and practical inferences, which can be critically evaluated using argumentation schemes. By the use of such schemes, it is possible to account for both components of the “expressive force,” namely an evaluation and an action-related dimension. We maintained that both dimensions can be investigated in terms of presumptions, leading to a stereotypical evaluation and a typical communicative act performed by such words (denigrating the interlocutor, for instance). This twofold (evaluative and pragmatic) presumption explains the complex nature of the expressive meaning.

106

F. Macagno and M. G. Rossi

We maintain that emotive meaning can be considered as the result of an evaluative inference (Nunberg, 2017) that has become commonly associated with the use of an ethical word. Emotive meaning can be represented by reconstructing the micro-arguments involved in the use of an ethical word through patterns of arguments called argumentation schemes. The schemes are used for reconstructing the content of the intended inferences, whose presence is commonly identified through pragmatic maxims. In this sense, our focus is not on how the existence of an intended inference is signaled, but on how it can be reconstructed independently of its being automatic or systematic. This analysis has two relevant characteristics. First, it accounts for specific and limited types of conclusions that can be drawn from the use of an ethical term (accounting for the stability and at the same time the variability and ineffability of evaluative judgments that speakers can draw from the use of an ethical term). Second, since argumentative inferences can become automatically (heuristically) processed, it can explain the limited defeasibility of emotive meaning. The description of emotive meaning as automatized inferences can explain also the historical variability of ethical terms. Ethical words can be quasi-defined, namely associated with a new or different emotive meaning. This quasi-definition is claimed to consist in the use of the target term in specific contexts that can evoke emotions or intense value judgments. In this fashion, the ethical word is presented as the premise of similar or related evaluative conclusions. The re-contextualization ends when the new inferences become associated more stably with the use of the ethical word, namely when the implicit premises that before were potentially controversial become part of the “bundle of topoi” associated with the term. On this perspective, connotation is regarded as a set of inferences commonly drawn from habitual uses of a term. The repeated use of the quasi-defined term in similar contexts reinforces the association between the word and the emotion or the value judgment, namely the specific “prior context” that is carried with the ethical word (the context that the linguistic expression creates) and is used for its interpretation (Kecskes, 2013, p. 135; Mey, 2006). Clearly, this inferential contextualization is possible because the premises warranting the evaluative conclusions are taken for granted by the speaker and not challenged by the interlocutors or hearers. On this perspective, the act of introducing a new connotation of words (quasi-definition) becomes an act of manipulating the common ground (Macagno, 2012b, 2015; Macagno & Walton, 2014). Acknowledgments  This work was supported by the Fundação para a Ciência e a Tecnologia (research grants no. SFRH/BPD/115073/2016 and PTDC/FER-FIL/28278/2017).

References Anderson, L., & Lepore, E. (2013). What did you call me? Slurs as prohibited words. Analytic Philosophy, 54(3), 350–363. https://doi.org/10.1111/phib.12023. Anscombre, J.-C., & Ducrot, O. (1983). L’argumentation dans la langue. Bruxelles, Belgium: Pierre Mardaga.

Inferential Patterns of Emotive Meaning

107

Aristotle. (1991). Topics. In J. Barnes (Ed.), The complete works of Aristotle (Vol. I). Princeton, NJ: Princeton University Press. Arnauld, A., & Nicole, P. (1662/1996). Logic or the art of thinking. (J. V. Buroker, Ed.). Cambridge, UK: Cambridge University Press. Atlas, J. D., & Levinson, S. (1981). It-clefts, informativeness and logical form: Radical pragmatics (revised standard version). In P. Cole (Ed.), Radical pragmatics (pp. 1–62). New York, NY: Academic Press. Bach, K., & Harnish, R. (1979). Linguistic communication and speech acts. Cambridge, MA: MIT Press. Bentham, J. (1824). The book of fallacies. London, UK: John and H.L. Hunt. Ben-Ze’ev, A. (2000). The subtlety of emotions. Cambridge, MA: MIT Press. Bianchi, C. (2014a). Slurs and appropriation: An echoic account. Journal of Pragmatics, 66, 35–44. https://doi.org/10.1016/j.pragma.2014.02.009. Bianchi, C. (2014b). The speech acts account of derogatory epithets: Some critical notes. In J.  Dutant, D.  Fassio, & A.  Meylan (Eds.), Liber Amicorum Pascal Engel (pp.  465–480). Genève, Switzerland: University of Geneva. Blakemore, D. (2015). Slurs and expletives: A case against a general account of expressive meaning. Language Sciences, 52, 22–35. https://doi.org/10.1016/j.langsci.2014.06.018. Blanchette, I., & Richards, A. (2004). Reasoning about emotional and neutral materials  – Is logic affected by emotion? Psychological Science, 15(11), 745–752. https://doi. org/10.1111/j.0956-7976.2004.00751.x Capone, A. (2010). What can modularity of mind tell us about the semantics/pragmatics debate? Australian Journal of Linguistics, 30(4), 497–520. https://doi.org/10.1080/07268602.2010.52 2971. Capone, A. (2011). Default semantics and the architecture of the mind. Journal of Pragmatics, 43(6), 1741–1754. https://doi.org/10.1016/j.pragma.2010.11.004 Capone, A. (2013). Explicatures are NOT cancellable. In A. Capone, F. Lo Piparo, & M. Carapezza (Eds.), Perspectives on linguistic pragmatics, perspectives in pragmatics, philosophy & psychology (Vol. 2, pp. 131–151). Cham, Switzerland: Springer. Capone, A. (2014). Indirect reports, slurs, and the polyphonic speaker. Reti, Saperi, Linguaggi, 1(2), 299–318. https://doi.org/10.12832/78931. Capone, A. (2016). The pragmatics of indirect reports: Socio-philosophical considerations. Cham, Switzerland: Springer. Carston, R. (2002). Linguistic meaning, communicated meaning and cognitive pragmatics. Mind and Language, 17(1&2), 127–148. https://doi.org/10.1111/1468-0017.00192 Carston, R. (2004). Truth-conditional content and conversational implicature. In C. Bianchi (Ed.), The semantics/pragmatics distinction (pp. 65–100). Stanford, CA: CSLI Publications. Cato, A., Crosson, B., Gökçay, D., Soltysik, D., Wierenga, C., Gopinath, K., et al. (2004). Processing words with emotional connotation: An FMRI study of time course and laterality. Journal of Cognitive Neuroscience, 16(2), 167–177. https://doi.org/10.1162/089892904322984481. Chen, S., & Chaiken, S. (1999). The heuristic-systematic model in its broader context. In S. Chaiken & Y. Trope (Eds.), Dual-process theories in social psychology (pp. 73–96). New York, NY: Guilford Press. Clark, H. (1977). Inferences in comprehension. In D. LaBerge & J. Samuels (Eds.), Basic processes in reading: Perception and comprehension (pp. 243–263). Hillsdale, NJ: Erlbaum. Clark, H. (1996). Using language. Cambridge, UK: Cambridge University Press. Clark, H., & Brennan, S. (1991). Grounding in communication. In L.  Resnick, J.  Levine, & S. Teasley (Eds.), Perspectives on socially shared cognition (pp. 127–149). Washington, D.C.: American Psychological Association. Croom, A. (2011). Slurs. Language Sciences, 33(3), 343–358. https://doi.org/10.1016/j. langsci.2010.11.005. Croom, A. (2014). The semantics of slurs: A refutation of pure expressivism. Language Sciences, 41, 227–242. https://doi.org/10.1016/j.langsci.2013.07.003

108

F. Macagno and M. G. Rossi

Cruse, A. (1992). Antonymy revisited: Some thoughts on the relationship between words and concepts. In A. Lehrer & E. F. Kittay (Eds.), Frames, fields, and contrasts (pp. 289–306). Hillsdale, NJ: Lawrence Erlbaum Associates. Damasio, A. (1994). Descartes’ error: Emotion, reason, and the human brain. New York, NY: Avon Books. Dancy, J. (2013). Meta-ethics in the twentieth century. In M. Beaney (Ed.), The Oxford handbook of the history of analytic philosophy (pp. 729–749). Oxford UK: Oxford University Press. Dascal, M., & Wróblewski, J. (1988). Transparency and doubt: Understanding and interpretation in pragmatics and in law. Law and Philosophy, 7(2), 203–224. https://doi.org/10.1007/ BF00144156. De Sousa, R. (1987). The rationality of emotion. Cambridge, MA: MIT Press. Dewey, J. (1938). Logic-the theory of inquiry. New York, NY: Henry Holt and Company. Druckman, J. (2002). The implications of framing effects for citizen competence. Political Behavior, 23(3), 225–256. https://doi.org/10.1023/A:1015006907312 Ducrot, O. (1979). Les lois de discours. Langue Française, 42, 21–33. Ducrot, O. (1984). Le dire et le dit. Paris, France: Minuit. Ducrot, O. (1993). Les topoi dans la “Théorie de l”argumentation dans la langue. In C. Plantin (Ed.), Lieux communs, topoi, stéréotypes (pp. 233–248). Paris, France: Kimé. Entman, R.  M. (1993). Framing: Toward clarification of a fractured paradigm. Journal of Communication, 43(4), 51–58. https://doi.org/10.1111/j.1460-2466.1993.tb01304.x Garza-Cuarón, B. (1991). Connotation and meaning. Berlin, Germany: De Gruyter. Geach, P. T. (1965). Assertion. The Philosophical Review, 74(4), 449–465. Giora, R. (2003). On our mind. Salience, context, and figurative language. Oxford, UK: Oxford University Press. Giora, R. (2008). Is Metaphor Unique? In R. Gibbs (Ed.), The Cambridge handbook of metaphor and thought (pp. 143–160). New York, NY: Cambridge University Press. Greene, J., & Haidt, J. (2002). How (and where) does moral judgment work? Trends in Cognitive Sciences, 6(12), 517–523. https://doi.org/10.1016/S1364-6613(02)02011-9. Grice, P. (1975). Logic and conversation. In P. Cole & J. Morgan (Eds.), Syntax and semantics 3: Speech acts (pp. 41–58). New York, NY: Academic Press. Haidt, J. (2001). The emotional dog and its rational tail: A social intuitionist approach to moral judgment. Psychological Review, 108(4), 814–834. https://doi.org/10.1037/0033-295x.108.4.814. Haidt, J., & Hersh, M. (2001). Sexual morality: The cultures and emotions of conservatives and liberals. Journal of Applied Social Psychology, 31(1), 191–221. https://doi.org/10.1111/ j.1559-1816.2001.tb02489.x. Hamblin, C. L. (1970). Fallacies. London, UK: Methuen. Haviland, S., & Clark, H. (1974). What’s new? Acquiring new information as a process in comprehension. Journal of Verbal Learning and Verbal Behavior, 13(5), 512–521. https://doi. org/10.1016/S0022-5371(74)80003-4 Hedger, J. (2012). The semantics of racial slurs: Using Kaplan’s framework to provide a theory of the meaning of derogatory epithets. Linguistic and Philosophical Investigations, 11, 74–84. Hjelmslev, L. (1969). Prolegomena to a Theory of Language. Madison, WI: The University of Wisconsin Press. Hom, C. (2008). The semantics of racial epithets. The Journal of Philosophy, 105(8), 416–440. https://doi.org/10.5840/jphil2008105834. Hom, C. (2010). Pejoratives. Philosophy Compass, 5(2), 164–185. https://doi. org/10.1111/j.1747-9991.2009.00274.x. Jaszczolt, K. (2005). Default semantics. Oxford, UK: Oxford University Press. Kahneman, D. (2003). A perspective on judgment and choice. American Psychologist, 58(9), 697–720. https://doi.org/10.1037/0003-066X.58.9.697. Kahneman, D. (2011). Thinking, fast and slow. New York, NY: Macmillan Publishing Company. Kecskes, I. (2003). Situation-bound utterances in L1 and L2. Berlin, Germany/New York, NY: Mouton de Gruyter.

Inferential Patterns of Emotive Meaning

109

Kecskes, I. (2008). Dueling contexts: A dynamic model of meaning. Journal of Pragmatics, 40(3), 385–406. https://doi.org/10.1016/j.pragma.2007.12.004 Kecskes, I. (2013). Intercultural pragmatics. Oxford, UK: Oxford University Press. Kecskes, I., & Zhang, F. (2009). Activating, seeking, and creating common ground: A socio-­ cognitive approach. Pragmatics & Cognition, 17(2), 331–355. https://doi.org/10.1075/ pc.17.2.06kec Kecskes, I., & Zhang, F. (2013). On the dynamic relations between common ground and presupposition. In A. Capone, F. Lo Piparo, & M. Carapezza (Eds.), Perspectives on linguistic pragmatics, perspectives in pragmatics, philosophy & psychology (Vol. 2, pp. 375–395). Cham, Switzerland: Springer. Keltner, D., & Lerner, J. (2010). Emotion. In S. Fiske, D. Gilbert, & G. Lindzey (Eds.), Handbook of social psychology (pp. 317–352). New York, NY: Wiley. Kerbrat-Orecchioni, C. (1977). La connotation. Lyon, France: Presses Universitaires de Lyon. Kovecses, Z. (2015). Where metaphors come from: Reconsidering context in metaphor. Oxford, UK: Oxford University Press. Leech, G. (1981). Semantics: The study of meaning (2nd ed.). Harmondsworth, UK: Penguin. Leech, G. (1983). Principles of pragmatics. London, UK: Longman. Loewenstein, G., & Lerner, J. (2003). The role of affect in decision making. In R. Davidson, K. Scherer, & H. Goldsmith (Eds.), Handbook of affective science (pp. 619–642). Oxford, UK: Oxford University Press. Macagno, F. (2012a). Presumptive reasoning in interpretation. Implicatures and conflicts of presumptions. Argumentation, 26(2), 233–265. https://doi.org/10.1007/s10503-011-9232-9 Macagno, F. (2012b). Reconstructing and assessing the conditions of meaningfulness: An argumentative approach to presupposition. In H. Ribeiro (Ed.), Inside arguments: Logic and the study of argumentation (pp. 247–268). Newcastle upon Tyne, UK: Cambridge Scholars Publishing. Macagno, F. (2015). Presupposition as argumentative reasoning. In A. Capone & J. Mey (Eds.), Interdisciplinary studies in pragmatics, culture and society (pp. 465–487). Cham, Switzerland: Springer. Macagno, F. (2017). Defaults and inferences in interpretation. Journal of Pragmatics, 117, 280–290. https://doi.org/10.1016/j.pragma.2017.06.005 Macagno, F., & Walton, D. (2008). Persuasive definitions: Values, meanings and implicit disagreements. Informal Logic, 28(3), 203–228. https://doi.org/10.22329/il.v28i3.594. Macagno, F., & Walton, D. (2010). What we hide in words: Emotive words and persuasive definitions. Journal of Pragmatics, 42(7), 1997–2013. https://doi.org/10.1016/j.pragma.2009.12.003 Macagno, F., & Walton, D. (2013). Implicatures as forms of argument. In A. Capone, F. Lo Piparo, & M.  Carapezza (Eds.), Perspectives on pragmatics and philosophy (pp.  203–225). Cham, Switzerland: Springer. Macagno, F., & Walton, D. (2014). Emotive language in argumentation. Cambridge, UK: Cambridge University Press. Macagno, F., Walton, D., & Tindale, C. (2014). Analogical reasoning and semantic rules of inference. Revue Internationale de Philosophie, 270(4), 419–432. https://doi.org/10.3917/ rip.270.0419. Mey, J. (2006). Pragmatic acts. In K. Brown (Ed.), Encyclopedia of language and linguistics (pp. 5–11). Oxford, UK: Elsevier. Molino, J. (1971). La connotation. La Linguistique, 7(1), 5–30. Nunberg, G. (2017). The social life of slurs. In D. Fogal, D. Harris, & M. Moss (Eds.), New work on speech acts. Oxford, UK: Oxford University Press. Patterson, D. (2005). Interpretation in law. San Diego Law Review, 42, 685–710. Perelman, C., & Olbrechts-Tyteca, L. (1951). Act and person in argument. Ethics, 61(4), 251–269. Petty, R., & Cacioppo, J. (1986). The elaboration likelihood model of persuasion. Advances in Experimental Social Psychology, 19, 123–205. https://doi.org/10.1016/S0065-2601(08)60214-2 Richard, M. (2008). When truth gives out. Oxford, UK: Oxford University Press.

110

F. Macagno and M. G. Rossi

Richard, M. (2012). Precis of when truth gives out. Philosophical Studies, 160(3), 441–444. https://doi.org/10.1007/s11098-011-9792-4 Ritchie, D. (2006). Context and connection in metaphor. New York, NY: Palgrave. Rossi, M. G. (2012). On the nature of moral conflicts. In P. Graziani, L. Guzzardi, & M. Sangoi (Eds.), Open problems in the philosophy of sciences (pp. 147–159). London, UK: College Publications. Samet, J., & Schank, R. (1984). Coherence and connectivity. Linguistics and Philosophy, 7(1), 57–82. https://doi.org/10.1007/BF00627475. Schlenker, P. (2007). Expressive presuppositions. Theoretical Linguistics, 33(2), 237–245. https:// doi.org/10.1515/TL.2007.017. Sniderman, P. M., & Theriault, S. M. (2004). The structure of political argument and the logic of issue framing. In W. Saris & P. Sniderman (Eds.), Studies in public opinion: Attitudes, nonattitudes, measurement error, and change (pp. 133–165). Princeton, NJ: Princeton University Press. Sperber, D., & Wilson, D. (2002). Pragmatics, modularity and mind-reading. Mind and language, 17(1&2), 3–23. https://doi.org/10.1111/1468-0017.00186 Stevenson, C. (1937). The emotive meaning of ethical terms. Mind, XLVI(181), 14–31. https://doi. org/10.1093/mind/XLVI.181.14 Stevenson, C. (1938a). Ethical judgments and avoidability. Mind, XLVII(185), 45–57. https://doi. org/10.1093/mind/XLVII.185.45 Stevenson, C. (1938b). Persuasive definitions. Mind, 47, 331–350. https://doi.org/10.1093/mind/ xlvii.187.331. Stevenson, C. (1944). Ethics and language. New Haven, CT: Yale University Press. Stump, E. (2004). Boethius’s “De topicis differentiis”. Ithaca, NY and London, UK: Cornell University Press. Vendler, Z. (1963). The grammar of goodness. The Philosophical Review, 72(4), 446–465. von Wright, G. (1963a). Norm and action. London, UK: Routledge & Kegan Paul. von Wright, G. (1963b). Practical inference. The Philosophical Review, 72(2), 159–179. von Wright, G. (1963c). The varieties of goodness. London, UK: Routledge. von Wright, G. (1972). On so-called practical inference. Acta Sociologica, 15(1), 39–53. https:// doi.org/10.1177/000169937201500104 Walton, D. (1995). Argumentation schemes for presumptive reasoning. Mahwah, NJ: Routledge. Walton, D. (1999). One-sided arguments: A dialectical analysis of bias. Albany, NY: SUNY Press. Walton, D., & Krabbe, E. (1995). Commitment in dialogue. Albany, NY: State University of New York Press. Walton, D., Reed, C., & Macagno, F. (2008). Argumentation Schemes. New York, NY: Cambridge University Press. Wheatley, T., & Haidt, J. (2005). Hypnotic disgust makes moral judgments more severe. Psychological Science, 16(10), 780–784. https://doi.org/10.1111/j.1467-9280.2005.01614.x. Whiting, D. (2013). It’s not what you said, it’s the way you said it: Slurs and conventional implicatures. Analytic Philosophy, 54(3), 364–377. https://doi.org/10.1111/phib.12024. Williamson, T. (2009). Reference, inference and the semantics of pejoratives. In J.  Almog & P.  Leonardi (Eds.), The philosophy of David Kaplan (pp.  137–158). Oxford, UK: Oxford University Press.

Seniors, Older People, the Elderly, Oldies, and Old People: What Language Reveals about Stereotypes of Ageing in Australia Keith Allan, Réka Benczes, and Kate Burridge

Researchers in an Australian Research Council funded project titled ‘The Cultural Model of Ageing: The Australian Conceptualisation of the Third Age’ created a SurveyMonkey questionnaire for speakers of Australian English that sought to identify the characteristics of reference sets for five noun phrases commonly used to label older Australians. 654 self-selected participants were asked which one of the five noun phrases was best matched to each of 25 characteristics found frequently in the media. The results are reported in this paper, and they are interesting. The notion of semantic stereotypes is reviewed in preparation for the discussion of such results as the following. The NP seniors is associated with positive personal characteristics of health and well-being such as ‘like to travel’, ‘lead an involved and active life’, ‘are vibrant and full of purpose’. The NP older people is also associated with positive characteristics, but somewhat less so than seniors and more socially (other) oriented. Older people are seen to ‘benefit the workforce through their experience’, ‘have wisdom and can always be turned to for advice’, ‘play an important role in their extended family’s life’. By contrast, the characteristics of those typically referred to by the NP the elderly are negative in the sense that the referents are incompetent or impose a burden on society, cf. ‘are frail and fall more often’, ‘are often victims of mental and physical abuse’, ‘are unable to look after themselves and depend on others for help’. The referents of the NPs old people and oldies have no particular set of characteristics assigned to them; perhaps that is why they only figure in the one (negative) characteristic ‘are tight-fisted with money’ K. Allan (*) · K. Burridge Monash University, Melbourne, Australia e-mail: [email protected] R. Benczes Institute of Communication and Sociology, Corvinus University of Budapest, Budapest, Hungary © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 F. Macagno, A. Capone (eds.), Inquiries in Philosophical Pragmatics, Perspectives in Pragmatics, Philosophy & Psychology 28, https://doi.org/10.1007/978-3-030-56696-8_6

111

112

K. Allan et al.

that itself is not strongly associated with any one of the five NPs: 28% old people, 23% the elderly, 20% older people, 14.5% oldies, 14.5% seniors. These results are discussed in some detail in the light of differences among different age groupings of the participants.

Overview The Australian Research Council funded project titled ‘The Cultural Model of Ageing: The Australian Conceptualisation of the Third Age’ is investigating linguistic aspects of the way in which Australians talk and think about ageing.1 In this paper we report on what they say about the older members of their nation. There is, of course, no precise onset to old age but it is appropriate to name it as 60 on the basis that 60 years of age is the earliest eligibility for a (state/territory) government Seniors Card (http://‌www.australia.gov.au/content/seniors-card). The end point is, necessarily, the decease of the older person. The oldest recorded Australian died at 114.4  years (Christina Cock, 25/12/1887–22/5/2002)  – which gives a potential maximum span of around 55 years as an older Australian (48% of a life-span!). Our language of ageing project created a SurveyMonkey questionnaire (https://‌www.surveymonkey.com/r/5X9BRGZ) for speakers of Australian English that sought to identify the characteristics of reference sets for five noun phrases commonly used to label older Australians, namely old people, older people, oldies, seniors, and the elderly. 654 self-selected participants were asked which one of the five noun phrases was best matched to each of 25 characteristics found frequently in the media. The task was to replace X with one of the five noun phrases old people, older people, oldies, seniors, the elderly in the following 25 statements: X are vibrant and full of purpose. X prove an increasing economic burden on society. X are poor. X serve as role models and are respected members of society. X are more conservative in attitude and thinking. X are more likely to cause road accidents. X are wealthy and like to spend their money. X are reliable. X are more likely to suffer from depression. X are often victims of mental and physical abuse. X play an important role in their extended family’s life. X are tight-fisted with money. X are often lonely and bored. X like to travel. 1  This research was supported under Australian Research Council’s Discovery Projects funding scheme, project number DP 140102058.

Seniors, Older People, the Elderly, Oldies, and Old People…

113

X are more susceptible to illnesses and disabilities. X benefit the workforce through their experience. X are conscious about their health and lead a healthy lifestyle. X are forgetful. X have a sense of satisfaction with their life. X contribute economically to society. X are unable to look after themselves and depend on others for help. X have no sex lives. X have wisdom and can always be turned to for advice. X are frail and fall more often. X lead an involved and active life. The list of statements was randomized and the order in which the expressions appeared under each statement was varied. The SurveyMonkey results are reported in this paper, and they are interesting because the different meanings ascribed to them are unexpectedly coherent. In brief overview, the NP seniors is associated with positive personal characteristics of health and well-being such as ‘like to travel’, ‘lead an involved and active life’, ‘are vibrant and full of purpose’.2 The NP older people is also associated with positive characteristics, but somewhat less so than seniors and, furthermore, is more socially (other) oriented. Older people are seen to ‘benefit the workforce through their experience’, ‘have wisdom and can always be turned to for advice’, ‘play an important role in their extended family’s life’. By contrast, the characteristics of those typically referred to by the NP the elderly are negative in the sense that the referents are incompetent or impose a burden on society, cf. ‘are frail and fall more often’, ‘are often victims of mental and physical abuse’, ‘are unable to look after themselves and depend on others for help’. The referents of the NPs old people and oldies have no particular set of characteristics assigned to them; perhaps that is why they only figure in the one (negative) characteristic ‘are tight-fisted with money’ that itself is not strongly associated with any particular one of the five NPs: 28% old people, 23% the elderly, 20% older people, 14.5% oldies, 14.5% seniors. (Percentages are normally adjusted to the nearest whole number.) Respondents were asked to assign themselves to gender and age groups. We have not assessed the data for any gender differences, though 69% of respondents were female. But in this paper we do consider the responses of different age groups. These are ≤30 years (55 participants), 31–50 years (55 participants), 51–70 years (313 participants), ≥71 years old (231 participants). These we might respectively call ‘the young people’, ‘the middle-aged’, ‘the younger old’, and ‘the older old’.

2  There is a flourishing of recent acronyms referring to active and affluent older persons: GOFER “genial old fart enjoying retirement”; GLAM “greying, leisured, affluent, married”; WOOP “well off older person”; WOOF “well off older folk”. Perhaps most popular is SKI “spending the kids’ inheritance” (e.g. the SKI Club claims ‘we are all about “Spending the Kids Inheritance”, and having a great time before that inevitable day comes’ (http:// http://www.skiclubaustralia.com.au/ about-the-ski-club/)

114

K. Allan et al.

Note that people we are calling ‘the younger old’ includes those up to 9 years junior to the official onset of old age at 60. To say that ‘the NP seniors is associated with positive personal characteristics of health and well-being’ whereas the NP the elderly is applied to people who ‘are incompetent or impose a burden on society’ suggests there are semantic stereotypes, and we wish to say something about such stereotypes. We hold that: The meaning of a language expression e is not a well-defined set of properties necessarily found in every denotatum for the expression, it is rather a minimum set of stereotypical facts about the typical denotatum. (Allan, 2006: 153; see also Allan, 2001: 333)

(Denotation is the relation between language expression e and things or events in the model of the world and time spoken of; a denotatum is potentially a referent – something spoken or written about, see Allan, 2013.) Like prototype semantics, stereotype semantics is an alternative to theories of meaning that postulate a checklist of properties to be satisfied for the correct use of the expression e (Fillmore, 1975: 123). For example, the denotatum of bird is expected to be bipedal, have feathers, and be capable of flight. But: there are several species of flightless birds (e.g., emus, ostriches, and penguins); a downy chick and a plucked chicken are featherless but nonetheless birds; a one-legged bird and a mutant three-legged bird are also birds. So the notion of a checklist of essential properties for the denotatum of e is problematical. Flightlessness is a fact that should appear in the lexicon entries for emu, ostrich, and penguin, for example, (see Allan, 2011 for how this is achieved) to condition an appropriate interpretation of the ostrich flew by. Prototype semantics differs from stereotype semantics in that it selects a particular denotatum or a particular sense as the most typical exemplar for a lexeme. Categories such as Rugs and Carpets (corresponding to the lexemes rug and carpet) or Cups and Mugs (corresponding to cup and mug) seem to merge gradually one into another, although prototypical exemplars of each category are clearly different (Labov, 1978). For instance, a tomato is both a fruit and a vegetable, but mostly it is used like a vegetable. Its characteristics are included in the internalized (cognitive) model of the tomato and contribute to the semantic frame of tomato. Prototype theory has no means of capturing such facts. Even the poorest examples of a category remain within the category; the fact that an onion is a poor exemplar of a vegetable for a tested population (Battig & Montague, 1969) does not change the subjects’ perception that it is a vegetable. The prototype in no way defines the category – no one believes that a vegetable is a carrot, though Battig & Montague 1969 claim the carrot to be the prototypical vegetable. Furthermore, categories are independent of related categories with respect to prototypicality. For instance, a guppy is a good exemplar of the category Pet Fish, but it is not a good exemplar of either of the categories Pet or Fish. Let us assume that a prototype is as good an example as can be found for the purpose in hand. This implies that the prototype is selected as superior to other denotata. A stereotype is not, hence the description ‘typical denotatum’ in our definition. How is ‘a (stereo)typical denotatum of e’ distinguishable from ‘as-good-an-exemplar-as-can-be-found among the class of things denoted by e’? Presumably, the stereotype properly includes the prototype. For instance,

Seniors, Older People, the Elderly, Oldies, and Old People…

115

whatever the stereotype of vegetable may be, it properly includes the prototype carrot and the peripheral onion (Battig & Montague, 1969). The stereotypical vehicle includes the prototypical car and/or bus together with the peripheral horse-drawn wagon. If this is correct, then we should favour the stereotype in giving the semantics of language expressions. In §§2–4 it will be shown that certain characteristics are assigned to the typical referent of the three NPs seniors, older people and the elderly. The other two NPs mentioned above, oldies and old people, will be discussed in §5. The final §6 presents conclusions.

Seniors Overall, participants attributed to the referents of seniors positive personal characteristics of being outgoing, healthy, active, wealthy, and happy. ( 1) 67% of participants claim that seniors ‘like to travel’. (2) 66% of participants claim that seniors ‘lead an involved and active life’. (3) 59% of participants claim that seniors ‘are vibrant and full of purpose’. (4) 57% of participants claim that seniors ‘are wealthy and like to spend their money’. (5) 55% of participants claim that seniors ‘are conscious about their health and lead a healthy lifestyle’. (6) 52% of participants claim that seniors ‘contribute economically to society’. (7) 49% of participants claim that seniors ‘have a sense of satisfaction with their life’. Seniors like to travel, (1), entails that seniors are willing and able to travel which in turn entails that seniors have the time, health, vitality, and economic resources to travel; furthermore, willingness to travel is purposive intentional behaviour. All these entailed traits are entirely consistent with the attributes of seniors ascribed in (2), (3), (4), and (5). These facts surely confirm that the survey results are, thus far, valid. Indeed, we shall see that our survey results all demonstrate a comparable validity. Germane to this attributed characteristic is the fact that a Senior’s Card sometimes (on trains, trams, buses, and ferries) enables travel at a discount price. The card (conceptually at least) can metonymically stand for the action it is involved in (object for action  – card for travel), and this metonymical asociation might have influenced the selection of seniors with liking to travel. The overall response of 67% reported in (1) is supported for each age group surveyed, but although every age group assigned the same kind of significance to seniors liking travel vs other characteristics of seniors, the results show that older people hold this belief more firmly than younger people: ≤30  years 45%; 31–50  years 63%; 51–70  years 65%; ≥71 years 75%. What this shows is that as one ages, Australians are more likely to assign the attribute of liking to travel to those they call ‘seniors’.

116

K. Allan et al.

Like (1), (2) identifies something that seniors like to do: seniors lead an involved and active life. This is obviously strongly consistent with (1), indeed it is entailed by both (1) and (2) and itself implies (3) – because seniors who lead an involved and active life are very likely to be vibrant and full of purpose. Once again this attribute is more significant as we get older, though the figures are not so disparate between the age groups as for (1). It is quite likely that, if seniors lead an involved and active life, they will be healthy (see (4)) and have a sense of satisfaction with their life (see (7)). So once again there is a consistency and coherence among survey responses. Seniors lead an involved and active life was assessed by 52% of those 30 years and under and roughly 68% for those over 30; more precisely, 31–50  years 67%, 51–70 years 66%, ≥71 years 70%. 59% of participants claimed that seniors are vibrant and full of purpose, (3). This was the view of 31% of respondents 30 years and under, 58% of those 31–70, and 68% of those over the age of 70. If you are vibrant and full of purpose you will probably have a sense of satisfaction with your life (see (7)) and it may well lead you to be conscious about leading a healthy lifestyle (see (4)) and also lead an involved and active life (see (2)). 57% of participants claim that seniors are wealthy and like to spend their money, (4). (4) unquestionably facilitates (1) though it is not, strictly speaking, a precondition for (1) because although it is almost impossible to travel without any money, the traveller does not need to be wealthy and nor to enjoy spending money. (4) enables (6), but otherwise this characteristic of having wealth that one is happy to spend is independent of the other characteristics of seniors except in being a highly desirable attribute. Perhaps it is only to be expected that people beyond the age of 50 (older Australians) are less enthusiastic about this assessment of seniors than the other traits listed in (1) to (7): ≤30 years 40%; 31–50 years 51%, over 50 years, 59.5%. 55% of participants claim that seniors are conscious about their health and lead a healthy lifestyle. This confirms that ‘healthy ageing’ is attributed to seniors. Furthermore, healthy ageing facilitates characteristics (1), (2), (3), and (7). Only 29% of young people (≤30 years) hold this assessment; 60% of the middle-aged (31–50 years) and slightly fewer, 57% of those over the age of 50. (6), seniors contribute economically to society, focuses on a generous, socially aware characteristic by contrast with the potentially more selfish (4), seniors are wealthy and like to spend their money; though as we have said, (4) (at 57%) enables (6) (at 52%). Increasing age leads our respondents to favour this trait among seniors: ≤30 years 35%; 31–50 years 45%, 51–70 years 51%, ≥71 years 58%. Finally, 49% of participants claim that seniors have a sense of satisfaction with their life, which is a manifestation of ‘successful ageing’.3 This can reasonably arise from all the characteristics identified in (1)–(6). It is interesting that this trait is ascribed to seniors more strongly by the middle-aged than by either the young or the

 The term successful ageing entered the gerontological literature (and general academic debate) in about 1987 (Rowe & Kahn, 1987), to denote the idea that an ever-increasing number of older people were leading an active and healthy lifestyle and were still contributing to society.

3

Seniors, Older People, the Elderly, Oldies, and Old People…

117

old: 27% of the young (≤30 years), 48% of the younger old (51–70 years), 52% of the older old (≥71 years), but 58% of those aged 31–50 years. Ranged from higher to lower ranking, young people rank seniors as (a) leading an involved and active life, (b) liking to travel, (c) being wealthy and liking to spend their money, (d) contributing economically to society, (e) being vibrant and full of purpose, (f) being conscious about their health and leading a healthy lifestyle and, finally, (g) having a sense of satisfaction with their life. The middle-aged rank seniors as (a) leading an involved and active life, (b) liking to travel, (c) being conscious about their health and leading a healthy lifestyle, (d) being vibrant and full of purpose and having a sense of satisfaction with their life, (f) being wealthy and liking to spend their money and, finally, (g) contributing economically to society. The younger old rank seniors as (a) leading an involved and active life, (b) liking to travel, (c) being wealthy and liking to spend their money, (d) being vibrant and full of purpose, (e) being conscious about their health and leading a healthy lifestyle, (f) contributing economically to society and, finally, (g) having a sense of satisfaction with their life. The older old rank seniors as (a) liking to travel, (b) leading an involved and active life, (c) being vibrant and full of purpose, (d) being wealthy and liking to spend their money, (e) contributing economically to society, (f) being conscious about their health and leading a healthy lifestyle and, finally, (g) having a sense of satisfaction with their life. Across all age groups, seniors are primarily seen as liking to travel and leading an involved and active life. Young people see the next most prevalent characteristics of seniors as being wealthy and liking to spend their money and contributing economically to society. The middle-aged put these characteristics last. The younger old have them third and fifth, and the older old fourth and fifth. All groups except the middle-aged place the characteristic of having a sense of satisfaction with their life last. But, as we have said already, overall, participants attributed to the referents of seniors positive personal characteristics of being outgoing, healthy, active, wealthy, and happy.

Older People The six characteristics assigned to the referents of the NP older people have them as socially responsible benefactors who have wisdom and experience to share, though they are also likely to be conservative in outlook. (8) 61% of participants claim that older people ‘benefit the workforce through their experience’. (9) 54% of participants claim that older people ‘have wisdom and can always be turned to for advice’.

118

K. Allan et al.

(10) 47% of participants claim that older people ‘play an important role in their extended family’s life’. (11) 47% of participants claim that older people ‘are reliable’. (12) 42% of participants claim that older people ‘serve as role models and are respected members of society’. (13) 38% of participants claim that older people ‘are more conservative in attitude and thinking’. Older people benefit the workforce through their experience, (8), is a trait wholly consistent with characteristics (9), (10), (11), and (12). Although a positive trait, it has no direct connection with any of the characteristics assigned to seniors in (1)– (7). The same is generally true of all characteristics of older people identified in (8)–(13), revealing unambiguous community differentiation in the applicability of the labels seniors and older people. (8) implies that older people have experience that they typically utilise to benefit the workforce, the focus is on them as do-­ gooders. Their experience is echoed in (9) and their benefit as role models is explicit in (12). The young (≤30 years) and the older old (≥71 years) are both 56% on this trait whereas those 31–50 are 63% and the younger old slightly higher at 64%. (9), 54% of participants claim that older people have wisdom and can always be turned to for advice, is similar to (8) except that the focus is on the owning of wisdom as a resource that can be tapped (rather than its beneficial outcome). (9) is potentially a prerequisite for (10) and is perhaps a function of (11). This trait is ascribed to older people more strongly by the middle-aged, 60%, than other age groups; the young only 36%, the younger old 55%, and the older old 56%. 47% of participants claim that older people play an important role in their extended family’s life, (10). (10) is obviously comparable with (8) and (9); it differs from (8) in focusing on the family instead of the workplace, and (9) will often be presupposed with respect to (10)  – though only emotional or physical assistance might be called on in (10) if there is no particular call on the wisdom of the older person. It is most likely that all of (8), (9), and (10) presuppose (11). For this characteristic the various age groups show a bell curve: ≤30 years 36%, 31–50 years 52%, 51–70 years 54%, and the older old 40% (30% of them ascribed this trait to seniors). 47% of participants claim that older people are reliable, (11), which is most likely presupposed by all of (8)–(10). This reliability is significant when older people are turned to for their experience and wisdom. For this trait, once again the age groups show a bell curve, but it peaks with the middle-aged instead of the younger old: ≤30 years 33% (also to seniors 33%), 31–50 years 58%, 51–70 years 52%, and ≥ 71 years 42% (43% to seniors). (12), 42% of participants claim that older people serve as role models and are respected members of society. (8)–(11) offer some reasons for (12). Only 22% of young participants ascribed this characteristic to older people whereas 38% ascribed it to seniors and another 38% to the elderly. The middle-aged also ascribed this trait predominantly to seniors: 31–50 years 38% to older people but 47% to seniors, and the older old (≥ 71 years) 39% to older people and 38% to seniors; only those aged

Seniors, Older People, the Elderly, Oldies, and Old People…

119

51–70 assign this trait most definitely to older people, 48%; this could be aspirational. The final trait assigned to older people is unconnected with the other characteristics assigned to them: 38% of participants claim that older people are more conservative in attitude and thinking. It depends on your world view whether this is a positive or negative characteristic, but unlike those in (8)–(12), it is a personal, not socially-oriented, trait  – which perhaps goes some way to explaining its being ranked only 38% overall. Even so, for all groups this trait was assigned to older people by a large margin, even if younger people (≤50 years) are the firmest believers (perhaps believing themselves comparatively progressive): ≤30  years 40%, 31–50 years 47%, 51–70 years 35%, and ≥ 71 years 39%. Young people rank older people as (a) benefiting the workforce through their experience, (b) being more conservative in attitude and thinking, (c, d) having wisdom and can always be turned to for advice and playing an important role in their extended family’s life, (e) being reliable, and (f) serving as role models and being respected members of society. The middle-aged rank older people as (a) benefiting the workforce through their experience, (b) having wisdom and can always be turned to for advice, (c) being reliable, (d) playing an important role in their extended family’s life, (e) being more conservative in attitude and thinking, and (f) serving as role models and being respected members of society (47% assigned this trait to seniors). The younger old rank older people as (a) benefiting the workforce through their experience, (b) having wisdom and can always be turned to for advice, (c) playing an important role in their extended family’s life, (d) being reliable, (e) serving as role models and being respected members of society, and (f) being more conservative in attitude and thinking. The older old rank older people as (a, b) benefiting the workforce through their experience, and having wisdom and can always be turned to for advice, (c) being reliable (43% assigned this property to seniors), (d) playing an important role in their extended family’s life, (e, f) serving as role models and being respected members of society (38% assigned this property to seniors), and being more conservative in attitude and thinking. With the exception of the trait older people are more conservative in attitude and thinking, the referents of the NP older people are socially responsible benefactors who have wisdom and experience to share with others within and outside the family. The characteristics of reliability and serving as role models, which can be regarded as personal strengths, were sometimes significantly assigned to seniors.

The Elderly The characteristics of those typically referred to by the NP the elderly are negative in the sense that the referents are poor in health, wealth, and competence; in consequence they tend to impose a burden on society.

120

K. Allan et al.

( 14) 68% of participants claim that the elderly ‘are frail and fall more often’. (15) 68% of participants claim that the elderly ‘are often victims of mental and physical abuse’. (16) 66% of participants claim that the elderly ‘are unable to look after themselves and depend on others for help’. (17) 61% of participants claim that the elderly ‘are more susceptible to illnesses and disabilities’. (18) 60% of participants claim that the elderly ‘prove an increasing economic burden on society’. (19) 51% of participants claim that the elderly ‘are more likely to suffer from depression’. (20) 51% of participants claim that the elderly ‘are often lonely and bored’. (21) 45% of participants claim that the elderly ‘are more likely to cause road accidents’. (22) 42% of participants claim that the elderly ‘are poor’. (23) 40% of participants claim that the elderly ‘have no sex lives’. (24) 39% of participants claim that the elderly ‘are forgetful’. 68% of participants claim that the elderly are frail and fall more often than younger people, (14). This is entirely consistent with the generally bleak view of old age asserted of referent of the NP the elderly in all of (14)–(24). (14) identifies physical incompetence instead of mental weakness or poor health. But it is an important part of a comprehensive image of decrepitude. Interestingly the older old have the lowest figure on this and, unsurprisingly, the youngest group has the highest: ≤30 years 75%, 31–50 years 69%, 51–70 years 73%, ≥71 years 59%. Notably the younger old have a strongly pessimistic view. The same overall percentage of 68% of participants claim that the elderly are often victims of mental and physical abuse, (15). This is a somewhat shocking indictment of the society in which the weak and disadvantaged are mistreated, or at least are believed to be mistreated. So the elderly not only suffer from their own weakness but are bullied by those stronger and more powerful than them. When we look at the responses of the different age groups we see once again that the younger old hold a strongly pessimistic view, in (15) the strongest at 73%: ≤30 years 60%, 31–50 years 64%, 51–70 years 73%, ≥71 years 58%. Once again there is a large decrease among the older old. Is this because they are responsive to their actual experience whereas the younger old are in fearful anticipation? In (16) 66% of participants claim that the elderly are unable to look after themselves and depend on others for help. (16) identifies a burden that the elderly impose upon their families and society because of physical or mental incompetence. It is an outcome of (14), (17), and (18) and lays the elderly open to the kind of abuse referred to in (15). When we compare the attitudes of difference age groups, the younger ones believe in (16) far more strongly than the older old, for whom there is a massive drop-off: ≤30 years 75%, 31–50 years 80%, 51–70 years 71%, ≥71 years 52%. One gets the impression that the young and middle-aged are anticipating a burden, whereas the older old don’t believe they constitute such a burden.

Seniors, Older People, the Elderly, Oldies, and Old People…

121

61% of participants claim that the elderly are more susceptible to illnesses and disabilities. (17) once again goes to an aspect of the decrepitude that is perceived to be a concomitant of old age, an outcome of which is (16). Here the younger cohort (≤30–50 years) at 69% exceeds figures for the older cohort which once again dips with the older old: 51–70 years 63%, ≥71 years 55%. In (18) 60% of participants claim that the elderly prove an increasing economic burden on society. This is more specific than (16) by invoking the cause as (22). There is an additional factor in that the burden imposed by the elderly is not only present but increasing  – a result of the aging population as the result of human beings generally living longer than they used to. With (18) it is the middle-aged who take the most pessimistic view and yet again the older old are far less pessimistic: ≤30 years 60%, 31–50 years 74%, 51–70 years 65%, ≥71 years 51%. 51% of participants claim that the elderly are more likely to suffer from depression, (19). No bases are specified for the judgment rendered here, but one can reasonably presume that the decreased health, wealth, and quality of life along with risk of mental and physical abuse ascribed to the elderly would be commonly accepted causes leading to depression, cf. all of (14)–(18) and (20)–(24). Among the various age groups, the middle-aged are most pessimistic with respect to this characteristic, the young least (but they assign this trait principally to older people), and the older old less so than the younger old: ≤30 years 29% (but 33% to older people), 31–50 years 64%, 51–70 years 58%, ≥71 years 46%. Overall 51% of participants also claim that the elderly are often lonely and bored, (20), giving good reason for (19). Once again the middle-aged hold this negative view of the elderly most strongly, but this time the older old (whom one might expect to know best) believe it least of all: ≤30  years 49%, 31–50  years 64%, 51–70 years 54%, ≥71 years 45%. 45% of participants claim that the elderly are more likely to cause road accidents. (21) presents the elderly as a physical danger to themselves and to others, with ensuing mental anguish. Although some other characteristics ascribe to the elderly certain mental, economic and physical impositions on others, this is the only one that refers to the potential to injure severely and even fatally. All age groups assess this (standard deviation is just over 2%): ≤30 years 42%, 31–50 years 44%, 51–70 years 47%, ≥71 years 43%. In (22), 42% of participants claim that the elderly are poor. (22) identifies the personal circumstance, economic poverty, that potentially gives rise to (18), being an economic burden on others. (22) may evoke compassion in some participants but irritation in others. Among the various age groups, the young and the older old ascribed this property more to old people: ≤30 years 29% (but 35% to old people), 31–50 years 50%, 51–70 years 48%, ≥71 years 34% (but 39% to old people). Once again the middle-aged were most pessimistic. 40% of participants claim that the elderly have no sex lives, (23). Such research as there is suggests that ‘there is a raft of evidence suggesting that older people are sexually active’ (http://‌ theconversation.com/sex-desire-and-pleasure-in-later-life-­ australian-­womens-experiences-35,725); however, that is not what many people believe. Among the different age groups the middle-aged believe (23) most firmly,

122

K. Allan et al.

but all groups ascribe this with almost equal strength to old people: ≤30 years 29% (but 40% for old people), 31–50 years 50% (and 37% for old people), 51–70 years 42% (and 36% for old people), ≥71 years 35% (but 37% for old people). The standard deviation for the elderly is 9% whereas that for old people is 1.7%. In (24), 39% of participants claim that the elderly are forgetful, possibly marking the onset of dementia, detrimental to the individual and an imposition on those around them. As with (23), we find that this characteristic is almost equally ascribed to old people: ≤30 years 22% (but 44% for old people), 31–50 years 41% (and 37% for old people), 51–70 years 42% (and 36% for old people), ≥71 years 35% (but 37% for old people). The standard deviation for the elderly is 9.2% whereas that for old people is 3.7%. The young rank the elderly as (a) being frail and falling more often, (b) often victims of mental and physical abuse, (c) being more susceptible to illnesses and disabilities, (d) being unable to look after themselves and depending on others for help, (e) prove an increasing economic burden on society, (f) more likely to suffer from depression, (g) being often lonely and bored, (h) more likely to cause road accidents, (i, j) having no sex lives, and being forgetful, and (k) being poor. The middle-aged rank the elderly as (a) being unable to look after themselves and depending on others for help, (b) prove an increasing economic burden on society, (c, d) being frail and falling more often, and being more susceptible to illnesses and disabilities, (e, f, g) often victims of mental and physical abuse, and more likely to suffer from depression, and being often lonely and bored, (h, i) being poor, and having no sex lives, (j) more likely to cause road accidents, and (k) being forgetful. The younger old rank the elderly as (a, b) being frail and falling more often, and often victims of mental and physical abuse, (c) being unable to look after themselves and depending on others for help, (d) prove an increasing economic burden on society, (e) being more susceptible to illnesses and disabilities, (f) more likely to suffer from depression, (g) being often lonely and bored, (h) being poor, (i) more likely to cause road accidents, (j, k) having no sex lives, and being forgetful. The older old rank the elderly as (a, b) being frail and falling more often, and being unable to look after themselves and depending on others for help, (c) being more susceptible to illnesses and disabilities, (d, e) often victims of mental and physical abuse, and prove an increasing economic burden on society, (f) being often lonely and bored, (g) more likely to cause road accidents, (h, i, j) more likely to suffer from depression, and being poor, and having no sex lives, and (k) being forgetful. The characteristics ascribed to referents of the NP the elderly are negative in that the referents are purportedly poor both in health and wealth, they are weak and vulnerable to abuse, bored and depressed, forgetful and lonely, dangerous drivers, economically and physically dependent on others and, consequently, a burden on society.

Seniors, Older People, the Elderly, Oldies, and Old People…

123

Oldies and Old People The colloquial NP oldies was the least popular of the five NPs on offer on the SurveyMonkey questionnaire. The highest figure it attracted was 24% from the youngest group (≤30 years) for the positive attribute oldies are vibrant and full of purpose. The next highest score was 17% from the older old in respect of the negative attribute oldies are tight-fisted with money, perhaps because of attraction between colloquial oldies and the colloquial adjective tight-fisted. In fact this particular item, X are tight-fisted with money, favoured the following overall scores: old people 28%, the elderly 23%, older people 20% and both seniors and oldies 14% (with a standard deviation of 5.8% across the five NPs). This is the only attribute that favours the NP old people above the other NPs, albeit weakly. We might reasonably conclude that, like referents of the elderly, referents of old people are typically viewed somewhat negatively. This view is supported from some of the findings in §4, where certain age groups preferred old people to the elderly or else found them quite similarly applicable in respect of three statements: (25) Old people are forgetful ≤ 30 years 44%, 31–50 years 37%, 51–70 years 36%, ≥71 years 37%. (26) Old people have no sex lives ≤ 30 years 40%, 31–50 years 37%, 51–70 years 36%, ≥71 years 37%. (27) Old people are poor ≤ 30 years 35%, 31–50 years 22%, 51–70 years 37%, ≥71 years 39%. Although we have described (25)–(27) as negative they may alternatively express a certain sympathy for aged Australians whereas a statement like the elderly are more likely to cause road accidents admits no such possibility.

Conclusions This part of our research under the Australian Research Council funded project titled ‘The Cultural Model of Ageing: The Australian Conceptualisation of the Third Age’ sought to identify the characteristics of reference sets for five noun phrases commonly used to label older Australians, namely old people, older people, oldies, seniors, and the elderly. To our surprise we discovered that a majority of speakers who completed our SurveyMonkey questionnaire assigned rather specific characteristics to the referents of seniors, older people and the elderly, were much less decisive about old people, and not at all decisive about oldies. From the ≤30 year olds (55 participants) there was 0% agreement that oldies are reliable, oldies are victims of mental and physical abuse, oldies benefit the workforce through their experience, oldies are conscious about their health and lead a healthy lifestyle, oldies contribute economically to society, oldies are unable to look after themselves and depend on others for help (there was also 0% for older people).

124

K. Allan et al.

This reinforces the conclusion that referents of the NP oldies are not ascribed firm characteristics – a point further strengthened in the next paragraph. From the middle-aged (55 participants) there was 0% agreement that oldies prove an increasing burden on society, oldies are more likely to suffer from depression, oldies / old people serve as role models and are respected members of society, the elderly are wealthy and like to spend their money, oldies / older people are often victims of mental and physical abuse, oldies / seniors are more susceptible to illnesses and disabilities, oldies / the elderly benefit the workforce through their experience, the elderly contribute economically to society and the elderly lead an involved and active life. There was 0% agreement from younger old (313 participants) that old people like to travel. The older old (231 participants) had 0% agreement on nothing. Leaving oldies aside, there are the positive ascriptions to referents of seniors and older people and negative ascriptions to referents of the elderly and, with much less certainty, to old people. The referents of seniors were generally attributed with the positive personal characteristics of being outgoing, healthy, active, wealthy, and happy. Thus: (28) The typical4 denotatum of senior is a person of around 60 years or more who has the personal characteristics of being outgoing, healthy, active, wealthy, and happy. The referents of the NP older people were generally attributed with the positive characteristics of being socially responsible benefactors who have wisdom and experience to share with others within and outside the family. Thus: (29) The typical denotatum of older person is around 60 years or more who is a socially responsible benefactor possessing wisdom and experience to share with others within and outside the family, and who is likely to be conservative. By contrast, referents of the NP the elderly are perceived negatively: they have poor physical and mental health, they are weak and vulnerable to abuse, bored and depressed, forgetful and lonely, they are a danger on the road, they are poor and therefore economically as well as physically dependent on others and, consequently, the elderly are a burden on society. Thus: (30) The typical denotatum of elderly person is around 60 years or more who is in poor physical and mental health, weak and vulnerable to abuse, bored and depressed, forgetful and lonely, a potential danger on the road, poor in wealth and therefore both economically and physically dependent on others and, consequently, a burden on society.

4  On the use here of ‘typical’ instead of ‘stereotypical’, see the discussion in §1 where we say that meaning is ‘a minimum set of stereotypical facts about the typical denotatum’.

Seniors, Older People, the Elderly, Oldies, and Old People…

125

Referents of the NP old people were not strongly characterised in the manner of those of the NPs seniors, older people, and the elderly; but they are attributed – perhaps somewhat sympathetically – with the mildly negative characteristics of being tight-fisted, poor, forgetful, and having no sex life. Thus: (31) The typical denotatum of old person is around 60 years or more who is poor, tight-fisted, forgetful, and without a sex-life. There, but by the grace of God go we.

References Allan, K. (2001). Natural language semantics. Oxford, UK/Malden, MA: Blackwell. Allan, K. (2006). Stereotype semantics. In E.  K. Brown (Ed.), Encyclopedia of languages and linguistics (Vol. 14, 2nd ed.) (pp. 153–155). Oxford, UK: Elsevier. Allan, K. (2011). Graded salience: Probabilistic meanings in the lexicon. In K. M. Jaszczolt & K. Allan (Eds.), Salience and defaults in utterance processing (pp. 165–187). Berlin, Germany/ Boston, MA: De Gruyter. Allan, K. (2013). Referring to 'what counts as the referent': a view from linguistics. In A. Capone, F.  L. Piparo, & M.  Carapezza (Eds.), Perspectives on linguistic pragmatics (pp.  263–284). Cham, Switzerland: Springer. https://doi.org/10.1007/978-3-319-01014-4_10 Battig, W. F., & Montague, W. E. (1969). Category norms for verbal items in 56 categories: A replication and extension of the Connecticut category norms. Journal of Experimental Psychology Monograph, 80(1): 1–46. https://doi.org/10.1037/h0027577. Fillmore, C. J. (1975). An alternative to checklist theories of meaning. In C. Cogen et al. (Eds.), Proceedings of the first annual meeting of the Berkeley Linguistics Society (pp.  123–131). Berkeley, CA: Berkeley Linguistics Society. Labov, W. (1978). Denotational structure. In D. Farkas, W. M. Jacobsen, & K. W. Todrys (Eds.), Papers from the Parasession on the Lexicon (pp. 220–260). Chicago, IL: Chicago Linguistics Society. Rowe, J.  W., & Kahn, R.  L. (1987). Human aging: Usual and successful. Science, 237(4811), 143–149. https://doi.org/10.1126/science.3299702

How to Be Impolite (or Worse) in an Artificial Auxiliary Language Alan Reed Libert

There is a very large group of languages which have been consciously created by one or more people, unlike natural languages such as English and Chinese. These languages are known as artificial languages (henceforth ALs) or constructed languages. Many ALs, such as Esperanto, were created as a means of international communication, i.e. as auxiliary languages, and it is these languages which will looked at in this paper.1 Artificial auxiliary languages are instances of conscious linguistic engineering: in some cases a natural language is changed, as in various ALs which are simplified versions of Latin (e.g. Latino sine Flexione); in other cases an AL is created from parts of several or many natural languages (e.g. Esperanto) or from scratch. However, in many cases ALs also involve social and/or psychological engineering. A motive behind the creation of Esperanto and many similar languages was to increase international understanding; we might therefore expect encouragement of positive messages in them and discouragement of negative or rude ones, either at an official or an unofficial level. An example of this is in Hogben’s (1943) book on Interglossa; he says (p. 106), “The need for the strong imperative will be rare, except in history books. An international auxiliary of peaceful communication is not for generals or for conversation with the cat”. Earlier on the same page he refers to this imperative as “the impolite imperative”; it involves the “polite imperative” (p. 105) with peti tu ‘please (you)’ omitted. Nevertheless, Hogben gave the impolite imperative, allowing for the possibility of being impolite in this way. 1  There are various other kinds of artificial languages, e.g. those created in connection with works of fiction (such as Klingon).

A. R. Libert (*) University of Newcastle, Callaghan, NSW, Australia e-mail: [email protected] © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 F. Macagno, A. Capone (eds.), Inquiries in Philosophical Pragmatics, Perspectives in Pragmatics, Philosophy & Psychology 28, https://doi.org/10.1007/978-3-030-56696-8_7

127

128

A. R. Libert

The basic questions in this research are the following: 1. Are there the means to be linguistically impolite or worse (e.g. racist) in ALs in general or in a particular AL? 2. Is being rude (racist, etc.) explicitly forbidden or discouraged in some ALs? 3. If so, are AL users still rude, etc.? 4. Is linguistic impoliteness different in ALs than in natural languages? The present short survey will not give definitive answers to all of these, but will give some indications in relation to them. There may be a tension between wanting people to be nice in one’s AL (social engineering) and wanting an AL to be a “real” language. See the quotations below: Swearwords and profanities: An [sic] useful part of language, as well as necessary for translation. Unless provided in sufficient quantity and quality, they will be imported or coined without care for the original vision.2 If your language can’t express the opinions of your enemies, if they can’t say what they want to tear you down, then you’re trying to create a world where they can’t express their thoughts. I can’t find any good racial slurs in Esperanto or Lojban, likely because the people who use these languages aren’t the sort to use them- but lacking expressivity means the language is crap. If you make a utopian language, they’ll use it in utopia- so, nowhere at all.3

In the a priori (i.e. built from scratch) language aUI, it may be impossible to create and hence utter ethnic slurs or other derogatory words, at least in principle. Consider the passages below: When Dr. Weilgart served at Xavier University in New Orleans, a Southerner told him: ‘Washington Carver may have been a gentleman and a scholar: in my book he is still a Nigger.’ Analyzed into aUI, it would read: Dr. Sarver may have been a wise-know-man (nU-gUw-u) and a good man (r-u), he is still a bad-black-man (yr-ybi-u) or: mal-tropic-man (yr-ia-u).’ Invective insinuations become verifiable self-contradictions. (Weilgart, 1979: lii) While ‘I’ reserve good words for myself and my friends, I wrap my enemies in loaded invectives, calling them Niggers, Kikes or Gooks. This means that evil is implicit in their name. If he ‘is’ a Kike, I need not prove that he is bad. I once heard in the South the question: ‘Is your lover a Nigger?’ (If asked [in aUI]: ‘is he a black monster?’ the girl could have answered, ‘One of his ancestors came from the tropics, he is handsome & good.’ or to that explicit question, she could even have answered ‘No.’) In aUI each invective must start with honest ‘yr-’ = no-good, bad. Then in transparent truth I could have to prove & verify, why he is yr-u = a bad man. I can easily command ‘Kill the Kikes,’ because they are bad, & the command rhymes in alliteration. But in aUI each command, so necessary in war, contains ‘r’ = good. Kill the Kikes = vyrov yru = make-non-live (good deed!) those ‘no-good people’! Then you can ask me: ‘Why is it good to kill them? Why are they no good so that it is good to kill them? […] Thus in aUI Invectives & Commands become statements of evil &good, verifiable both. We speak transparent truth. (Weilgart, 1979: xlv–xlvi)

2  https://gdoc.pub/doc/e/2PACX-1vT4-VIep_uTa4Np7Kz66MYhdHf-bRHVgsoRcOEhXA4i Xvygvh3nMdhKlKvCThyZrUHUj48nlI08_Vcw. 3  https://ask.slashdot.org/story/15/04/08/162231/ask-slashdot-what-would-a-constructedlanguage-have-to-be-to-replace-english.

How to Be Impolite (or Worse) in an Artificial Auxiliary Language

129

If I understand correctly, because aUI words must be made up of one or more semantic atom-words, the only way of creating an aUI equivalent to a slur against Black people is to say something like ‘bad-black-person’, and one might then wonder why he is bad. However, there can be invectives in aUI: Weilgart (1979: 45) says, “INVECTIVES start explicitly with Yr- (bad-)”. One might imagine an extreme social engineering language in which one could not even use invectives (because they would not exist). As an a priori language, aUI is (at least on the surface) structurally quite different from natural languages such as English. Let us now turn to an AL which is built mainly from natural languages (such ALs are called a posteriori languages), Esperanto. There certainly are insults in Esperanto, as can be seen from the list of words in Corsetti and Lowenstein (2009), some of which are given below; note that all of these were “actually used in spoken Esperanto” (p.  160) in the period 1985–1990:4 bugren! - (bugriĝen) bugger off! (p. 175) [bugri ‘to bugger’] estu almozulo! - become a beggar! (p. 176) estu fikota! - get fucked!, get lost! (p. 176) estu mortigota! - get killed! get lost! (p. 176) fak-delegito!  - specialized delegate! (euphemism for «fekaĵo» [‘shit’] refers to a system of representatives within the Esperanto movement) (p. 177) fekero - lump of shit, stupid person (p. 177) fekfikanto - shit fucker, sodomist (p. 177) fermu vian faŭkon! - shut your gob (faŭko is usually only used of an animal’s mouth) (p. 177) fik-klaŭno - fucking clown, idiot who wants to fuck all the time (p. 177) forfikiĝu! (forfiku!) - fuck off! (p. 178) [fiki ‘to fuck’] forfurzu! - fart off! (p. 178) forpisu! - piss off! (p. 178) iru al la diablo! - go to the devil! (p. 179) iru feki! - go and shit! (p. 179) iru fiki! - go and fuck! (p. 179) kaco! - prick! (p. 180) piĉen! - cuntwards, get lost! (p. 185) rasisto ‘racist’ (p. 187) senkaculo - man with no prick, useless person (p. 188) senkaĉulo - person with no porridge (euphemism for «senkaculo») (p. 188) Most of these would not be unusual in natural languages, but there are some other Esperanto insults that are unique to it: Corsetti & Lowenstein (2009):

4  The glosses and explanations are Corsetti and Lowensteins’s; my additional explanations are in square brackets.

130

A. R. Libert

blinkenbergo - person who opposes Esperanto (after delegate in the Unesco General Assembly in 1954) (p. 174) bofrontido - traitor (after De Beaufront, an early speaker of Esperanto, who became a supporter of Ido5) (p. 174) volapugisto - stupid person (humorous combination of «volapukisto» … and «pugo» = arse) (p. 190) [volapukaĵo - Volapük, something incomprehensible p. 1916] The same may be true of bonantagulo; below is part of the Wiktionary entry for this word: bonantagulo […] Etymology bonan tagon (“good day”) + -ulo (“person’) […] 1. (slang, derogatory) someone whose Esperanto is limited to a few basic phrases like bonan tagon (“good day’)7 There is also supposedly a more substantial difference between Esperanto insults and those of natural languages: Readers will notice the complete absence [among the terms collected by them] of insults based on national stereotypes. Again, this has its basis in the values of the Esperanto speaking community, whose members usually take a positive interest in other countries and peoples, and feel obliged to assert the equality of minority groups. There is also the practical fact that when using Esperanto, a speaker or writer is likely to be addressing a genuinely international public, so it is neither socially acceptable nor possible in practice to use the equivalents of such words as «Yank» or «Polack». Another factor explains the paucity of insults aimed at people who do not fit in with the reigning norms. Esperantists themselves are aware of belonging to a deviant social group and tend to be more tolerant than the average member of society. (Corsetti & Lowenstein, 2009: 165–6)

However, humans being what they are, I would not be surprised if somewhere one might find racist (etc.) terms used even by Esperantists. In any case, one might need such terms to translate fiction, as they could occur in the speech of some characters. My guess is that, if we dig deeply enough, and far enough back in time, we could find racist terms in Esperanto (and other ALs of a similar age). This assertion is supported by the following fact: although racist terms have been seen as unacceptable in normal society for some time, this is not true of homophobic terms – it is only within the last couple decades that they have become unacceptable. We should therefore be able to find homophobic terms in fairly recent Esperanto, and this turns out to be the case. For example, geja ‘gay’ is not homophobic, but a related term, gejaĉo, is, as shown by the Wiktionary entry for the latter term: 5  Ido is a revised version of Esperanto. There was hostility between the Esperanto movement and the Ido movement. 6  Volapük is an artificial auxiliary language created slightly before Esperanto. Its morphology is rather complicated. 7  https://en.wiktionary.org/wiki/bonantagulo#Esperanto.

How to Be Impolite (or Worse) in an Artificial Auxiliary Language

131

gejaĉo […] Etymology From geja (“gay”, adjective) + -aĉ- (“pejorative”) + -o […] 1. (offensive, derogatory) faggot8 This entry then gives an example of the use of this word involving a remark apparently made by Alessandra Mussolini, “Pli bone faŝisto ol gejaĉo” (‘Better a fascist than a faggot’). Note, however, that this is a translated remark, i.e. an Esperantist did not say this, and, in fact, it seems difficult to find examples of its use by Esperantists, perhaps because geja ‘gay’ itself is a relatively new term (as is English gay in the relevant sense), and, as I have stated, homophobic language is no longer seen as acceptable. There is also the term pederasto, part of whose Wiktionary entry is given below: pederasto […] 1. pederast […] 2. (offensive, derogatory) fag […] Synonym: gejaĉo 3. (archaic) homosexual9 One of the examples of its use in the second sense is the following: “Ja mi ne estas pederasto. Kaj estas rekomendinde al chiu pederasto memori tion.” ‘I sure am no fag. And it would be best if every fag remembered that.’ Since this appeared in an Esperanto newsgroup (soc.culture.esperanto), it seems that it is not a translation, i.e. it is a homophobic statement made in Esperanto by an Esperantist. Among other derogatory terms in Esperanto is viringo: viringo […] Etymology From viro (“man”) + -ingo (“holder”), a pejorative pun on virino [‘woman’] […] 1. (offensive, derogatory) promiscuous woman; cumdumpster, slut, whore 2. (offensive, derogatory) objectionable woman; bitch, cunt, slut10 Note that many or most ALs may lack insults, etc. not because of any desire by the designer to eliminate or restrict them, but because their vocabulary was not fully developed, and/or because they were not used enough for such terms to develop. Of course one does not need to use racial slurs to say or write racist remarks in an artificial language (or any other language); consider the following situation: En la numero el julio-aŭgusto 2016, diversaj anoj de SAT akuzis artikolon en la gazeto fare de Roger Condon, nome “Zamenhof pravis,” pri antisemitisma kalumnio. Specife, ili kritikis sekcion, kiu kulpigis la judojn pri la Holokaŭsto:

 https://en.wiktionary.org/wiki/geja%C4%89o.  https://en.wiktionary.org/wiki/pederasto#Esperanto. 10  https://en.wiktionary.org/wiki/viringo#Esperanto. 8 9

132

A. R. Libert

“Ĉu ne la judoj mem, per siaj kolerema konduto kaj ekstremaj troigoj, startigis la nazian maŝinon, kiu tuj decidos pri ilia ekstermado en la naziaj koncentrejoj?” — Roger Condon en Sennaciulo, marto-aprilo 2016, paĝo 711 (‘In the July-August 2016 issue [of the Esperanto journal Sennaciulo], various members of the SAT [Sennacieca Asocio Tutmonda ‘World Non-National Association’] accused an article in the magazine written by Roger Condon entitled “Zamenhof was right” of anti-­Semitic slander. Specifically, they criticized the section which blamed the Jews for the Holocaust: “Did not the Jews themselves, by their testy behavior and extreme exaggerations, start the Nazi machine which would immediately decide on their extermination in the Nazi concentration camps?” — Roger Condon in Sennaciulo, March-April 2016, page 7’)

One might also note that in Nazi Germany itself Esperanto was used in support of Nazi ideology (although eventually many Esperantists and the Esperanto movement suffered greatly under the Nazis): In the October 1933 Germana Esperantisto six pages were devoted to an Esperanto-­ language translation of a speech by Hitler, and the Esperanto Union of Saxon Teachers published a four-page leaflet, La Nova Germanlando (The New Germania), which was distributed to 70 countries in 10,000 copies. Using statistics in an attempt to prove that Jews dominated German public life, this pamphlet, with its clear anti-Semitic bias, must be numbered among the most odious publications ever produced in the language of Zamenhof. (Lins, 2016: 104)

Conclusion In spite of the efforts of some AL designers, it is possible to be rude, racist, homophobic, etc. in artificial auxiliary languages. In fact, it might be necessary to allow this to happen if such languages are to become usable human languages: people are often rude, and sometimes racist; while one might regret this, it is part of human behavior, and cannot easily be eliminated in a functioning language. However, racist language may occur less in artificial languages than in natural ones, although this is probably due to the type of people who decide to learn these languages rather than anything in the languages themselves.

References Corsetti, R., & Lowenstein, A. (2009). Esperantists do it internationally: Insults and similar words in the international language Esperanto. Interlinguistica Tartuensis, 9, 159–191. Hogben, L. (1943). Interglossa. Harmondsworth, UK: Penguin Books. Lins, U. (2016). Dangerous language  — Esperanto under Hitler and Stalin. (translated by H. Tonkin). London, UK: Palgrave Macmillan. Weilgart, J. W. (1979). aUI: The language of space (4th ed.). Decorah, IA: Cosmic Communication.

11

 https://eo.wikipedia.org/wiki/Sennaciulo#Akuzo_pri_antisemitisma_kalumnio.

Pragmemes at the Market Place Alessandro Capone

Introduction Should we consider language as a set of formal rules that allow for interpretation or should we also consider language as a social artefact that is embedded in culture? In this paper, I give some answers to these important questions and I attempt to integrate linguistics as formal rules applicable to the sentential level with a kind of linguistics that is attentive to societal phenomena. In particular, I deal with pragmatic interpretation that goes beyond what is literally and conventionally said and normally requires some kind of expansion (for example, my niece sends me the photo of a jumper and says ‘Do you like it?’. Part of pragmatic competence is to understand whether she is asking for advice, should she buy it, or whether she wants to give me a present for which it is important that I say that I like it, in which case she or her mother will buy it). Contextual information is decisive in deciding how to take this inference. Pragmatics is interested in the interpretation of utterances in their respective contexts and in a cultural setting. As Capone (2010) says, Once we leave aside the social dimension of language, what language is for, its various and multiple functions, the relationship between the expression of thought and the social and institutional facts about society, the propagation and transmission of culture through language, we are left with a very meagre, impoverished and surely aseptic vision of linguistics (Capone, 2010, 2861).

We assume that utterances in linguistic interaction have both an intrinsic intentionality, in that they express states of the mind, and a derived intentionality in that they are linguistic objects that express the intentionality of the states of mind that precede such utterances and motivate them (see Haugh and Jaszczolt (2012), which this view is an elaboration of). On this view, intentionality is both a dimension that is A. Capone (*) Università di Messina, Messina, Italy © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 F. Macagno, A. Capone (eds.), Inquiries in Philosophical Pragmatics, Perspectives in Pragmatics, Philosophy & Psychology 28, https://doi.org/10.1007/978-3-030-56696-8_8

133

134

A. Capone

expressed by an utterance and something that motivates the proffering of the utterance, without which the message could not have its reason d’être. The context is what reveals large portions of meaning and has, according to Gross (2001), an illuminating function, in so far as it allows hearers to interpret deictic elements or relationships such as, for example, the argumentative but even the syntactic relations of the text, as argued at a later point in this paper through an analysis of sentential fragments. The context can be considered as the set of objects, participants and events that are placed or situated around or inside the event in question, the communicative event, that is to say, the event that creates or is created by the utterance (see Hymes, 1962) and that normally provides the referents of the deictic elements (e.g. this; that table) through a procedure that may require an indexical or demonstrative gesture, as we point the index finger towards the object or even use a look. The communicative event is created by the act of telling a story or, in general, by a linguistic act/linguistic game intended not only as a single utterance, but as a series of utterances related to the same communicative function. We may call this, following illuminating terminology by van Dijk (1980), who coined the term ‘macrostructure’, the macro speech act. The macro speech act is inferred, normally, by deleting information that is not crucially important and by subordinating information that cannot be deleted to propositions that can subsume it (Van Dijk, 1980), but also by keeping track of conventional ways to arrange utterances in sequences with a certain structure, such that, given the structure, one can recognize the main goal underlying it (Schegloff, 1988; Haugh, 2015). If texts are seen in this way, the notion of illocutionary action must be related to sequential effects and I mainly agree with Haugh and Jaszczolt (2012) that the overall intention at the level of implicature and explicature should take into account sequential information, textual plans, and contextual effects. The illocutionary force, in such a view, cannot be calculated utterance by utterance, but must be related to the inferred illocutionary act that is performed by using a structured sequence of utterances. Such intentions are called by Haugh and Jaszczolt (2012) ‘higher order intentions’. A lot of what I say in this paper is compatible with their view and capitalizes on it. A communicative event’s occurring presupposes a communicative competence that goes beyond knowledge of syntax and that, normally, consists in the ability to follow cultural and textual rules that allow the language user to connect the communicative event to a wider textual and cultural context. As Mey says, a propos of communicative competence, in addition to being ratified by the addressee, “the act will have to be executed in accordance with society’s customs and practices, including all the written and unwritten rules for how to comport oneself in a particular society and how to interact with one’s partners” (Mey, 2016a, 32). Normally, events are considered things by philosophers given that they occupy a portion of space and time, which cannot be occupied at the same time by a different event (see Higginbotham, 2000); but in a socio-pragmatic paper like this one, it may be useful to consider a communicative event as something that is constructed by a community of speakers/language users by resorting to socio-pragmatic practices and rules, as something that exists because of social intentionality and not only

Pragmemes at the Market Place

135

because of individual intentionality (as Mey, 2010 says, communicative events like pragmemes require uptake and collaboration). While I can run on my own and this fact constitutes an individual event, even if no participation by other people is required, a communicative event always presupposes a community of speakers, at least a dyad of participants, the collaborative construction of utterances or of actions, etc. Normally, however, the context does not end with the inclusion of referents through the use of interpretative procedures (which Goffman, 1981 calls ‘the alternative channel’), but includes the culture in which a given communicative event is inserted and which serves to resolve/dissolve important interpretative nodes. Culture enriches the understanding of utterances through conventions of use, which allow the language user/interpreter to construct a communicative event in a global sense not only from the bottom to the top, but, on the contrary, from top to bottom (see Recanati, 2004; Kecskes, 2010, Kecskes, 2008, on the contrary, tries to mitigate this view by stressing the importance of linguistic semantics in starting the interpretation process). Such norms or other rules underlie what the philosopher Wittgenstein would call ‘language games’. Among the possible enrichments of the semantic and syntactic unit (the sentence) that serve to construct (or collaboratively construct or reconstruct) the language game, we could have the determination of the illocutionary force of a sentence; other possible enrichments could be the result of a range of interpretive options that require the comparison of options so that the interpretative choice will be responsive to the cultural background in which a communicative (linguistic) event can be seen. The considerations on pragmemes can be seen as the basis for our analysis of language use at the market places. Such uses, in fact, are situated, both in the situation of utterance and in the cultural background which provides a set of norms for interpretation, as well as a set of social values (see Wong, 2010) which determines or motivates the behaviour of the language user both in linguistic production and in interpretation. Values are the things that can provide an ethical dimension to interpretation and establish the limits which interpretation cannot transcend. This paper can be seen as an exploration of language use seen from the perspective of vendors and customers at the market place. There was a period in which sellers at the market could capitalize on crowds of clients and, therefore, had to find ways to attract their attention, advertise their products and finally sell them. The market in those times was a considerable resource of wealth and, thus, vendors had to take advantage of their position and exploit linguistic resources in order to adapt (in the sense of Verschueren, 1999 and Mey, 2001) to the changing circumstances of life. Language use could be seen as a way to advance the social status of vendors and to procure them wealth, which then would have allowed their families to educate their children and exploit new opportunities for social mobility. This paper, thus, can be seen in the light of socio-pragmatics, from which it derives its reason d’être, and can introduce us to the notion of emancipatory language. Linguistic resources can be at the disposal of people who want to pursue social climbing and have to develop specialized techniques for gleaning the advantages of their work.

136

A. Capone

Structure of the Paper In this paper, I deal with socio-pragmatics and pragmemes and propose that we need to take into account societal and cultural considerations in order to develop a persuasive theory about utterance meaning and language use. In particular, I deal with pragmemes and propose that they depart considerably from speech act theory, which is more or less a kind of armchair linguistics, and they can be combined with the notion of language games (Wittgenstein, 1953) and with the notion of the triadic structure of language (Wong, 2010). In this view, culture plays a predominant role. As Wong says: To attain a more holistic picture of language, one would need to see beyond form and even meaning to consider the cultural context of the language (or pragmemes) under study. This is because, as numerous studies have argued persuasively, people use language to express not only meaning but, consciously or subconsciously, culture as well – values, attitudes, biases, etc. A pragmeme is therefore best viewed as a “culturally situated speech act” and when we study pragmemes, we should ideally study them with respect to the culture in which they are used (Wong, 2010, 2942).

In this paper, I explain why culture plays a predominant role in the pragmeme and should be a component of the definition of this type of unit of language use. I provide certain examples of pragmemes which require the notion of pragmatic competence and clearly show that a certain cultural dimension is important in determining conventions of use. Then I consider a special kind of pragmemes: pragmemes at the market place. Pragmemes at the market place can be seen in the light of Verschueren’s and Mey’s (2001) ideas on adaptability. Language users develop ways to adapt to reality (and to the changing circumstances they are confronted with) by making use of pragmatic acts that exploit contextual considerations and, given that they are to be seen as responses to the changing circumstances of life, may establish a path of language use and norms that constitute precedents from which subsequent language users can inherit ways to adapt to the circumstances of life.1 Sellers at the market develop techniques for shouting, surprising their clients, attracting attention, advertising their products through imaginative and poetic ways. These clearly require techniques, social practices and rules of use that are very different from those involved in other forms of communication in that they are to be seen as responses to specific circumstances, specific purposes, and specific needs to adapt to the new

 Capone (2016b) writes:

1

The pragmeme is particularly useful, for these cases, as it is a case of language use that creates a precedent. To my knowledge, the notion of utterance, language game, or speech act is not flexible enough to account for this part of critical discourse analysis – how new norms are negotiated and emerge out of interaction. The meaning of an utterance, therefore, cannot be isolated from what can be done with it and from its potential future uses. (Capone 2016b, xix).

Pragmemes at the Market Place

137

circumstances of life. I propose that the poetic function (Jakobson, 1960)2 in these language games merges with other functions (e.g. the phatic, the argumentative and the conative function). Both the poetic and the phatic function are subordinated to the conative function. I consider cases of expansion, in which knowledge of the cultural background and the situational context can guide interpretation and, in particular, the calculation of syntactic relations, which is made particularly difficult because we are confronted with fragments. I then consider pragmemes and deontic modality, arguing that the cultural context ensures that the right interpretation is arrived at through cultural ethical norms (readers who belong to different cultures notice that literal translations of these utterances, separated from inferences due to cultural norms, produce no similar interpretations in their languages). Next, I consider the pragmeme ‘selling fish at the market place’. We can see that a particular pragmeme has its own norms that are dedicated to it and determined by the situational needs of the seller. The seller specializes in selling this product (tuna fish and fish in general) and, thus, best knows how to advertise it. Why should there be these forms of specialization? Probably the answer is that the seller knows which questions the client has for him and he is prepared to answer them in advance, having developed poetic techniques for this task. I finally provide a series of common features which pragmemes of selling things at the market place display. We may consider them the topoi of the specific pragmemes. I am aware that there is a lot more work to be done, but this is not always easy because this kind of language use is gradually being lost due to changes in the structure of society. The data collected by anthropologists do not suffice in any way because they are focused on folklore more than on language use. The attempt to focus on the core features of pragmemes or, in the case in point, the pragmemes of selling fish at the market place is thwarted by the scarcity of the data. 1. Pragmemes To say that pragmemes are linguistic acts in a context is like proposing an old idea with new terms. But this is, more or less, what is said in the literature (e.g. Mey 2002 or Capone, 2005). Capone tried to solve these problems in the Introduction to Allan, Capone, Kecskes (Allan et al., 2016), but it seems that an exhaustive answer to these problems is still to be found. To say that the context is something that has a top-down influence (from top to bottom) offers only a few partial answers to the problems I raised/mentioned. It is true that the context is active even before the utterance is proffered (in that it has a predictive power) and the hearer sometimes already knows in advance which linguistic act will be proffered and must not make 2  According to Jakobson (1960), the poetic function is an integral part of the study of linguistics. He says:

Poetics deals with the problems of verbal structure, just as the analysis of painting is concerned with pictorial structure. Since linguistics is the global science of verbal structure, poetics may be regarded as an integral part of linguistics (1960, p. 350).

138

A. Capone

any interpretative effort (in the sense of deriving the meaning and function of the sentence from the constituents and their syntactic composition); so much so that Montminy (2010) called Mey’s approach ‘anti-intentionalist’, a position which would certainly deserve discussion. Sometimes we know very well what will be said later and what an utterance, correctly anticipated, means, In fact, we may call this phenomenon noted by Mey (2001) based on the expectations raised by scripts ‘anticipation’. But for many statements, the interpretative process does not proceed this way and it is worth constructing the meaning from the bottom up to the top; this does not exclude an influence from the context to the bottom. Indeed it is reasonable to think that the standard way to proceed is this: in parallel, both the function of interpreting the constituents and their syntactic composition takes place, as well as the function of contextualizing the syntactic constituents and the global result of the syntactic composition. This view is compatible with both the socio-pragmatic view of Mey (2001) and the cognitivist view of Carston (2002) and Jasczolt (Jaszczolt, 2016). Furthermore, it is an attempt to reconcile Mey’s ideas with Kecskes’, as the latter author holds a special place for conventionality and compositionality. One of the features that characterizes pragmemes is that they are situated. This feature, of course, allows for top-down effects that percolate from the context in the presence of contextual clues that allow us to connect the text with the context; see Gumperz, 2001 and Dascal, 2003. As Mey says: Speech acts, in order to be effective, have to be situated. That is to say, they both rely on, and actively create, the situation in which they are realized. Thus, a situated speech act comes close to what has been called a speech event in ethnographic and anthropological studies (Bauman and Sherzer, 1974): speech as centered on an institutionalized social activity of a certain kind, such as teaching, visiting a doctor’s office, participating in a tea-­ ceremony, and so on. In all such activities, speech is, in a way, prescribed: only certain utterances can be expected and will thus be acceptable; conversely, the participants in the situation, by their acceptance of their own and others’ utterances, establish and reaffirm the social situation in which the utterances are uttered and in which they find themselves as utterers. (Mey, 2001: 219)

But this, again, does not say much about pragmemes, in addition to saying from the very beginning that what is interpreted, in fact, is not only the result of the semantic composition of the sentence, but the result of the composition of various utterances that constitute a text in a language game. The language game, at least in part, consists in putting together utterances that, in the end, function as a single one, are coherent and have a very precise textual order (the macro speech act theorized by Van Dijk, 1980). The language game and the pragmeme are very similar. But what really identifies the pragmeme is that it is a triadic type of unit and consists of a unit (a small piece) of text (a linguistic game), a linguistic and extra-linguistic context (the set of textual sequences that surround the utterance and the objects present in the environment to which the deictic and pronominal elements of the discourse are anchored through a pragmatic interpretative work) and, finally, a cultural context, which serves to enrich and often fully determines the meaning of a phrase (or a sentential fragment) or an utterance (see Wong, 2010). We often happen to notice that the same sentence in different languages ​​has different connotations or

Pragmemes at the Market Place

139

serves different uses, although the phrasal meaning is basically the same or very similar (See the analysis of the term ‘police’ by Kecskes, Messina lectures 2017). The pragmeme often coincides with those units Kecskes (2010) called ‘situation-­ bound’ (related to the situation), as it plays a unique role in the context and cannot work independently of the influence of the situation, which shapes its meaning.3 But pragmemes are not only linked to particular situations (ritual events or recognizable communicative events as they occur in various situations), but are linked to the cultural context. If it were not for Anna Wierzbicka (2006) and her school, perhaps we would not even hear about cultures and the fact that different cultures may require different speech acts (Capone, 2005 talked about the pragmeme ‘recommending’ and explained that what might seem like the same action has different implications in Italian and English culture).4 The role played by culture in interpretation has been highlighted by philosophers and linguists of the past, such as Plato and Rask, although they never provided such in depth analyses. Proceeding from the notion of pragmeme, we arrived at the notion that different languages ​​may require different lists of linguistic acts, although they certainly share a sub-section of the list and there are possibly speech acts that justify a claim to universality  - and this news seems to show that the notion does some work and is useful in illuminating the differences among languages: linguistics deals not only with the universal traits of language, as stressed by Chomsky’s research, but also with the differences between languages which ​​ attest to a remarkable conceptual richness. This seems to express well the Searlian concept (Searle, 1979) that we will always find the semantic resources to say what we want to say. 2. Examples of pragmemes An interesting example of a pragmeme is the following: 1. Chi mi talii? (language used: Sicilian from Palermo). (lit. What do you look at me). (non-lit. Why are you looking at me?) In context, the sentence could take on the meaning of ‘Why are you looking at me?’. But it is not this (minimal) transformation that elicits our attention. The real problem is that an utterance of this kind is often considered (or provides) a reason to quarrel. It is true that the addressee can always answer something like ‘Why, is it forbidden to watch?’, but this answer would not remove the reason for the potential 3  Kecskes (2010, 2891) says that they are “formulaic expressions called ‘situation-bound utterances’ (...)”. They “are highly conventionalized, prefabricated pragmatic units whose occurrences are tied to standardised communicative situations” (also see Coulmas, 1981). 4  To recommend someone, in English, is to praise that person for his objective qualities and to describe those qualities so that some use can be made of them. In Italian, ‘raccomandare’ is to advance a person promoting him or her to a certain job, regardless of that person’s qualities. That amounts to using an illegitimate influence on job procedures.

140

A. Capone

quarrel/conflict (in fact, it would have a confrontational import). The problem is that while the hearer certainly has the right to look, the duration of the look or the way the person was looked at can be offensive. This article on pragmemes at the market place is basically a chapter in socio-­ pragmatics. We are interested in studying pragmemes in relation to a society and its culture. The pragmemes are, in fact, societal language units, not only because certain social rules provide for orderly participation in the interaction, but also because the pragmemes are a form of social actions with clear and intentional effects. Social intentionality which is expressed by pragmemes is something that goes well beyond individual intentionality (see Capone, 2005), although favoured by contextual effects; individual intentionality may be projected by idiosyncratic statements. In the case of pragmemes, intentionality is projected by conventions of use, by public use of an utterance in context, by preceding praxis (what would be called ‘precedents’ in the context of the law), by the fact that the sentence has a public dimension. Notice how an offense is aggravated by its public dimension. What we could withdraw as a joke, if said in private, can no longer be treated in the same way, when it is spoken in public. It is the conventions of use that are sensitive to the public dimension of the utterance and that can transform a friend into our worst enemy, even if he makes a gaffe, even if his/her behaviour was not completely intentional. The intention will be attributed to the public dimension of the language game. Consider a person who attends a funeral to praise a deceased friend (Capone, 2010). Well, no matter how hard he tried to be honest, he couldn’t be, because the public situation requires that if something is said about a deceased friend, this must be seen as praise. Private intentionality succumbs to public intentionality, determined by linguistic rules, preceding praxis, and the historical and cultural context. 3. Pragmemes at the market-place as socio-pragmatics. After this preamble on pragmemes, their public effects, and their conventions of use, let us return to the main issue of this paper, which is pragmemes in the market place. To use the words by the philosopher Austin, speech acts serve to change the world (to do things with words), to modify and produce actions on the part of the participants, both on the basis of the social relationships that exist between the participants (and which contribute to conferring certain deontic attributes on them by virtue of which someone has the right to see his or her speech act satisfied, while we expect another to show uptake), or on the basis of wider persuasive speeches (the weapon of rhetoric, the pragma-dialectic dimension or argumentation (van Eemeren and Grootendorst, 1984; Macagno and Walton, 2015)). In a market place, where so many social classes merge and where social (and even racial) segregation is impossible, as sellers expect to make a profit from all their clients, regardless of their social class, their ethnic origin and religious persuasion, we expect the most effective weapon to be that of persuasion. This is why vendors sharpen their ingenuity, often based on techniques they have inherited from their predecessors (from preceding praxis), to capture the goodwill of the clients. The most important thing, at least in the markets I saw when I was a child, was to get the attention of the customers, to

Pragmemes at the Market Place

141

get thick circles of customers around the seller, often placed on a very high chair used to attract the largest number of people by getting a natural amplification of the voice (staying up high could also be used to better control what happens). The one who ‘bandia ‘or ‘vandia ‘(shouts) (the use of ‘v’ or ‘b’ depends on the transcription of the local dialect) must first obtain the attention of the public (of the clients) and must do so through special appellations or terms of address (Signora, Signorina), which are often placed in sequence and sometimes exploit the inversion technique, which seems to recall broader poetic techniques. In the transactions of the market place, the use of the NP ‘Signuritta’ (lit. Small lady) is really striking. This is a diminutive syntactic form and corresponds to a pragmeme, if we accept the considerations by Mey, 2010 on reference (or the act of referring) intended as pragmeme. How is it possible that the use of a single word, without a verb or other predicate to be saturated, can count as a pragmeme (the social act of referring to something)? Although some philosophers (eg Searle, 1979) treat reference (the act of referring to someone) as a linguistic act, they do not clarify how a linguistic act can be performed without a predicate being saturated by an NP (assuming that ‘Signuritta’ is a description containing conceptual materials, we expect that the act of reference occurs an NP that matches the content of the description). In the case of the description/term of address ‘Signuritta ‘(here conceptual materials ensure that the object should satisfy the description in virtue of having qualities as dictated by the description), the speaker looks for the person he is talking about in the world, but at the same time he wants to establish a certain relationship with her. We may want to call such NPs interactional descriptions, meaning that their form refers not only to a referent according to a conceptual rule, but also reminds us that the NP has an interactional dimension (a word’s meaning is its use, according to Wittgenstein (see Mey 2016)). This interactional dimension is provided, presumably, through connotations, that is to say memories of past uses of the NP (see Kecskes, 2014 on this; Kecskes is concerned with the speaker’s or the hearer’s prior experience which has a bearing on interpretation and constitutes a kind of egocentrism (Kecskes, 2010, Kecskes, 2008)). While it is possible for a sub-section of the lexicon to be linked to the dimension of egocentrism, I am not sure that the connotations of certain words like ‘Signuritta’ are best explained by making recourse to the speaker’s egocentrism. It is possible that these connotations too belong to the public dimension of language as a community of language users share the same past experiences. As Wittgenstenian scholars may say, past uses of words remain impressed in our minds and may happen to have an influence on interpretation. As Kecskes says, words like ‘police’ may be associated with traumatic events, which may predominate in shaping meanings. Past uses of a word count as well, as the word becomes linked with associations which may be positive or negative (depending on the past experience of the speaker) and may determine the direction of global interpretation. Words which have an interactional dimension are normally discourse markers such as ‘well’, ‘But’, ‘you know’, etc. But these usually have an adverbial function and are placed, grammatically speaking, at the periphery of a sentence, determining the function which a sentence may serve in being placed in a certain

142

A. Capone

discourse. It is quite out of the norm that an NP should have a discourse function, which would be more naturally associated with an adverbial phrase serving to connect previous and subsequent discourse; however, it is not unusual in languages like Italian for NPs to acquire discourse functions, as in the case of diminutives for names (e.g. I was called ‘Sandrino’ by my parents as a child; then, I progressed to Sandro); but we have to discuss this case, as it appears to be not devoid of theoretical import. A woman who is called ‘Signuritta’, in the context of the market and, in particular, of vendor/ client interaction, must see the vendor as someone who tries to satisfy her, who tries to meet her demands, but cannot offer her a lower price (or satisfy some other type of request), because otherwise he would not gain anything); in other words, the form of the appellative has strong connotations. This NP, used as an appellative and very close to a proper name in that it refers rigidly to a referent who has human features and also plays the role of addressee (thus, it has second person features) is certainly indicated for market transactions, since it presupposes that the person in question has requested a lower price, and also assumes that the vendor cannot meet (much to his regret) this request. At the same time, the vendor requests (but does not assume) that the customer meet his request given that a reasonable price was proposed. Clearly the whole presupposes a negotiation, a tug-of-­ war, a situation in which the participants are willing to bargain, but at a certain point the bargaining stops. The trader, however, instead of giving a long explanation, uses the term of address ‘Signuritta’ and somehow mitigates the effect of the ceased bargaining. Furthermore, ‘Signuritta ‘indicates (as if it were a marker of speech) that the shopkeeper is going to explain the cessation of bargaining with a series of reasons which are conclusive. The trader feels compelled to justify himself and to introduce a long argument in his defense. At the same time, he assumes that the customer has the right to complain, to want to proceed further in the negotiation. Today it is really difficult to record broad and poetic speeches in open-air markets, apart from some exceptions that I reported in previous articles. But at the time when I was a child, the vendor at the market was really a poet in that he used formal poetic techniques. Not that he created a poetic and structured repertoire of metaphors and effects like inversion, parallelism, metaphors, etc. Probably, the mind was trained by what he had already heard and learned, and often what was already known was repeated (For example, to say that there is a sale, the seller pretends to be crazy and says variations of topoi like ‘I have gone mad, I sell everything, everything at very low price’. This sentence, which I have extracted on the basis of real examples of language use, can have, in real life, hundreds of variations and it would be very useful to study them all comparatively, if it were not for the fact that the market is disappearing due to the threat and competition of the shopping malls and that the vendors, discouraged by the financial crisis and by the lack of crowds of customers, do not find it useful any longer to stick to the tradition of the ‘vandiari’ and to put it into practice. What can be studied now is only the pale memory of what we have seen when children. Very frequently, especially in the Palermo markets, you can hear the obsessive repetition of a sentence like:

Pragmemes at the Market Place

143

2. Tri melanciani un euro, tri melanciani un euro, tri melanciani un euro. (three aubergines a euro) The final (rhematic) section of each of these utterances (un euro) is obviously more heavily emphasized, with the effect that it is not always easy to understand the content of the whole statement. I give up explaining the triadic structure of the sentence (why is it repeated three times and not four? Is this a religious value, with reference to the Trinity?). But it remains to be explained that in the Palermo markets this kind of utterances (the specification of the price) has this kind of repetitive structure. The repetition does not seem to me due to the fact that a person would find it difficult to understand the price of the product, but it seems to be explained by the fact that, for the most part, the crowd of these markets, unlike for example that of the city of Barcellona (Sicily) where I was born and raised, where the crowd stands in a circle around the seller and can be called a ‘standing crowd’, is an itinerant crowd. As more people pass and different people pass at slightly different times, the statement not only creates an amplification on a sound level through an increase in acoustic energy (note that these statements are uttered by shouting at the top of the sellers’ lungs), but also an amplification that follows the passers-by moment by moment. It is appropriate to say, following Goffman (1981) on Forms of talk, that the speaker, besides being the author and principal (he who is responsible for the intentions of the linguistic action) is the sounding box, the one who lends his voice, and who follows the speakers with his voice step by step. We have already made it clear that today’s markets are a pale reminder of yesterday’s. However, given my description of what happens, it is possible to imagine a market where the speakers are busy in a choral action. The different roles that different voices would assume in a choral utterance manifests itself through the recruitment of different voices. The totality of the products of the market constitute the whole of the items the ‘vandiaturi’ (those who shout) can talk about. The possibility that the overall rhythm depends on individual sequences is not excluded; however, it is not easy to imagine in practice what happened in the past. The attention of sociologists has focused on individual cases, but not on the overall action, and in any case has not been attentive to linguistic and conversational details, considering these events only manifestations of popular uses and traditions. The popular tradition, on the contrary, would need to be studied from the point of view of the conversational structure, of non-­ monotonic inferences and of the concatenative structures between single communicative events, of which, unfortunately, we know very little. That there can be chains is clearly shown in the article by Capone (2018), which points out that a seller utters the sentence ‘Arancia di vaniglia’ (Vanilla Orange), to which another seller responds ‘chi tra li corna lu pigghia’ (which takes it among the horns). The example is a case in which one of the sellers wants to show off his poetic ability by responding to what another has said. It is clear that there can be concatenations between the sentences/ utterances, and this can happen both at the level of the meaning and at the level of the rhythmic structure. Utterances at the market place are often fragments (of sentences/ utterances), and as such must be reconstructed based on the context (physical, co-text and socio-­

144

A. Capone

cultural context). They require a process of rational expansion (see the very ­interesting book by Stainton, 2006 on sentential fragments). Consider the following fragment 3. ‘Tutti cosi senza piccioli’ ((We sell) everything without money). The compulsory expansion consists in arranging this apparently modified noun phrase ‘tutti cosi’ (everything), apparently modified by ‘senza piccioli’ (without money), into a tree structure in which there is at least a noun phrase and an unexpressed verb (vindemu (sell)) which then together with the NP ‘Tutti cosi’ is a VP which is a sister of the initial (unexpressed) NP. This unexpressed NP (0 subject or pro both in Sicilian and in Italian) saturates the (unexpressed) predicate (VP), which must be reconstructed through contextual clues and presumably corresponds to ‘sell’. It is interesting to note that the adverbial phrase ‘senza piccioli’ cannot modify’ ‘Tutti cosi’, as it would attribute the predicate ‘senza piccioli’ to ‘tutti cosi’ (the result of the semantic composition would be roughly ‘all things that have no money’, which makes no sense). The meaning intended by the speaker, instead, should be something like ‘Here we sell everything without money’ (another plausible expansion would be ‘Here you can buy everything without money’). In other words ‘senza piccioli’ modifies ‘si vendi’ (we sell). We need a semantico/pragmatic (structural and functional) expansion not only to get a sentence, but also to accurately establish what modifies what and add syntactic nodes to the right nodes, regardless of what the surface structure seems to suggest. In my expansion, I have not only added a pronoun (the choice of a pronoun compared to a noun is something arbitrary, although justifiable), but I introduced the adverb ‘Here’ in the structure of the sentence to expand; this adverb, needless to say, serves to convey the comparative (rhetorical) structure of the statement: Here (which conversationally implicates “but not in the other stalls”) you buy (or: we sell) everything at a low price. Note that in my expansion I also modified the adverb ‘senza piccioli’ (without money) with ‘a basso prezzo’ (at a low price), which is what the seller really means in a hyperbolic style. The seller uses one of the most important rhetorical figures (hyperbole), but one that is also very widespread and frequently used at the level of the local dialect (For example ‘Non vinni nuddu’ (nobody came) turns out to mean ‘Few people came’). This means that the market has been his learning environment where he has acquired many of the speaking and argumentative techniques necessary to amaze and attract customers. Note that at the level of the video from which this fragment was taken. (https://www.youtube.com/watch?v=i6woVaxIWYA), the show represented by the active participants repeats the scheme (script) which I discussed earlier, an apparently mad seller who sells his goods at a very low price (in this case, apparently without making money). The style of the performance activates this script. One of the participants shouts apparently disconnected phrases like ‘Maria!’. Not only is the script invoked by some explicit or implicit utterances (‘I feel unwell. Ahi’), but it is represented in a theatrical way (see the utterance which is proffered in a very loud voice ‘Put me on facebook ‘). The utterance ‘Maria’, which is often uttered, may have a double interpretation: (a) it may indicate surprise (or the expectation that customers will be surprised by prices) ; (b) it could be an utterance in

Pragmemes at the Market Place

145

which a very common proper name is used instead of a generic pronoun (see the use of the name ‘Mary ‘ for woman in Tok Pisin (Capone, Forthcoming). Both interpretations can be credible, although I prefer the former because of its conventionality. The former interpretation is also more compatible with the script ‘I have become mad and I sell everything’. The sentences proffered at the market as well as being phrasal fragments can be considered conversational, discursive and argumentative fragments. Consider the following, repeated many times 4. ‘Non ci va nuddu da capu’ (https://www.youtube.com/watch?v=BlF6AR6mGZ8&list=RDi6woVaxIWYA &index=2) This sentence/utterance, in itself very cryptic, becomes much less cryptic when inserted in an argumentative sequence. At least other fragments must be put together to solve the puzzle. The speaker is making a comparison between his shop and a competing one that may be up the street. If one were not aware of these comparative language games, one would certainly not be able to understand the pragmatic import of the utterance and one would not be able to connect it to the conversational sequence. This clearly proves that the view by Schegloff (1988) and Haugh (2015) that actions are not realized through individual utterances by speakers but that they emerge in discourse and should be understood by reference to sequential context is a sacrosanct tenet of pragmatics. 5. Aiutatimi a fari scindiri chisti pi ca sutta (help me to let these people come over here) Si vende tutti cosi. (Here we sell everything) Sentiti a mia. (listen to me) Si vendi tuttu. (here we sell everything) Aiutu. (Help). Nun ci vaci nuddu da capu (Nobody goes to the stall up there) Si vendi tuttu. (here we sell everything) .....

146

A. Capone

U mandarinu dda un euro qua 50 centesimi (mandarines there cost a euro here they cost 50 cents) Allora chissà che c’è sutta. (Then God knows what there is behind that) Si vendi tuttu. (here we sell everything) The comparison is made with some other stall and it is said that while nobody goes up there, everything is sold down here. An exhortative ‘Help me get these people down here’ is also used. The obsessively repetitive structure is due to an important structural/functional reason. Those who have lost track of previous or subsequent utterances would not be able to understand the argumentative role of the utterance. This is why one needs to recycle conversational fragments many times. And given the argumentative structure, there is not only the fragment of the sentence in question, but also the others that serve to complete the argumentative structure to be considered.

Pragmemes and Deontic Modality Consider the following statement: 6. O picciriddu (for the small child) proffered by a vendor who brings fish with a wheelbarrow to Ballarò market (Palermo). To say that this utterance is meaningful is to say little. The utterance requires a considerable and intelligent expansion work and is the demonstration of how various areas of knowledge must interact in order to bring out the meaning of/ from a short sentence (in this case, a fragment of a sentence). The phrase that holds the PP together (prepositional phrase) and underlies the correct interpretation of the sentence is something like this: NP VP (PP) where the PP represents a lower node than the VP. The fragment recovers its dignity as a sentence through an embedding operation that allows one to obtain the VP (verb phrase), which then saturates the subject constituted by the unexpressed NP (e.g. You can give this (fresh) fish to your child). The expansion operation, however, is not automatic, but requires knowledge of encyclopaedic and cultural information. In fact, the PP is involved in the saturation of the predicate by the subject, as it is part of a node subordinate to the VP and it is not part of a node subordinate to the NP. Those who truly understand the fragment (6) must know that (a) Ballarò market does not guarantee high quality for the fish; (b) those who are interested in quality must interact with the seller and clarify that the fish is intended for categories of

Pragmemes at the Market Place

147

people who are quite fragile, such as old people and children; (c) that there is an ethical dimension in the behaviour of sellers, so that, if fish is required for old people and children, they will act accordingly and meet consumer demand. The real problem of interpretation is that (6) is provided on a voluntary basis and that it is not a response to a real request (or a request for information). Therefore, it is not clear what status (6) has. Does the utterance (6) work like a way of advertising a product and provide an implicit form of argument? Or is it to be seen as an answer to an implicit request? In the former case, it does not seem that there is necessarily an ethical dimension in the behaviour; indeed, the act of advertising a product could require an unethical action, that is to say something that is not necessarily true and need not persuade consumers to buy the advertised product (the addressees may be prepared to concede that a lie was told, if they think the speaker is not trustworthy. And why should someone they do not know be trustworthy?). In the latter case, one could glimpse an ethical dimension, but one wonders why the seller refers to this ethical dimension, given that no one asks him a question or a request. An implicit request from a generic consumer does not make much sense, as the final consumer or the addressee may not be the one we are confronted with (we imagine that old people and children have some problems buying goods in such a crowded market). And here comes into play the Everything considered inference of pragmatics (Cummings, 2009). Cummings doubts that the theory of mind module may in all cases provide standardized responses to interpretative problems and that standard inferential mechanisms are especially restricted to a particular cognitive mechanism that resides in a given module. According to her, various types of information interact and pragmatic interpretation is the result of an inferential process that is also based on reasoning and related to general intelligence (e.g. deduction processes). Since an interpretative hypothesis is more reasonable than another, we choose the former. It would appear that, in this case, a solution is provided due to an inference to the best explanation. The two possible interpretations must be evaluated in a comparative manner (and if there were a third one, this should also be evaluated) and the interpretation actually chosen (voluntarily even if perhaps implicitly) would be the one that generates least conflicts with the information that we consider accepted, taken for granted, both by the speaker and by the addressee. The interpretation of this utterance requires a temporal collocation, but also a place in space and culture. It seems that this interpretation predominates here, in this historical moment and in this culture. The ethical dimension, in fact, is mainly due to the cultural context in which the speakers are immersed. When you go to the market, you expect so much dishonesty (for example the worst oranges are placed at the bottom of a box; if they tell you that a box weighs 5 kilos, you can expect a maximum of 3 kilos; furthermore, the price is rounded off with some 9 that is written in a way similar to zero); but when it comes to old people and children, this dishonesty is reduced to a minimum. The place is very important, as we expect this culture to be a function of the place where we find ourselves. The place also provides coordinates for interpretation including parameters for moral conduct and interpretation that depends of values.

148

A. Capone

The Pragmeme ‘Selling the Fish’ at the Market Place. Consider the following statement: 7. Guarda (Look at it) Questo pesce è locale (this is a local fish) Veni veni (Come come) Guarda (Look at it) A tunnina viva e ch’è bella (tuna fish is alive how beautiful) Ca cipuddata (with fried onions) Ca cipuddata è (it is with fried onions) A tunnina, viva viva è (tuna fish is alive alive) Viva viva (alive alive) Na rosa viva viva (a rose that is alive alive) A pigghiau na signurina (a young lady took it) Na signurina a pigghiau (It a young lady took) A tunnina (tuna fish) Viva viva e bella (alive and beautiful) Viva viva e bella (alive and beautiful) U culuri du mari (the colour of the sea) Pinna gialla è (it is yellow fin) Pinna gialla (yellow fin) Pinna gialla è (yellow fin it is) È pinna gialla (It is yellow fin) Che bella sta tunnina (how beautiful this tuna fish is) Si po manciari cruda (one can eat it raw) Scalau (the price went down) Scalau (the price went down) Ca cipuddata (with fried onions) Questa si può baciare (one can kiss it)

Pragmemes at the Market Place

149

(the translations are literal). This statement was taken from the Corpus: Mercati storici siciliani (Centro regionale per l’inventario, la catalogazione e la documentazione grafica, fotografica, aerofotografica, fotogrammetrica e audiovisiva dei beni culturali e ambientali, 2006). First of all we note that the utterance is collaborative, as two people collaborate on the same text and can be defined co-authors and co-principals (to use Goffman’s categories). We have differentiated the two individuals or actors that collaborate in the same utterance through the use of bold characters. We have also underlined the parts in Italian, differentiating them from the prevalent use of the local Sicilian dialect. The use of Italian is a case of code-switching and could indicate not so much a formal occasion, but the fact that the seller recognizes various layers in the composition of customers, ranging from the less educated to the more educated and wealthy. (This technique is reminiscent of Shakespeare’s techniques which seek to involve various layers of his audience, providing more earthy shows land for the less educated). Goffman (1981), in a seminal paper, pointed out that similar acts of code switching could be related to participation and, in particular, to the composition of the addressees. The utterance can be seen as mainly an argumentative act. The seller tries to persuade the potential customers to buy the fish and uses conventional expedients. When he says that the fish was taken by a young lady, there is basically an ambiguity: was the fish eaten or bought by a young lady? Probably the seller makes the point that the young lady bought and ate the fish and that her particularly delicate stomach was able to digest it without problems. The structure of the utterance/language game is as follows: (a) The seller (or rather one of the sellers) invites those present to look, to observe the fish. (b) The seller specifies the best cooking method (with the onions) (c) The seller specifies the qualities of the fish in a subjective manner; (d) The seller specifies the qualities of the fish in an objective manner. (e) The seller uses various metaphors. (f) The seller makes comparisons (this is better with respect to x). (g) The seller relates the quality of the product to the type of customer he has in mind. (h) G.  The seller makes gestures that emphasize and accompany the argument produced. (i) The seller produces very unusual metaphors (you can kiss the fish), which attract attention. (j) The seller makes use of local cultural elements (for example the mention of the young lady who bought the fish), being sure that the conventions can facilitate the interpretation of the text. As regards point h), we refer to an image.

150

A. Capone

Image h)

Conclusion In this article, I have produced an analysis of pragmemes by emphasizing the fact that they rely on a cultural background which is the input of interpretation rules. The pragmemes at the market place, on the one hand, work like poetry and follow strict rules; on the other hand, they serve the purpose of achieving freedom of expression, that best allows you to advertise products to sell (creating curiosity and attracting the attention of potential customers). Rules and linguistic innovation are the basis of pragmemes at the market place. Following the rules, market vendors also follow the rule that rules must be broken and the purpose of innovation and expressive freedom can be pursued. Conventionality and expressive freedom are the basis of our linguistic culture, the tension between them allows the language to improve itself functionally, change, restructure itself and even give itself new rules. The sellers of the market place are fundamentally poets, in the sense that one of the functions of the texts presented is the poetic one. The poetic function, in these cases, is not prevalent, but is almost always subordinated to the argumentative and the conative function (as well as the phatic function of opening a communication channel between the one who sells and the one who buys). The interweaving of argumentative, conative, phatic and poetic functions is very important. Generally, these functions are considered separately, but it is clear that in everyday speech they merge together and

Pragmemes at the Market Place

151

s­ upport each other. The argumentation act on its own is not enough, but as in advertising, it needs metaphors and poetry. Indirect messages, often and willingly, work better than direct messages, as they involve a cognitive effort and this cognitive effort is subordinated to memorization, increasing the impact that the memory of the product can have on the human mind. In the same way as, in an advertisement for Daikin air conditioners, butterflies were used to suggest that the air produced was particularly light and healthy, thus adding a poetic dimension to the language of advertisements, often even in market discourses the non-literal sense is used to mentally engage addressees in an inferential mechanism that has the purpose to enhance the argument and create effects that literal discourse could never achieve. This way, the argument gets impressed in people’s minds. So, when the fish seller insists that the tuna fish has the colour of the sea and is alive, we are confronted with messages that are, from a literal point of view, blatantly false and therefore require the reconstruction of the (true) intentions of the speaker. ‘Viva’ means that the tuna is fresh (not literally alive), just caught. That tuna has the colour of the sea means that it has a colour that indicates the freshness of the fish. When the seller says that it (the tuna fish) can be kissed, the utterance probably has a greater emotional impact as it intersects the plane of love with that of food. The strangeness of the metaphor ensures that it is remembered, that it does not go unnoticed by the passerby (and potential buyer). Of course, if they told us to kiss a fish, we would be taken by disgust, due to the memory of the strong smell generally emanating from it (especially from fish that is not fresh enough). But if the speaker tells us that this fish can be kissed, he probably suggests that it is so fresh that it does not emit bad smells. Obviously the comparison is between kissing people and kissing tuna. The experience of kissing people who, for various reasons, emit bad smells is evoked. Therefore, the metaphor has a strong connotative impact. And it is the connotations that justify it, which ensure that the metaphor works, and that the speaker’s intentions are correctly reconstructed. It is really striking that my ideas on the adaptability of language in connection with the poetic function (clearly displayed by pragmemes) intersect with considerations on linguistics and poetry by Roman Jakobson (1960): Sometimes we hear that poetics, in contradistinction to linguistics, is concerned with evaluation. This separation of the two fields from each other is based on a current but erroneous interpretation of the contrast between the structure of poetry and other types of verbal structure: the latter are said to be opposed by their “casual”, designless nature to the “non-­ casual”, purposeful character of poetic language. In point of fact, any verbal behaviour is goal-directed, but the aims are different and the conformity of the means used to the effect aimed at is a problem that evermore preoccupies inquirers into the diverse kinds of verbal communication (Jakobson, 1960. 351).

Mutatis mutandis, we see that the considerations by Jakobson are echoed by those put forward by Verschueren (1999) and Mey (2001). The adaptability of language, which the latter authors accept and are proponents of, has deep effects in favouring the unification of poetics and linguistics. Poetic techniques imply an effort by practitioners to improve their mastery of poetic art. The adaptability of language (as demonstrated in this paper on pragmemes at the market place) ensures that language users develop (probably subconsciously) specific responses to

152

A. Capone

specific problems. Pragmatics is part of the picture. At the time in which Jakobson wrote, pragmatics did see a great expansion and much less of it was known to linguists (this does not mean that all linguists are favourable to pragmatics these days). But pragmatic expansions and indirectness can be seen as conscious responses on the part of language users to societal or practical problems. Thus, at least part of language (if language use is admitted to the picture) can be seen as not designless, but as purposeful. The purpose of having developed pragmatic machinery often lies in producing responses to environmental and societal problems. Those who are opposed to these considerations on the grounds that Jakobson was only speaking of langue and not of parole, may have to accept that gradually pragmatic information seeps into language and may become part of langue or full-blown linguistic competence (see the cases in which pragmatic inferences become frozen (Horn, 1989)). In this paper, I have mainly clarified previous considerations on pragmemes, taking into account the triadic view of language (see Wong, 2010), in which cultural information plays an important role. I then provided various cases of pragmemes to analyse, including pragmemes at the market place. I have argued that the notion of social praxis plays an important role in determining the content of a certain global speech act. Both context and culture are involved in this process of shaping meaning.

References Allan, K., Capone, A., & Kecskes, I. (Eds.) (2016). Pragmemes and theories of language use. Cham, Switzerland: Springer. Allan, K. (2018). Getting a grip on context as a determinant of meaning. In A. Capone, M. Carapezza, & F. L. Piparo (Eds.), Further advances in pragmatics and philosophy (pp. 177–201). Cham, Switzerland: Springer. Austin, J. L. (1962). How to do things with words. Oxford, UK: Clarendon. Bernstein, B. (1971). Class, codes and control. The structuring of pedagogical discourse. London, UK: Routledge. Bauman Richard & Joel Sherzer (Eds.) (1974). Explorations in the ethnography of speaking. Cambridge, UK: Cambridge University Press. Capone, A. (2005). Pragmemes (a study with reference to English and Italian). Journal of pragmatics, 37(9), 1355–1371. https://doi.org/10.1016/j.pragma.2005.01.013. Capone, A. (2010). On pragmemes again. Dealing with death. In A. Capone (Ed.), Perspectives on language use and pragmatics (pp. 149–168). Muenchen, Germany: Lincom. Capone, A. (2016). Introduction. In Allan, K., Capone, A., & Kecskes, I. (Eds.), Pragmemes and theories of language use (pp. 1–2). Cham, Swtizerland: Springer. Capone, A. (2018a). Pragmemes again. Lingua, 209, 89–104. https://doi.org/10.1016/j.lingua .2018.04.004. Capone, A. (2018b). Embedding explicatures in implicit indirect reports. Simple sentences and substitution failure cases. In A. Capone, M. Carapezza, & F. L. Piparo (Eds.), Further advances in pragmatics and philosophy (pp. 97–136). Cham, Switzerland: Springer. Capone, A. (Forthcoming). Pragmatics and philosophy. Connections and ramifications. Cham, Switzerland: Springer. Carston, R. (2002). Thoughts and utterances. Oxford, UK: Blackwell. Coulmas, F. (1981). Conversational routine. Explorations in standardised communicative situations and pre-patterned speech. The Hague, Netherlands: Mouton De Gruyter.

Pragmemes at the Market Place

153

Cummings, L. (2009). Clinical pragmatics. Cambridge, UK: Cambridge University Press. Dascal, M. (2003). Interpretation and understanding. Amsterdam, Netherlands: Benjamins. Goffman, E. (1981). Forms of talk. Philadelphia (PA): University of Pennsylvania Press. van Eemeren, F. H., & Grootendorst, R. (1984). Speech acts in argumentative discussions: A theoretical model for the analysis of discussions directed towards solving conflicts of opinion. New York, NY: De Gruyter. Goldberg, S. C. (2019). When a speaker is reported as having said so. In A. Capone, M. Garcia-­ Carpintero, & A.  Falzone (Eds.), Indirect reports and pragmatics in the world languages (pp. 133–147). Cham, Switzerland: Springer. Gross, S. (2001). Essays on context-sensitivity and its philosophical significance. New York, NY/ London, UK: Routledge. Gumperz, J. (2001). Interactional sociolinguistics. A personal perspective. In D.  Schiffrin, D. Tannen, & H. Hamilton (Eds.), The handbook of discourse analysis (pp. 215–228). Oxford, UK: Blackwell. Haugh, M., & Jaszczolt, K. (2012). Speaker intentions and intentionality. In K. Allan & K. Jaszczolt (Eds.), The Cambridge handbook of pragmatics (pp.  87–112). Cambridge, UK: Cambridge University Press. Haugh, M. (2015). Im/politeness implicatures. Berlin, Germany: De Gruyter. Higginbotham, J. (2000). On events in linguistic semantics. In J.  Higginbotham et  al. (Eds.), Speaking of events (pp. 49–79). Oxford, UK: Oxford University Press. Horn, L. (1989). A natural history of negation. Chicago, IL: Chicago University Press. Hymes, D. (1962). The ethnography of speaking. In T.  Gladwin & W.  C. Sturtevant (Eds.), Anthropology and human behavior (pp. 13–53). Washington, DC: Anthropological Society of Washington. Jacobson, R. (1960). Linguistics and poetics. In T. Sebeok (Ed.), Style in language (pp. 350–377). Cambridge, MA: MIT Press. Jaszczolt, K. (2016). Meaning in linguistic interaction. Oxford, UK: Oxford University Press. Kecskes, I. (2008). Dueling contexts. A dynamic model of meaning. Journal of Pragmatics, 40(3), 385–406. https://doi.org/10.1016/j.pragma.2007.12.004. Kecskes, I. (2010). Situation-bound utterances as pragmatic acts. Journal of Pragmatics 42(11), 2889–2897. https://doi.org/10.1016/j.pragma.2010.06.008. Kecskes, I. (2014). Intercultural pragmatics. Oxford, UK: Oxford University Press. Macagno, F., & Walton, D. (2015). Classifying the patterns of natural arguments. Philosophy & Rhetoric, 48(1), 26–53. https://doi.org/10.5325/philrhet.48.1.0026. McDowell, J. (1998). Meaning, knowledge and reality. Cambridge, MA: Harvard University Press. Mey, J. (2001). Pragmatics. An introduction. Oxford, UK: Blackwell. Mey, J. (2010). Reference and the pragmeme. Journal of Pragmatics 42(11), 2882–2888. https:// doi.org/10.1016/j.pragma.2010.06.009. Mey, J. L. (2016a). Pragmatics seen through the prism of society. In A. Capone & J. L. Mey (Eds.), Interdisciplinary studies in pragmatics, culture and society. Cham, Switzerland: Springer. Mey, J. L. (2016b). Why we need the pragmeme, or speech acts and its peripeties. In K. Allan, A.  Capone, & I.  Kecskes (Eds.), Pragmemes and theories of language use (pp.  133–140). Cham, Switzerland: Springer. Recanati, F. (2004). Literal meaning. Cambridge, UK: Cambridge University Press. Sacks, H. (1992). Lectures on conversation. Oxford, UK: Blackwell. Schegloff, E. (1988). Presequences and indirection: Applying speech act theory to ordinary conversation. Journal of Pragmatics, 12(1), 55–62. https://doi.org/10.1016/0378-2166(88)90019-7. Searle, J. (1979). Expression and meaning. Cambridge, UK: Cambridge University Press. Stainton, R. (1998). Quantifier phrases, meaningfulness “in isolation”, and ellipsis. Linguistics & Philosophy, 21, 311–340. Stainton, R. (2006). Words and thoughts. Oxford, UK: Oxford University Press. Van Dijk, T. (1980). Macrostructures. An interdisciplinary study of global structures in discourse, interaction, and cognition. London, UK: Routledge.

154

A. Capone

Verschueren, J. (1999). Understanding pragmatics. London, UK: Arnold. Wierzbicka, A. (2006). English: Meaning and culture. Oxford, UK: Oxford University Press. Wittgenstein, L. (1953). Philosophical investigations. Oxford, UK: Blackwell. Wong, J. (2010). The triple articulation of language. Journal of Pragmatics 42(11), 2932–2944.

Pragmatics of Self-Reference Pronouns in Capital Trials Krisda Chaemsaithong

Introduction Pronouns have been shown to be a key resource for negotiating one’s position and (dis) aligning with others (De Fina, 1995; De Marco and Palumbo, 2016; Gardelle and Sorlin, 2015; Goffman, 1981; Tang and John, 1999). This is so because the meaning associated with pronouns is not only referential but also pragmatic and dependent on the specific context of an utterance and the user’s goal. To illustrate, because of the semantic feature [speaker + at least one other], the pronoun we can convey empathy and solidarity (Wales, 1996, p. 84), or distance the speaker from what is being said (Goffman, 1981), depending on whom the speaker intends to include or exclude as part of “at least one other”. It has been shown, for example, that politicians use we to identify themselves with supporters by emphasizing the supposed homogeneity between the groups. Conversely, they also use the same pronoun to foreground differences between them and whom they consider enemies or out-­group members, thereby signaling disaffiliation (Bull and Fetzer, 2006; Proctor and Su, 2011; Wilson, 1990). The present study focuses on first-person pronouns along with their grammatical variants (hereafter, I and we for short), and explores how they can be strategically manipulated for rhetorical effects in capital trials. In particular, this study attends to the closing speech in the penalty phase, a prime example of persuasive discourse representing the prosecution’s and defense’s last opportunity to secure the jury’s choice of death or life. Integrating Goffman’s (1981) concept of “footing”, or speaking position, in the analysis of ten closing statements from five capital trials, this study aims to answer the following questions: (1) How do lawyers present K. Chaemsaithong (*) Department of English, Hanyang University, Seoul, South Korea e-mail: [email protected] © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 F. Macagno, A. Capone (eds.), Inquiries in Philosophical Pragmatics, Perspectives in Pragmatics, Philosophy & Psychology 28, https://doi.org/10.1007/978-3-030-56696-8_9

155

156

K. Chaemsaithong

themselves through first-person pronouns in this monologic communicative event, and (2) To what extent are first-person pronoun choices intertwined with the lawyers’ ideological positions on death sentence and the individual on trial? As will be shown shortly, first-person pronouns are ideologically motivated and allow lawyers to foreground their participation as members of different social communities other than that defined by the current setting, thereby constructing an authoritative and credible self-image and aligning the jurors with their position on the death sentence. My inquiry into self-reference pronouns in this context contributes to deconstructing the dynamic, yet subtle, process in which lawyers perform and enact multiple aspects of self and speaking roles that serve to legitimize state killing for the prosecution, or to neutralize such an attempt for the defense. This paper begins by providing some background to the closing argument in the penalty phase of capital trials before proceeding to a detailed discussion of the theoretical insights that inform this study. The findings with respect to the participation alignments invoked by I and we, respectively, are then presented and discussed in detail.

Capital Trials, the Penalty Phase and Closing Arguments Capital trials differ from ordinary criminal trials in that they feature a bi-furcated procedure. In addition to the guilt phase found in ordinary criminal trials, where the jury weighs the evidence presented and witness testimony and judges culpability accordingly, a second phase—the penalty phase—follows, should the defendant be found guilty. In this phase, the same set of jury is entrusted with the task of weighing aggravating factors (reasons that would incline jurors toward the death sentence) presented by the prosecution against mitigating factors (reasons that would incline jurors toward the life sentence) presented by the defense. For a death sentence, the jury must determine that the aggravating factors outweigh mitigating ones, or else life imprisonment will be imposed. Occurring after the presentation of evidence of aggravating and mitigating circumstances and before the delivery of the sentence verdict, the penalty phase closing argument is not only the chronological and psychological culmination of a capital trial but also “the dramatic highpoint of the capital trial” (Burt, 2008, p. 903). This last appeal is not a mere summary that recounts what each witness has said. On the contrary, to solidify juror allegiance, it must move beyond facts and evidence to arguments emphasizing morality, justice and human motivation, show how the recommended sentence will serve the interests of society, and explain why life or death is consistent with widely held moral principles (Costanzo and Peterson, 1994, p. 126). A body of legal research speaks to the significance of a good closing argument. Amsterdam and hertz (1992) hypothesize that during ambiguous cases, closing arguments can be crucial in persuading the jury to vote one way or the other. Similarly, Levin et al. (2003) argue that cases can be won or lost during this phase.

Pragmatics of Self-Reference Pronouns in Capital Trials

157

Observing 66 closing arguments in Philadelphia, Walter (1988) concludes that the most successful lawyers are those who are most aware of their histrionic function, that is, “an awareness that summation is a performance” (p. 222) “as if it were the lawyer’s one aria in the opera of the trial” (p. 225). One lawyer commented during the interview: “My mission is to persuade the jury of the equity of my client’s cause…with rapprochement with the jurors, creating, hopefully, the feeling that we are a team, the court, the jury, and me as a lawyer” (p. 42). In short, “[m]uch of what [lawyers] do and say during the trial, including summation, is directed towards getting the jurors to like them” (p. 221). Focusing on closing arguments in ordinary criminal trials, Rosulek (2015, pp.183-185) report that first-person singular pronouns occur more than first-person plural pronouns, and that the prosecutors use fewer pronouns than the defense. The researcher notes that the defense is more prone to using an undefined we for which the context has to be used to determine the referents. Most insightful to the present study is that first-person plural references can include: lawyer and jurors, lawyer and side, lawyer and a witness, all lawyers, all participants, and all people. Conley (2016) finds that varieties in person reference can operate as both distancing and inclusive tactics among jurors, attorneys, and defendants. For instance, one juror whom Conley interviewed distanced himself from the capital defendant by using the indefinite and generalized subject you rather than the more common form I, as in “you think, well, that person, you know, they just need to be...eliminated” (p. 156). By using a generic you, the juror evades personal responsibility for the statement and for rendering the death verdict. In stark contrast, a juror may also position the defendant as an actual person when referencing their face-to-face meeting through the use of the inclusive we in describing their similar upbringings, as in “It was very interesting to meet Bobby, and he’s a very gentle soul, and we, as Bobby says, we’re two sides of the same coin. We were educated probably within a mile and a half of each other” (p. 156). It is argued that this management of proximity between the jurors and defendant through pronouns can stymie or construe empathy with the defendant and justify jurors’ decisions for death. All in all, past studies have shown that the lawyer’s construction and management of self and their relationships with the jurors are critical to the trial outcome, and that pronouns constitute a prime negotiating resource in such a process. I turn now to a detailed discussion of pronouns and footing.

Pronouns and Footing A pragmatic perspective views pronouns as doing more than referencing work. For one reason, pronouns possess a flexible referential range. Wales (1996, p. 63) argues that the English we is seemingly “limitless” in referents. This makes it difficult to pinpoint the referent in many cases but, at the same time, opens up a possibility for pragmatic manipulations. For instance, the speaker-oriented pronoun I can distance the speaker’s point of view when the anchorage shifts away from the speaker,

158

K. Chaemsaithong

thereby serving as a strategy for evasive criticism, as in “I shouldn’t do that, if I were you” (Wales, 1996, p. 70). In addition to their referential flexibility, pronouns can occur in media res, without any antecedents, in which case the interlocutor has to deduce the referent (Emmot, 1997). What this means is that a referent is to be indexically presupposed or created by making relevant aspects of the context, speaker intention or world knowledge, and calling those aspects into existence (Silverstein, 1976, p. 34). As Sacks (1995, p. 148f) points out, “‘[W]e’ can clearly refer to a category, which has as one of its crucial properties that no intention exists of listing the incumbents, and furthermore they’re not listable”. Because participants in the interaction often are members of different social groups at the same time, the employment of we may be addressing intersecting memberships which are not necessarily otherwise nameable or pre-defined, but rather get constructed or delineated in the discourse itself. To add to this creative function, pronouns often carry an ideological bent because they are deeply embedded in naming people and groups that the user wishes to (dis)affiliate with and, as such, imply relations of power. Pennycook (1994, p.174) calls this aspect the “politics of the pronouns”. If we claims authority and communality, it further constructs an ‘I/we’, ‘we/you’ and ‘we/they’ dichotomy. Thus these two pronouns must always be understood with reference to other assumptions about who is being defined as the we from which the you and they differ (Pennycook, 1994, p. 176). That pronouns allow for different configurations self-other positioning can be understood in the context of Goffman’s footing (1981), which refers to the roles that speakers assume in a particular setting. According to Goffman, speakers consistently adapt the ways they participate in conversation and shape their talk in relation to the material being presented and to anticipated reactions from the interlocutor. Conversational participation is, therefore, not a simple affair in which one party functions as a speaker (and the other party as the listener), but actually consists of different kinds of participants and speaking perspectives. Occurring quite frequently and continually, such shifts in speaking position embody “change in the alignment we take up to ourselves and to others present as expressed in the way we manage the production or reception of an utterance” (Goffman, 1981, p. 128). This is because we rarely speak just “as ourselves” but naturally and quite imperceptibly oscillate between various perspectives: our words also draw on issues about who is responsible for what we say, whose words we are using and who is saying it. In the courtroom, a lawyer may, thus, speak as an individual, as a representative of his team, as the client he represents (e.g. when he reproduces his client’s words), or on behalf of legal professionals as a whole. A lawyer may fulfill all these roles, but does not necessarily have to and may not do so simultaneously. When a particular footing is enacted by a pronoun, it has social and moral implications, including who is a member of what social group and what responsibility and expectations this membership involves. The pronoun I indicates a high degree of involvement of the speaker with a topic or commitment to “authorship”, but does not say anything about the speaker’s relationship to others. When switching from I to we, a speaker can give indications on authorship as well as identification with

Pragmatics of Self-Reference Pronouns in Capital Trials

159

others (De Fina, 1995, p. 384). By using self-reference pronouns to present oneself as, or to align with certain categories of people who are recognized as, knowledgeable or having official credentials, the speaker can benefit from “category entitlement” to knowledge claims and privileged experience associated with that person or group (Potter, 1996, p. 133). In effect, a speaker can assume an authoritative speaking position and manage accountability for what is being said. Note that this is contingent upon whether the domain of reference the user is trying to construct is accepted or not. If accepted, the individual will be treated as having such a social role or identity. If not, the individual will fail to claim that identity. With the above pragmatic potentials of pronouns in mind, I now turn to my investigation of the functions of pronouns in closing arguments in the penalty phase of capital trials, and the ways they enable lawyers to construct themselves and manage interaction with the jurors.

Data and Method The closing arguments under study are based on the official transcripts of five capital cases (34,783 words). All the cases involve multiple killings, with the exception of Case 5, a rape and murder case with one victim. Although from the same state, these closing arguments are from different counties and were delivered by different lawyers. This helps ensure that the patterns found are not the linguistic idiosyncrasies of only a few lawyers. The trials occurred within a relatively contemporary time frame, from the year 2000 to the present, during which there have been no significant changes in the legislature or positions of the Supreme Court regarding death penalty, which otherwise might affect the linguistic patterns. Methodologically, the corpus was analyzed first by using the software AntConc 3.3.5 m to identify and obtain the frequencies of first-person pronouns (along with their grammatical variants). Each instance was then carefully examined to exclude those pronouns that do not belong to the lawyer’s expression of self (e.g., in reported speech). All the frequency counts were then normalized to a common basis of 10,000 words to allow for comparison for different text lengths. The qualitative, discourse-oriented analysis was finally performed to classify the pragmatic functions of the pronouns and explicate the ways in which they are indexical of the underlying ideologies of the presenter, thereby aiding in self/other positioning and the shaping of jurors’ perceptions of the defendants and the death penalty.

Findings As shown in Table  1, the overall frequency counts indicate that first-person pronouns are an integral part of the construction of the closing arguments in capital trials. More specifically, as shown Table 2, first-person pronouns appear in all the

160

K. Chaemsaithong

Table 1  Distribution of first-person pronouns in the corpus (per 10,000 words) Prosecutor (hereafter “Pro”) 306.72 13.05 10.79 – 330.56 248.88 32.08 27.20 – 308.16

I Me My Mine Total We Us Our Ours Total

Defense (hereafter “Def”) 719.90 33.34 76.18 – 829.42 664.12 61.58 57.25 – 782.95

Table 2  Distribution of first-person pronouns in each trial

I Me My Mine Total We Us Our Ours Total

Case 1 Pro – – – – 0 28.27 11.33 11.33 – 50.93

Def 116.93 7.54 22.88 – 147.35 171.44 12.44 5.44 – 189.32

Case 2 Pro 36.65 4.08 – – 40.73 81.44 13.33 – – 94.77

Def 83.13 – 9.98 – 93.11 173.83 13.74 7.24 – 194.81

Case 3 Pro 82.62 6.70 6.70 – 96.02 61.87 – 10.72 – 72.59

Def 277.47 24.63 29.58 – 331.68 130.24 25.61 29.16 – 185.01

Case 4 Pro 89.98 – 4.09 – 94.07 27.40 – – – 27.40

Def 128.53 1.17 9.35 – 139.05 122.69 9.79 11.03 – 143.51

Case 5 Pro 97.47 2.27 – – 99.74 49.90 7.42 5.15 – 62.47

Def 113.84 – 4.39 – 118.23 65.92 – 4.38 – 70.30

cases, although I note with interest that the singular form does not appear in the prosecution’s speech in Case 1. This particular lawyer prefers displaying a collective stance to the exclusion of a personal one. Moreover, first-person pronouns appear to concentrate in the defense’s speech. That is, the defense uses both we and I more frequently than the prosecution does. Interestingly, such a distribution pattern is replicated at the level of individual trials, as Table 2 shows. What this means is that the defense views constructing the self as part of a collectivity (either including or excluding the jurors) as a more effective way to stamp authority and credibility onto their argument than overtly presenting their personal voice. With respect to the form, the subjective case predominates, while certain forms are scarce (such as the possessive forms mine and ours), as compared to their possessive determiner and objective case counterparts (my/me and our/us). This may be attributable to the lawyers’ perceived need to make clear and highlight the referent lexically, rather than pronominally. Having presented the quantitative findings, I turn now to the qualitative results, which go beyond sheer frequencies, and discuss when and why specific footings are invoked in relation to the presenter’s ideological basis.

Pragmatics of Self-Reference Pronouns in Capital Trials

161

Functional Roles of I Based on the data in my corpus, five footings emerge and are signaled by I: I as an Interlocutor: The closing argument is, objectively speaking, a monolog, and the institutional conventions strictly constrain the conversational roles of both the lawyer and jurors so that they cannot talk to each other like in ordinary talk. However, when assuming this footing or speaking role, lawyers can show explicit recognition of the presence of the jurors and connect to them, so as to pull them along with their argument, focus their attention, and guide them to interpretations. In effect, this role turns silent addressees into co-constructors of the discourse. This footing usually involves different kinds of speech acts, such as challenging (1a) and requesting (1b). Note in these examples the paired use of I and you, indicative of a dyadic relationship between the lawyer and the silent jury. (1a) The depravity, the horror, I would challenge you to think, have you ever heard of a murder that you’ve heard in the news…that was more heinous and more deserving of the death penalty than this case…I would challenge you to think of a crime you have ever heard of that is ore horrific…and more deserving of death penalty than the facts in this case. (Case 3 Pro) (1b) The defense said…that we’re going to show you photographs…I ask, of those 80 or so photos, that you remember three, Donna, Avery, and Sydne. (Case 2 Pro) Often co-occurring with you, the lawyers assuming this role also provide metadiscursive comments to the jurors. The prosecution in (2a), for example, uses “I mean”, whose function is to “indicate upcoming adjustments” (Schiffrin, 1987; Fox Tree and Schrock, 2002, p.  741) to reformulate the previous rhetorical question as “throwing a feather on a scale with a brick” and “one excuse after another”, thereby nullifying the other side’s argument. The defense in (2b) embarks on a clarifying move, assuring the jurors that their stance is not to excuse the murder. In doing so, the lawyer can give a more nuanced explanation of the defendant’s agency, from a violent murderer to a public nuisance. (2a) Did you hear anything that justifies the behavior that occurred on February 24, 2005? I mean it’s like throwing a feather on a scale with a brick...The abuse excuse...toxic parenting, bad report cards...His mother had chemotherapy...I mean it’s one excuse after another. (Case 3 Pro) (2b) Now, ladies and gentlemen, please don’t misunderstand. I’m not saying it’s okay to commit these kind of crimes; I’m just trying to illustrate that these kinds of crimes aren’t in a category of violent crimes. They’re crimes against the public order and crimes against decency. Roy Ward [defendant] was a public nuisance. (Case 4 Def) I as an Opinion Holder: In this role, first-person singular pronouns come to indicate the lawyer’s personal, rather than collective, authorial stance toward the truth of a proposition and the extent he wants to epistemically commit himself to it. The

162

K. Chaemsaithong

prosecutor in (3a), for instance, overtly signals disagreement with the mitigating factors presented by the other side, using “I don’t know” and “I don’t see”. Here, by declaring insufficient knowledge, the lawyer in fact prefaces his disagreement and dispreferred response to the other side’s claims (Diani, 2004; Tsui, 1991). In a similar way, the defense in (3b) expresses disagreement with the death sentence as the appropriate punishment, and further expresses his conviction in the claim that capital punishment (through “I’m sure”) will not bring an emotional closure to the victim’s family. (3a) So the mental claims, mental issues, those aren’t issues in this case…It’s very hard for me to comment on that at this point in time… If you find the defendant has proven that he maintained a period of gainful employment, that may be considered as a mitigating factor. I don’t know why this one was in here. In fact, it was placed in here, I don’t know what evidence was put on to support it. Because the evidence that we received…I don’t see how that mitigates murder. I don’t see how that has anything to do to reduce or excuse what he did to Kelly Eckart (Case 5 Pro) (3b) I don’t know that there’s any healing to be done in this case. I’m sure there is not. But I’m equally sure that what will not heal is a killing by the State. Because of all kinds of killing, whatever they are and whatever the justification or excuse, killing by the State is the most ceremonious. It’s ceremonious. We make it into a ceremony. (Case 3 Def) I as a Lawyer: Lawyers assume this role when they wish to position themselves as an authoritative figure on the issue being presented. By drawing upon their professional experience or knowledge, they achieve what Sacks (1995) calls “category entitlement”. In some cases, this entitlement claim is explicit and institutionally-­ warranted as in (4a), where the prosecution first starts off with his identity as a prosecutor to argue the person being tried is the worst criminal in the state (before subsequently constructing a collectivity with other prosecuting lawyers through we). In other cases, the knowledge claim is rhetorically created and negotiated as the discourse unfolds, as in (4b), in which the defense vouches for the defendant’s “real” character through personal relationship and familiarity during the past 18 months as a defense lawyer. (4a) In my career in law enforcement in this community, we have had at least 125 murders....Of those 125 or so, no murder even comes close to the murders committed by [defendant]. (Case 3 Pro) (4b) The Danny that committed these crimes is not the true Daniel Wilkes, not the true Daniel Wilkes that I’ve gotten to know over the past 18 months. I submit to you and my message to you is that this is the Daniel Wilkes that myself and Barbara Williams have gotten to know. (Case 2 Def) I as a Lay Person: This role enables the lawyer to step out of the institutional identity and transforms himself into a member of society just like the jurors, thereby

Pragmatics of Self-Reference Pronouns in Capital Trials

163

creating solidarity with them. This role also disguises any argumentative attempts he is trying to achieve. Interestingly, this footing does not emerge frequently in the corpus, but is found in both side’s speech. In (5a), speaking as a parent, the prosecution supports the argument that the defendant could have made a choice to not kill despite the problematic childhood, but he did not choose that choice, and thus he deserves a death sentence. In contrast, although similarly speaking from the same parental position, the defense lawyer in (5b) first expresses empathy for the losses brought about by the crime, just to subsequently discourage that emotional reaction and the death sentence, as he shifts into an institutional role in the concessive clause headed by “but”. (5a) You know, I’ve got a year-old daughter who watches MTV, and she wonders why she can’t have her own jet and why she doesn’t have a house like the rappers on MTV. Life’s tough. You know what, deal with the situation that you’re in. And because he had a tough background, it was harder for him to make choices. He still had a choice. (Case 3 Pro) (5b) And I’m also a parent, ladies and gentlemen, like most of you are, and when I saw these photographs and when I read about this case, if this was my child, I’d want to kill him. There’s no doubt in my mind I’d want to kill him (indicating), but that’s not what our system is all about, ladies and gentlemen. (Case 4 Def) I as Another Individual: In contrast to the other roles described above, this use of first-person singular pronouns shifts the reference from self to another individual. Found in both the prosecution’s and defense’s speech, this footing allows the lawyer to identify with a position that is not his own and to perform another character in such a way that fits with his argument. In both examples below, the lawyers assume the position of the defendant and speak as that character. In (6a), by assuming the role of the defendant, the prosecutor invalidates the mitigating factors presented by the other side. In (6b), the defense can explain what drives the defendant to committing an illegal act as a mitigating factor. (6a) This is called the abuse excuse. I had a tough childhood. So don’t sentence me to death. It’s the same thing that’s been running through this case since it started. How do I get myself out of the problem that I’m in? How do I make my penalty the least that it can possibly be? He’s thirty-two years old when this crime was committed. (Case 3 Pro) (6b) The few criminals that I’ve ever met...did not act in a moment’s passion, but planned out their torture and their activity, their attitude was always it’s you folks out there against me. I know what’s going to happen to me. You’re going to kill me if you catch me, and therefore, I’m going to be as brutal as I perceive that society has been to me. It’s amazing to me that some of the attitudes that criminals have. (Case 3 Def)

164

K. Chaemsaithong

Functional Roles of We We is found to enact seven different footings, as discussed below. We as a Team: This role is assumed when we references the presenter along with his team. In much the same way as “I as an opinion holder”, this speaking perspective allows the presenting lawyer to declare his side’s united view on the suitable punishment and to show his side’s commitment to the validity of a proposition (7a). In addition, the lawyer can bestow authority upon his side by claiming possession of exclusive knowledge, experience, and evidence pertaining to the case, as exemplified in (7b) (7a) We don’t believe the mitigation that we presented on his behalf should outweigh the horrificness of this crime, but we believe that his life should be spared and that the penalty of life without parole will be sufficient to punish him. The penalty of never, ever getting out of prison will be sufficient to punish him. (Case 4 Def) (7b) You’ve heard how that murder was an intentional murder. We talked about the intent to take the shoe lace out of Kelly’s shoe, and deliberately remove it and tie it around her neck. (Case 5 Pro) In the above examples, by speaking on behalf of the team, the lawyer is not a mere “animator” of the statement, but is a “principal” (Goffman, 1981) who represents the entire side as responsible for the message. Everyone in that side views the issue in the same direction and orchestrates with each other in the presentation of the case. We as Legal Professionals: This footing contributes to forming a collectivity that extends beyond just the presenter’s side to include members of the same profession (who may have the same responsibility and standpoint). Consequently, the lawyer can add authority to a particular statement and enhances their authority on the issue being discussion. As in (8a), the defense lawyer shifts into this role to seeks empathy for the defender as a whole when representing a capital defendant. (8a) I’ve been doing criminal defense work for 23 years. It’s never easy for any of us to take a case… (Case 4 Def) Sometimes, this professional collectivity may include the other side. In effect, the presenter can erase opposition between the two sides, and in doing so, the lawyer can present a statement as being agreed upon by both sides, as in (8b). (8b) We are all talking about anger. We are all talking about leaving marriage. (Case 1 Def) We as a Group with Jurors: Including the jurors, the presenting lawyer aims to construct a single homogeneous unit. Both the prosecution and defense shift into this role when presenting a statement as if the presenter and the jurors share the same experience, knowledge, and information. In these cases, it is worth stressing that this homogenous unit (and by extension the knowledge and experience) is not necessarily a socially recognized group, but is instead created in that moment. In

Pragmatics of Self-Reference Pronouns in Capital Trials

165

effect, the jurors are construed as being in agreement with the lawyer’s argument. Notably, this role is initiated in the presence of a saying or mental verb. In (9a), the prosecution emphasizes the violence and the aftermath of the murder by positioning the jurors as being aware of the fact that the case involves multiple casualties, while in (9b), the defense is reminding the jurors of the shared knowledge of the term “neuron”, a metaphor used to describe the troubled childhood experience of the defendant as a target. (9a) Multiple killing dictates, tells us, Mr. Isom eliminated a family from this earth. (Case 1 Pro) (9b) Now we know in 1989 that Julia died, the person that he [the defendant] called mom...But heaven forbid you’re a neuron....We learned about it, didn’t we? If you’re a neuron, means you’re a target from everybody. (Case 1 Def) We as Courtroom Participants: In this capacity, the group being constructed does not only consist of the jurors, but also includes other participants in the courtroom, that is, the other side, judge, and witnesses. By involving everyone in the current context, the presenting lawyer creates a shared courtroom knowledge or experience. Such a shared experience may be construed negatively, as in (10a), where the number of service hours and the repetition of “anything” indicate not a fruitful session. Alternatively, the shared experience can be one that is intended to ensure justice and an appropriate sentence, as in (10b). (10a) The defense is permitted to say anything they want. That’s why we’re here for seven or eight hours listening to Dr. Clark because anything, anything, anything, anything can be presented to you. (Case 3 Pro) (10b) However, if you take Mr. Cooper’s argument, because we had a first phase and you found the defendant guilty, you take Mr. Cooper’s argument, then I don’t know why we even have a second phase...The law...requires that we have this phase and you be instructed by the judge on what mitigating circumstances are, and that mitigating circumstances are not excuses or justifications. (Case 4 Def) We as Society Members: By expanding the referential domain to social members outside the courtroom, including the local geographical setting or even the entire nation, the lawyer draws upon communally held social norms or values in this role. One of the consequences of this deictic center expansion is that the discourse becomes more bureaucratic, thereby making empathy a secondary concern. As (11a) illustrates, speaking from a State’s perspective, the prosecution is framing the person on trial as deserving of a death sentence because it involves multiple killings. (11a) Not every murder qualifies for the death penalty in our state...If you kill a person and kill another person at any time, that qualifies you for the death penalty. (Case 3 Pro) In contrast, assuming a similar collective position, the defense stresses that it is the society that should be responsible for the crime and what caused the defendant to commit the crime, thereby invoking empathy, as (11b) shows.

166

K. Chaemsaithong

(11b) If you think, considering your life versus Michael Bear’s life, we need to kill him, that’s your choice. We didn’t do much for him growing up but we’re ready to kill him. That’s your choice. (Case 3 Def) We as Human Beings: Referring generically to refer to no group in particular, this use of we foregrounds how a particular situation applies to everyone and belongs to everyone, as if it is common sense or universal, thereby elevating a proposition to a generality. The prosecution takes advantage of that commonality for the defendant to accept responsibility as human beings, but it is the defendant’s choice to commit a crime and therefore deserves a life sentence, as in (12a). (12a) He wouldn’t have had to kill Donna...He had a choice, like all of us do... Everyone of us are born with a clean slate. (Case 2 Pro) The defense, in contrast, stresses a common human proclivity for revenge, which contrasts with compassion, as in (12b): (12b) Basically what it [death penalty] does is it scratches that primordial, primitive part of us who wants to kill out of self-defense. (Case 3 Def) We as the Other Side: In this case, the deictic center is shifted to the other side, so that the presenter can argumentatively respond to the other side’s argument, or reanimate it in order to subsequently invalidate—even denigrate—the argument. This role was not found in the defense’s speech. In (13a) and (13b), the pronoun we pragmatically delimits the referent to the opposing side, as it is the defense that employed the so-called abuse excuse, and the prosecution shows disaffiliation with such a mitigating factor. (13a) When are we going to stop using the abuse excuse? (Case 5 Pro) ( 13b) Isn’t it time for him [defendant] to take some responsibility for his own behavior?...When do we become responsible for that? When do we stop crying about a tough childhood we have? (Case 3 Pro)

Conclusion The overall aim of this project is to analyze the strategic use of first-person pronouns and the footings activated by these pronominal forms, allowing the lawyers to achieve their communicative goals in the closing speech of the penalty phase. From a distributional perspective, the use of we is generally more frequent than I in my corpus, and is found in the defense’s speech much more than the prosecution’s. The quantitative findings overall support previous studies that argue that the closing argument is a performance, with the presenting lawyer switching between different positions. However, the findings stand in contrast to Rosulek’s study (2015), which reports more singular-first person pronouns than plural ones in the closing arguments of ordinary criminal trials. What this means is that lawyers in the penalty phase tend to exploit the very ambiguity and indeterminacy of we to achieve particular pragmatic functions by (dis)associating themselves with a certain

Pragmatics of Self-Reference Pronouns in Capital Trials

167

collectivity. To a lesser extent, the lawyers also project different kinds of individual aspects into the speech. This study also identifies the footings that I and we can activate, but which have not been discussed in the existing literature. In line with previous research on pronouns, this study reveals that first-person pronouns do not form neat categorical divisions, but they can signal a variety of footings. Specifically, five footings for I and seven for we have been identified. Both I and we, thus, serve to designate a range of individuals moving outwards from the presenting lawyer himself to the presenter plus the jurors, other participants of the courtroom, social members, other individuals, and human beings as a whole. In addition, they can be called upon to perform other social actors and voice their positions as well. Each side is observed to make use of these pronouns to good effect: to claim knowledge and experience of the issue at hand, thereby creating solidarity with or distance from the defendant; to intervene into the text and leave authorial assessment and evaluation; and to interpersonally connect with other courtroom participants, social members, and humans in general. In some cases, the pronouns construct and foreground differences between the defendant and jurors, to the extent of deserving a death sentence. In more practical terms, while the study does not offer a predictive outcome as to which combination of footings will convince the jury, as this issue depends on many other factors (e.g. reference terms that may work synergistically with pronouns), the study does show how self-referencing facilitates the ensuing death sentence. With a view toward a more humanistic judicial system, the findings in this paper, therefore, have implications for training jurors and the public at large to be aware and more critical of the power the seemingly insignificant (but indeed complexly orchestrated) linguistic choices in constructing and shaping courtroom experiences and outcomes.

Appendix: Data Sources Case 1: State of Indiana v. Kevin Charles Isom (2013) Case 2: State of Indiana v. Daniel Ray Wilkes (2007) Case 3: State of Indiana v. Frederick M. Baer (2002) Case 4: State of Indiana v. Roy Lee Ward (2002) Case 5: State of Indiana v. Michael D. Overstreet (2000)

References Amsterdam, A., & Hertz, R. (1992). An analysis of closing arguments to a jury. New York Law School Law Review, 37, 55–122. Bull, P., & Fetzer, A. (2006). Who are we and who are you?: The strategic use of forms of address in political interviews. Text & Talk, 26, 3–37. https://doi.org/10.1515/TEXT.2006.002.

168

K. Chaemsaithong

Burt, M. (2008). The importance of storytelling at all stages of a capital case. UMKC Law Review, 77, 877–910. Conley, R. (2016). Confronting the death penalty: How language influence jurors’ in capital cases. Oxford, UK: Oxford University Press. Costanzo, M., & Peterson, J. (1994). Attorney persuasion in the capital penalty phase: A content analysis of closing arguments. Journal of Social Issues, 50, 125–147. https://doi. org/10.1111/j.1540-4560.1994.tb02413.x. De Fina, A. (1995). Pronominal choice, identity and solidarity in political discourse. Text, 15, 379–410. https://doi.org/10.1515/text.1.1995.15.3.379. De Marco, A., & Palumbo, M. (2016). Identity through discourse: The use of pronouns in the narration of Italian emigrants. In S. Guzzo & D. Britain (Eds.), Languaging diversity, Vol. 2: Variationist approaches and identities (pp.  19–38). Newcastle upon Tyne, UK: Cambridge Scholars Publishing. Diani, G. (2004). The discourse functions of I don’t know in English conversation. In K. Aijmer & A.  Strenstrom (Eds.), Discourse patterns in spoken and written corpora (pp.  157–171). Amsterdam, Netherlands: John Benjamins. Emmot, C. (1997). Narrative comprehension: A discourse perspective. Oxford, UK: Oxford University Press. Fox Tree, J., & Schrock, J. (2002). Basic meanings of ‘you know’ and ‘I mean. Journal of Pragmatics, 34, 727–747. https://doi.org/10.1016/S0378-2166(02)00027-9. Gardelle, L., & Sorlin, S. (2015). Personal pronouns: An exposition. In L. Gardelle & S. Sorlin (Eds.), The pragmatics of personal pronouns (pp. 1–23) (pp. 1–23). Amsterdam, Netherlands: John Benjamins. Goffman, E. (1981). Footing. In E. Goffman (Ed.), Forms of talk (pp. 124–159). Philadelphia, PA: University of Pennsylvania Press. Levin, F., Papantonio, M., & Levin, M. (2003). Closing arguments: The last battle. Pensacola, FL: Seville. Pennycook, A. (1994). The politics of pronouns. ELT Journal, 48, 13–18. Potter, J. (1996). Representing reality: Discourse, rhetoric, and social construction. London, UK: Sage. Proctor, K., & Su, L. (2011). The 1st person plural in political discourse—American politicians in interviews and in a debate. Journal of Pragmatics, 43, 3251–3266. https://doi.org/10.1016/j. pragma.2011.06.010. Rosulek, L. (2015). Dueling discourses: The construction of reality in closing arguments. Oxford, UK: Oxford University Press. Sacks, H. (1995). Lectures on conversation Vol. 1 (edited by G. Jefferson). Oxford, UK: Blackwell. Schiffrin, D. (1987). Discourse markers. Cambridge, UK: Cambridge University Press. Silverstein, M. (1976). Shifters, linguistic categories, and cultural description. In K.  Basson & H. Selby (Eds.), Meaning in anthropology (pp. 11–55). Albuquerque, NM: University of New Mexico Press. Tang, R., & John, S. (1999). The ‘I’ in identity: Exploring writer identity in student academic writing through the first person pronoun. English for Specific Purposes, 18, S23–S39. https://doi. org/10.1016/S0889-4906(99)00009-5. Tsui, A. (1991). The pragmatic functions of I don’t know. Text & Talk, 11, 607–622. https://doi. org/10.1515/text.1.1991.11.4.607. Wales, K. (1996). Personal pronouns in present-day English. Cambridge, UK: Cambridge University Press. Walter, B. (1988). The jury summation as speech genre: An ethnographic study of what it means to those who use it. Amsterdam, Netherlands: John Benjamins. Wilson, J. (1990). Politically speaking: The pragmatic analysis of political language. Oxford, UK: Blackwell.

Navigating Narrative Subjectivity in Schizophrenia: A Deictic Network Analysis of Narrative Viewpoints of Self and Other Linde van Schuppen, José Sanders, and Kobie van Krieken

Introduction When attempting to comprehend an enigmatic disorder such as schizophrenia, it is essential to gain an understanding of the way self, other and world are experienced by the people diagnosed with it (Sass and Parnas, 2007). Research into the structures of subjectivity that characterize schizophrenia may prove especially important since it is often construed as a ‘self-disorder’ (Mishara et al., 2014). Schizophrenic subjectivity, though, is difficult to study if one has had no experience of one’s own, and phenomenological research has largely been dependent upon a limited number of memoires (e.g. Sechehaye, 1951; Schreber, 1903/2000) as its main source. This paper starts from the idea that the linguistic study of schizophrenic narratives might provide a valuable source for studying (inter)subjectivity in people with a schizophrenia diagnosis. There has been a recent surge towards theories that see language and thought as tightly interwoven and therefore inseparable, or even construe language as the configurator of human-specific thought (for an overview see Hinzen, 2017). These views imply that studying language equals studying the mind. But even if their relationship should be less intertwined, language use can be seen as reflective of general cognitive functions like perception, intentionality, memory, and perspective-taking, among other phenomena (Zimmerer et al., 2017). Since language is the most significant vehicle of expressing and sharing experiences and the way language is used can reveal many things about the inner world of the Speaker, linguistic research on the stories and interactions of people with a schizophrenia diagnosis might contribute to our knowledge of schizophrenic (inter)subjectivity.

L. van Schuppen (*) · J. Sanders · K. van Krieken Centre for Language Studies, Radboud University, Nijmegen, Netherlands e-mail: [email protected]; [email protected]; [email protected] © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 F. Macagno, A. Capone (eds.), Inquiries in Philosophical Pragmatics, Perspectives in Pragmatics, Philosophy & Psychology 28, https://doi.org/10.1007/978-3-030-56696-8_10

169

170

L. van Schuppen et al.

We propose that the concept of ‘viewpoint’ or ‘perspective’ is especially suitable for studying (inter)subjective experience and cognition through language, since it can be studied as both a cognitive, experiential and as a linguistic phenomenon. This follows from the fact that all cognitive and perceptual acts are inherently performed by a body that can only be situated at one place at one time, thereby saturating experience with a viewpoint per definition.1 Viewpoint is fundamentally reflected in language use and story-telling, and the focus of a large body of linguistic literature (e.g. Dancygier and Sweetser, 2005; Zeman, 2017). In addition, viewpoint is of particular interest in research on schizophrenic language because people with a schizophrenia diagnosis are considered to have a reduced capacity to accurately understand and theorize about other perspectives (Corcoran et al., 1997; Brune, 2005; Langdon and Ward, 2009). Along these lines, cognitive deficits in recognizing and correctly representing and distinguishing different perspectives are expected to be reflected in language use (van Schuppen et al., 2019). Empirical research on viewpoint indicators in schizophrenic language is still in its infancy, however, and appears to be isolated from both theoretical (philosophical) accounts of subjectivity and the phenomenology of schizophrenia, making it hard to interpret these studies’ findings in an informed way. In this paper we establish connections between the philosophical literature on schizophrenia and linguistic research on the cognitive and pragmatic aspects of perspective-taking in discourse. We aim to advance the understanding of the nature and complexity of perspectivization issues in people with a schizophrenia diagnosis by analyzing the linguistic manifestation of viewpoint in an authentic conversational patient narrative. As such, its goal is not so much to systematically compare two groups of language users, i.e. schizophrenia diagnosed versus neurotypical people, but rather to investigate the potential that studying viewpoint in narrative has in enriching our current research on (inter)subjectivity and perspective taking in a variety of language users, and attempt to do so by connecting phenomenological theory to an empirical approach in a mutually beneficial way. As a result, our study departs from an exploratory angle, demonstrating that the proposed methodology can successfully pinpoint viewpoint anomalies in the narratives of non-neurotypical storytellers. In what follows, we first offer a review of both philosophical and linguistic research on viewpoint in schizophrenia and discuss a cognitive linguistic model for the analysis of viewpoint in oral stories.

Review of the Literature The literature addressing questions on the language, cognition and experience of people with a schizophrenia diagnosis seems to be divided into a theoretical body of work, grounded in philosophy, and an empirical body of work that has its roots in

1  This part of experience is sometimes indicated as the ‘mineness’, ‘what-it-is-likeness’ of experience (e.g. Zahavi, 2014).

Navigating Narrative Subjectivity in Schizophrenia

171

cognitive psychology and psycholinguistics. In this section we will briefly review a number of important findings from these different fields and argue for a need to integrate these different bodies of work. Philosophical Accounts of Schizophrenia Within the philosophy of psychiatry, theories from phenomenological psychiatry are becoming increasingly pronounced. One part of this field is trying to account for schizophrenic symptoms by describing the psychotic experience as stemming from a fundamental disturbance in the patient’s sense of self, or ‘ipseity’, explaining the different symptoms as originating in the same distortion of embodied, pre-reflective self-awareness (Sass, 2014; Parnas and Henriksen, 2013; Gallagher and Zahavi, 2008). These accounts have recently been criticized for being i) too individualistically oriented and ii) too detached from empirical reality in its reduction of the complexity and heterogeneity of the schizophrenic condition to one principle (see e.g. Kyselo, 2016; De Haan and Sanneke, 2010; commentaries by Lysaker and Hamm, 2015, Ratcliffe, 2015, and Fuchs, 2015, on Sass and Byrom, 2015). In elaboration on these comments, a number of accounts on the intersubjective dimensions of schizophrenia have been developed, which propose that self-disturbances, and the symptoms that accompany it, are fundamentally intersubjective in nature (Pienkos, 2015; Fuchs and Röhricht, 2017; Ratcliffe, 2017; van Duppen, 2016). This means that understanding psychotic experience necessarily requires an understanding of the interactive relations of the patient with other people. Fuchs and Röhricht (2017) for example, describe how people with a schizophrenia diagnosis often indicate that they experience a disengagement from their bodies and the world. Such a disembodied and alienated experience of one’s self can influence the inter-­ corporeal understanding that one has with other people, alienating the self from others and resulting in the loss of a shared world of pre-reflective meaning. Fuchs and Röhricht (2017, p. 133) construe this as a loss of ‘common sense’, i.e. the tacit knowledge and familiarity guiding the relations and interactions with others.2 This retraction from embodied interaction with others might result in the inability to navigate and flexibly shift between an ego- and allocentric perspective, implying that it might become harder to distinguish first and third person points of views (Fuchs and Röhricht, 2017, p. 130). According to this account, this development might climax in delusions, i.e., beliefs that are not, and sometimes cannot, be shared by others. Another account on intersubjectivity in schizophrenia suggests that a disruption of trust-structures between self and other could give rise to an inability to distinguish between different modalities of consciousness, resulting in psychosis 2  ‘If this embodied involvement in the world is disturbed as in schizophrenia, it will result in a fundamental alienation of intersubjectivity: the basic sense of being-with-others in a shared lifeworld gives way to a subtle “loss of natural selfevidence”, as Blankenburg (1971) has described it. Patients report that they feel isolated and detached from others, unable to grasp the natural, everyday meanings of the common world.’ (Fuchs and Röhricht, 2017, p. 133).

172

L. van Schuppen et al.

(Ratcliffe, 2017). Lastly, Pienkos (2015) and Sass and Pienkos (2015) describe various manifestations of intersubjective disruptions on the basis on different verbal descriptions that patients gave of their experiences, such as abnormalities of common sense and conventionality, pathological sense of empathy and openness, feelings of paranoia and of centrality, and perceptions of devitalization of others. These disturbances are suggested to be constitutive to a large variation of psychotic symptoms, and are expected to be reflected in the language use of schizophrenic patients (Pienkos and Sass, 2016). Notably, most of the above mentioned accounts seem to arise from a very narrow empirical foundation, often only referring to a few famous, professionally edited memoires or a small selection of quotes from interviews with patients. In addition to these phenomenological accounts that take the experience of the patient as directory and aim to understand the disorder ‘from within’, an account was developed that criticizes the division between language and thought that is often implied in cognitive science; here, schizophrenia is approached as primarily a disorder of language (Hinzen, 2017; Hinzen and Rosselló, 2015). Three main symptoms of schizophrenia are viewed as failures in language-mediated forms of meaning: disorder of speech perception (Auditory Verbal Hallucinations), abnormal speech production running without feedback control (Formal Thought Disorder), and production of abnormal linguistic content (Delusions). In this approach, findings of cognitive psychology, psychiatry and linguistics are brought together to gain a better understanding of schizophrenic symptoms (see also Zimmerer et al., 2017). Empirical Research on Language in Schizophrenia Although researchers have been interested in the way that people with schizophrenic symptoms use and understand language since the early twentieth century (Kraepelin, 1919), the language use of people with a schizophrenia diagnosis remains notoriously difficult to characterize. In 1961, Dr. Maria Lorenz formulated the problem in the following way: ‘“Schizophrenic language” can be correctly designated as both concrete and as abstract; as restricted; impoverished, and as fluid, overideational; as empty of meaning and as overinclusive of meaning; as resembling prelogical thinking and as metaphoric and symbolic’ (Lorenz, 1961: 603)

In the past fifty years, several studies have been conducted into the language of schizophrenia, analyzing writing samples, first person illness narratives, and/or (clinical) interviews with people who are diagnosed with schizophrenia. For example, a number of review studies found deficits in both language production, as well as comprehension (Delisi, 2001; Kuperberg, 2010), especially in the areas of syntax (Moro et al., 2015) and high-level semantics (Barrera et al., 2005). There has been a surge in the use of computational, automated methods of analysis, with several studies focusing on a decrease in sentence complexity and coherence (e.g. Willits et al., 2018; Elvevåg et al., 2007; Saavedra, 2010), as well as a decrease in narrative

Navigating Narrative Subjectivity in Schizophrenia

173

coherence (Allé et al., 2015; Lysaker et al., 2002; Holm et al., 2016) and incomprehensibility (Saavedra et al., 2009). A recent study by Bedi et al. (2015) even predicted psychosis development in 11 high-risk youths with 100% accuracy, using automated analysis of semantic coherence and two syntactic markers of speech complexity in interview transcripts. These results indicate that the study of schizophrenic language has a lot of (possibly diagnostic) potential, without being specifically grounded in understanding schizophrenia as a disturbance of subjectivity. Yet other studies do provide indications of the relation between language and subjectivity disturbances. These studies are concerned with language phenomena that are essentially manifestations of viewpoint, such as self and other indication by means of reference and pronoun distribution. This is of interest because references can only be understood with regards to Speaker (self) and Hearer (other) viewpoint. In order to make a reference understood, the Speaker must, for example, track what the Hearer knows, intents and believes, and take into account what she can see and hear. The same kind of ‘tracking’ is necessary for the Hearer to be able to understand reference to yet other subjects. The skillful navigation of such different subjective viewpoints, i.e. using reference in such a way that it is clear for the Hearer to understand who or what is being referred to, is therefore clearly pronounced in the way referents are linguistically expressed. In particular, pronoun use is indicative of the way that the Speaker makes the perspectives of the available subjects explicit. Interestingly, patients with ‘Formal Thought Disorder’, one of the main diagnostic criteria for schizophrenia, were for example found to display significantly more instances of ‘unclear reference’ (Rochester and Martin, 1979; Harvey, 1983; Chaika and Lambe, 1989), while deviancies in the use of demonstrative and personal references were also found to correlate with this disorder (Barch and Berenbaum, 1996). Following up on these early findings, more recent studies employed word count methodologies to shed light on the frequency of referential phenomena. These studies have resulted in a hybrid picture. For example, people with a schizophrenia diagnosis were found to make more (e.g. Hong et al., 2015; Buck and Penn, 2015) or rather less (Deutsch-Link, 2016) use of self-referential pronouns (‘I’) than people without the diagnosis. Compared to people with a mood disorder, they were found to make less use of first person pronouns (e.g. Fineberg et al., 2015). Other studies found an increase in use of the pronoun ‘you’ by schizophrenia patients (e.g. Watson et al., 2012), whereas Buck et  al. (2015) found that lower social cognition measures, characteristic of schizophrenia (Brune, 2005; Langdon and Ward, 2009), were significantly correlated with a decrease in second-person pronoun use as well as the overall use of pronouns. The attention for reference within schizophrenia can be understood in the light of the fact that reference is one of the most pronounced characteristics of pragmatics, i.e., the use of language in its communicative context. Pragmatic abilities, especially comprehending discourse and non-literal meaning, were found to be compromised in a large majority of people with a schizophrenia diagnosis (Bambini et al., 2016; see also Covington et al., 2005; Cummings, 2014). More specifically, patients seem to have trouble producing contextually appropriate speech, and inferring

174

L. van Schuppen et al.

context-­depending meanings, like metaphors and irony (Langdon et  al., 2002; Kiang et  al., 2007; Schettino et  al., 2010; Rapp et  al., 2014; Cummings, 2014; Pesciarelli et al., 2014; Tavano et al., 2008; Mossaheb et al., 2014). In addition, a number of studies found that patients have difficulty maintaining thematic coherence and respecting conversational conventions (Linscott, 2005; Perlini et al., 2012; Haas et al., 2014; Elvevåg et al., 2007); such pragmatic deficits have been linked to either executive functioning or Theory of Mind (Langdon et al., 2002; Brune, 2005; Champagne-Lavau and Stip, 2010; Thoma et al., 2009). To conclude, “half a century of investigations of language in SZ [schizophrenia] leave the impression that a principled notion is missing that would relate language to symptoms, but not that data are lacking that might support such a notion” (Hinzen, 2017, p. 219; see also Cardella, 2018). It seems that an interdisciplinary approach is opportune; we therefore aim for the connection between philosophical and linguistic approaches.

 onnecting Philosophical and Linguistic Research C on Schizophrenia In our view, both theoretical research in need of empirical grounding, and empirical work wanting for a theoretical framework within which to interpret patterns in language, can benefit from acknowledging the significance of the concept of perspective or viewpoint. Considering that this concept is a fundamental principle underlying and pervading subjective experiences, cognition, and language, it transcends the boundaries of the two disciplines and might therefore be able to bridge the gap between them. Each individual has a unique physical perspective, or viewpoint, from which objects, events and scenes are encountered. This physical perspective determines our subjective experiences, for it restricts perception. Our cognitive architecture is far less restrictive in its representation of perspective. Humans are cognitively able to ‘look beyond’ their physical perspective and to represent scenes and events from different angles. These angles need not be (purely) visual or spatiotemporal; perspective also encompasses a cognitive, emotional, and moral dimension (cf. van Krieken et al., 2017). Crucially, we are constantly aware of our own perspective but also of that of the persons we interact with (Sweetser, 2012). For example, an object within our reach does not have to be within reach for someone else, and in day-today interaction, we (subconsciously) continuously estimate from whose perspective the object is within reach and from whose perspective it is beyond reach. This default mode of representing multiple viewpoints implies that perspective is inherently intersubjective (Verhagen, 2005), enabling us to mentally align or contrast our own perspective with that of others as well as to share perspectives. The cognitive management of our own perspective and that of others is reflected in language. The notion of origo elucidates how managing perspectives is

Navigating Narrative Subjectivity in Schizophrenia

175

linguistically coordinated (Bühler, 1934/1982). Origo refers to a perspective that is deictically anchored in time and space: the I-here-now. This origo or deictic center is what makes deictic expressions, such as there and yesterday, meaningful. In interpersonal conversation, interlocutors need to take into account their own deictic center and that of their communication partner in linguistically structuring their utterances in a meaningful and comprehensible way. Deictic expressions only find their meaning in their relation to an origo. Importantly, the number of deictic centers playing a role in the conversation does not have to be restricted to two: interlocutors may introduce the deictic center of a third party into their talk. In addition, they may project that deictic center, as well as their own deictic centers, in other spatiotemporal domains such as the past or imagined future. The multiplication of origo’s increases the complexity of managing perspectives, requiring interlocutors to navigate their way through a network of interconnected viewpoints. This process can be described and analyzed in a cognitive linguistic model of perspective: a Deictic Navigation Network.

Deictic Navigation Network The Deictic Navigation Network is a model for the analysis of perspective in conversational discourse (van Schuppen et al., 2019). The DNN is based on the Basic Communicative Spaces Network which takes into account two fundamental principles of human communication (Sanders et al., 2012): first, that human communication is inherently multi-viewpointed, and, second, its manifestation in recursive, potentially infinite patterns. These principles are depicted in a series of cumulative figures (Figs. 1, 2 and 3), which show a simplified visualization of the DNN. In Fig. 1, the actual communicative setting, indicated as the Speech Act Domain, is represented in which Speaker and Hearer interact and establish the discourse. In Fig. 1  Speech Act Domain of the Deictic Navigation Network

176

L. van Schuppen et al.

Fig. 2  Projection of Narrative Domain in the Deictic Navigation Network

Fig. 3  Recursion of domains in the Deictic Navigation Network

this Speech Act Domain, the Speaker needs to take into account her own perspective (“I”, here, now) and that of the Hearer (“you”, here, now) in the linguistic encoding of her utterances, represented by (1) in the figure. Typical for interpersonal interaction is the tendency to narrate about events, people and situations that are neither present in, nor observable from, the Speech Act Domain. We often talk about dreams (Schredl, 2000; Olsen et al., 2013, 245), for example, as well as memories from the past and hypothetical scenarios in the future (van Krieken et al., 2019). In doing so, we cognitively and linguistically construct a domain distinct from the here-and-now Speech Act Domain, i.e., a Narrative Domain representing past, future, hypothetical or counterfactual events and experiences. In this Narrative Domain, the Speaker can project herself as well as others, among whom the Hearer. This Narrative Domain is represented on the left side of Fig.  2. The dotted grey arrows depict the projection of deictic centers from the Speech Act Domain onto the Narrative Domain.

Navigating Narrative Subjectivity in Schizophrenia

177

Narrating about people and events in the Narrative Domain requires the Speaker to distinguish between the perspectives of Speaker and Hearer in the Speech Act Domain and the perspectives of the subjects being talked about (2). For example, the Speaker might bring up a shared past experience, in which case four different perspectives need to be acknowledged: the Speaker’s perspective in the Speech Act Domain (“I”, here, now), the Hearer’s perspective in the Speech Act Domain (“you”, here, now), the Speaker’s perspective in the Narrative Domain (“I”, there, then), and the Hearer’s perspective in the Narrative Domain (“you”, there, then). Within the Narrative Domain, the perspectives of the persons being talked about  – Speaker, Hearer, and/or Others – need to be distinguished from one another (3). Crucially, each of the subjects in the Narrative Domain can be represented as an embedded Speaker. In such cases a new, embedded Speech Act Domain is construed: a Narrative Speech Act Domain. This is visualized in Fig. 3. The perspectives of the Speaker and the potential Hearer in this new, embedded domain, need to be distinguished from: one another (5); the perspectives represented in the Narrative Domain (4a); as well as the perspectives of Speaker and Hearer in the original Speech Act Domain (4b). This is a recursive and potentially infinite process: subjects represented in the embedded Speech Act Domain can create a new Narrative Domain (6) in which subjects and viewpoints are introduced, who can then be presented as embedded Speakers creating a new Narrative Domain, ad infinitum (Sanders et al., 2012). The embedding of Speech Act Domains and Narrative Domains within the original Narrative Domain multiplies the number of viewpoints that Speaker and Hearer in the Speech Act Domain need to track. Humans are unique in their ability to handle complex viewpoint structures, although this task increases in difficulty as the number of viewpoints increases (van Duijn et al., 2015). This task would presumably be particularly difficult for people with schizophrenia because of the perspective-­taking difficulties associated with the disorder, as outlined in the discussed phenomenological literature. In the remainder of this study we will make a first attempt to excavate occurrences of such difficulties and explore their nature by applying the Deictic Navigation Network. We will do so in a pilot study of a conversational story as told by a person diagnosed with schizophrenia.

Data Data for this analysis were obtained through a semi-structured narrative interview.3 The goal of the interview was to stimulate narrative production in the participant, thereby obtaining narrative stretches of interactive discourse. To this end, an 3  The interview was a pilot to a larger interview study that will be reported in future studies. The recruitment of the participant in this study was conducted in line with a research protocol that was waived from being assessed under the Medical Research Involving Human Subjects Act (Wet medisch-wetenschappelijk onderzoek met mensen / WMO) by the Committee on Research

178

L. van Schuppen et al.

interview was conducted according to the clean language method (Lawley and Tompkins, 2000), which is aimed at minimizing the influence of the interviewer on the verbal narrative production of the interviewee by evading interpretive and morally loaded content words like ‘problem’, ‘normal’, ‘strange’, and ‘false’, and asking for clarification without offering interpretation. The interviewer merely asked questions that stimulated the participant to continue producing narrative discourse such as ‘what happened next?’ These methods allowed the participant to talk freely in his own words about topics he chose to talk about and to structure his narrative in a way that he deemed appropriate. The interview was initiated with the question ‘Can you tell me something about your life, up until now?’ Three additional themes were explicitly touched upon: the daily/weekly routines of the interviewee, relationships with other people, and thoughts about the future. Also, the interviewer used narrative generating questions that were flexibly deployed when the interviewee did not answer in a narrative way in response to the main themes. These questions were, for example, ‘Can you tell me something about a memory that is important to you?’, ‘Can you tell me something about a time in your life when you learned a lot?’, ‘What was it like to grow up in [hometown X]?’ The audio of the interviews was recorded with the consent of the interviewee, transcribed, and anonymized. Both the audio and transcription were saved anonymously on a secure university server for analysis.

Analysis In our analysis, we show how viewpoint navigation takes place in a small and apparently simple narrative. In this particular stretch of narrative, the interviewee talks about his youth in the Netherlands, moving to a specific region in another country at the age of 18, and the subsequent return to his mother’s house in the Netherlands a couple of months later (please see the Appendix for the original transcript in Dutch and its English translation).

Involving Human Subjects (Commissie Mensgebonden Onderzoek / CMO of region ArnhemNijmegen (file nr. 2017–4007). In addition, the protocol was ethically approved of by the Ethics Assessment committee; Ethics Assessment Committee for the Humanities of Radboud University (file nr. 3625). The participant was an adult male, diagnosed with schizophrenia, who was not acutely psychotic at the time of the interview; he was not living in a clinical institute but independently, with several housemates, as arranged by a regional institute for protected living arrangements.

Navigating Narrative Subjectivity in Schizophrenia

179

Setting up the Narrative At the start of the narrative, the interviewer (henceforth: iter) invites the interviewee (henceforth: itee) to tell something about his teenage years. Nr. 1. 2. 3. 4.

Phrase I remember well We lived in [loc X] And my school was in [loc Y] So that was a bike ride of forty kilometers every day, back and forth

In phrases 1–4, the itee sets up a Narrative Domain that is spatiotemporally distanced from the time and space where the speech acts take place and where he and the iter are present at that moment. This domain is built up by the use of the past tense, separating the present here-and-now in which the itee “remember[s] well” (1) from the past in which he and other person(s) “lived in [x]” (2). Thus, the itee indicates that the story that is to follow has the form of a memory, performing a perlocutionary act that opens up a Narrative Domain of which the spatial coordinates are initially grounded in location [X]. The pronoun “we” (symbolized by ∈) includes, by presupposition, the Speaker’s deictic center ‘I’, now projected in the past. The location of living, [X], is in (3) contrasted to another relevant location: “and my school was in [Y]”. In (4) the itee here-and-now concludes (“so”) that this bi-­location resulted in the general practice of biking back and forth between [X] and [Y]: “so that was a bike ride of forty kilometers every day, back and forth” (see Sanders et al., 2012). These orientational elements describe the spatiotemporal setting of the Narrative Domain, helping the Hearer (itee) to construct a mental model of the story that is to come (Bower and Morrow, 1990) and anchor herself spatially and temporally within that story (Zwaan, 2016). This is achieved by aligning her deictic center or origo onto the Speaker (iter)‘s origo (Fludernik, 2003). The pronouns “we” and “my” indicate the presence of the narrative’s subject of consciousness (Sanders et  al., 2012). This projection is represented in Fig. 4.

Fig. 4  Projection of Speaker’s and Hearer’s deictic center from Speech Act Domain onto Narrative Domain (phrases 1–4)

180

L. van Schuppen et al.

Navigating Deictic Centers Now that the Narrative Domain has been set up, enabling the Hearer to project her deictic center into the Narrative Domain through an alignment of viewpoints, the itee expands the Narrative Domain through the introduction of events on the narrative time line. The projected deictic center in the Narrative Domain is salient, while the itee’s deictic center in the Speech Act Domain is latent. The itee continues to narrate about his school period: he bought a moped and was happy not to have to ride his bike any longer (6–10): Nr. 6. 7. 8. 9. 10. 11. 12. 13. 14.

Phrase When I was sixteen I got a …ah bought a moped with my own money myself And then I just drove eh from [loc X] to eehhh to [loc Y] So then I did not have to bike anymore So that was kind of nice then Ehhh until my eh sisters and brothers Yes are able to learn at a higher level And stuff like that Eh ehh I just did eh what they said

Phrases (11–14) include two noteworthy linguistic elements. In “until my eh sisters and brothers yes are able to learn at a higher level” (12), the itee shifts from past to present tense, continued by a past tense “I just did what they said” (14). Thus, pragmatics discourage an interpretation of this clause as representing the here-and-­ now of the Speech Act Domain and rather demand a temporally aligned representation of (12) in the narrative past. Secondly, in (14) the itee introduces a referent “they” without providing a specific deictic center to which it can be anchored. The actors ‘they’ are not introduced properly and the Hearer does not know who the itee is indicating. A resolution of this reference is very difficult on the basis of the information provided, causing a representation problem. Typically, “they” can be used in contexts where reference resolution is either inferable or unnecessary (Gundel et al., 2005). Here, however, only schematic inference could allow the listener to construct a possible ground in which both the referent and the content of what was said have to be reconstructed; for instance parents, teachers and/or maybe other responsible adults conjoining in a demand of prolonged school attendance. The itee then continues to narrate about leaving high school and starting to travel: Nr. 15. 16. 17. 18.

Phrase ehhhmmm I got out of Public Education Law then I was eighteen ehh went on a journey a journey around the world

Navigating Narrative Subjectivity in Schizophrenia Nr. 19. 20.

181

Phrase but any way I wounded up in [country Z] and ehhh there I worked

In (15–20), a shift from one spatiotemporal setting [loc X] to another [country Z] is indicated. Consequentially, there are now three spatiotemporal settings within the Narrative Domain, which encompasses the entire narrative stretch of the interviewee’s youth years which will end with his return home in (39); this is represented in Fig. 5. Having set up this Narrative Domain with its different locations, the itee can now meaningfully refer to specific locations as “there”, navigating his projected deictic center between them, for instance, in (20): ‘and ehhh there [country Z] I worked’.

Embedded Speech Act Domains The use of present tense presupposes the deictic center being present in the Speech Act Domain, which is by definition a here-and-now. Thus, a shift to present tense indicating events within the narrative requires the Hearer to build a representation of

Fig. 5  Projection of Speaker’s deictic center from Speech Act Domain onto Narrative Domain during four temporal stages at different spatiotemporal settings (1–58)

182

L. van Schuppen et al.

a Speech Act Domain that is embedded in the Narrative Domain, for example by means of a direct quote of what was said or thought at the time of the narrative now. For instance, in (35–37) the itee, having run out of money, is considering his next steps: Nr. 35. 36. 37. 38. 39.

Phrase Then I thought like Well in that case I must go home As much, − as much as I hate it So then I amI really had to adjust back home again

Built up by “thought”, a verb of cognition (van Krieken et al., 2017) and the use of present tense, phrases (36–37) are to be represented from an embedded Speech Act Domain in which the Speaker conveys thoughts that he had at the particular time and place (country Z) within the narrative. Pragmatic reasoning from the situation model that was built up from the previous utterances, forces the listener to infer, however, that the embedded Speech Act Domain does not mean that ‘home’ refers to the here-and-now location of this speech act; rather, it refers to [location X] where the narrative started. This is represented in Fig. 6. From the viewpoint of the narrative subject in the Narrative Domain, we follow in the expression of the itee’s thoughts. In (38) “So then I am-”, the itee is looking to formulate the next action in the chain of events that takes place, starting off with present tense but breaking off the clause and proceeding with past tense in (39) “I really had to adjust back home,” an evaluation of events on the narrative timeline. The abrupt navigation from the embedded Speech Act Domain to the Narrative Domain in combination with the semantics of “home” require the listener to pragmatically infer that indeed, at this point in the narrative, the itee has returned to the starting location [X]. The itee does not seem to explicitly mark this transfer, but it is probable, considering his earlier presented thoughts about having to go back home. Notice that it is not necessary for the itee to introduce us to the location of “home”, since he knows that he provided the iter with that information earlier; the information that home = location [X] is presumably shared knowledge at this point, and the deictic center shifts back from Country [Z] to location [X], all within the Narrative Domain.

Navigating Narrative Subjectivity in Schizophrenia

183

Fig. 6  Projection of Speaker’s deictic center on embedded Speech Act Domain (36–37)

Spatial Navigation The subsequent phrases, in which the Speaker explains his statement about having to “adjust back home” (39), require extensive spatial navigation of the deictic center. The itee wishes to illustrate his “adjusting back home” by presenting an anecdote. This is preceded by an interjection with present tense, quickly touching base with the iter in the Speech Act Domain here-and-now in clause (40): “I do remember”. Nr. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49.

Phrase I do remember I would visit my mother And she had a front- front- front yard and a back yard And ehh Then I sat in the front yard For I had a barrack there Where there’s steps Where you can sit down And ehh yes Then – Then - then you’d grab something to eat or something

184 Nr. 50. 51. 52. 53. 54. 55. 56. 57. 58.

L. van Schuppen et al. Phrase And then you would sit in the sun ehh Without getting a lot of attention But if you do that here in the city You will get attention of course And that was not desirable That I would sit there She said. Hah So, that is also understandable

After this introduction, he navigates from a generic event (41) “I would visit my mother”(Dutch literally: “[dan] then I went”, [dan] typically indicating something that would occur regularly) to a specific event in (43) “then I sat in the front yard” (Dutch literally: “[toen] then I sat down”, [toen] typically indicating a specific singular moment in time). Construing a representation of the information in these clauses is already effortful as a result of the transfer from generic-in-narrative-time to specific-in-narrative-time; yet, such construal will be even more effortful in the next few clauses. In (44), the itee states in a generic mode: “for I had a barrack there”. For (Dutch literally: [want]) indicates a conclusion in which the Speaker here-and-now is involved, again touching base with the iter (see Sanders et  al., 2012). Subsequently, it is not obvious how to navigate the deictic center guided by there; the preferred interpretation being to have it refer to the yard at “home”, but pragmatic reasoning from the bi-located situation model that has been built so far will lead the listener to interpret this instance of “there” as referring to Country [Z], back in time from the present point in the narrative. This deictic center’s navigation between the two locations is not marked linguistically, suggesting that the itee had a barrack in the front yard of his mother’s house, thereby derailing the comparison and confusing the iter. This might indicate that the Speaker, at least on this occasion, failed to linguistically distinguish two different viewpoints in his story, making it difficult for the Hearer to accurately interpret the deictic expression “over there”, possibly causing her to lose track of the storyline. In fact, it is not until the contrasting spatial clue “here” in (52) that the listener can confidently infer that the barrack was in fact to be represented in country [Z]. This is depicted in Fig. 7. Clauses (45–52) provide information about the habits surrounding barracks in Country [Z] by use of a general “you”, indicating that that the Speaker is explaining a common practice that was not limited to one particular moment in time, or to his own actions or that of another particular person. Figure 7 depicts the information in clause 52 separately in a Conditional Domain that is set up by “if” (Dancygier and Sweetser, 2005), representing that the validity of what happens to general ‘you’ in 53 is limited to the spatiotemporal conditions (“here”) that are set up by ‘if’. In doing so, this use of ‘you’ is more viewpointed than theoretical knowledge and an indicator of cultural knowledge about a generic practice or way of life, instead. The itee uses “here” in 52 to generically contrast this country here-and-now, specifically in the city, with country [Z]: “but if you do that here, in the city” because (53) “you will get attention, of course”.

Navigating Narrative Subjectivity in Schizophrenia

185

Fig. 7  Navigation of deictic center between Speech Act Domain, conditional domain, and Narrative Domain (45–52)

Introducing Other Speakers in the Narrative In clause 54–55 “and that was not desirable, that I would sit there, she said” again require to scrutinize on deictic navigation. Itee and iter are back here/home and there in (55) could be presumed to refer to a distant location back at the barrack in country [Z]. However, pragmatic inferencing requires to navigate back from this suggested location to “home”, in mother’s garden. This is depicted in Fig. 8. During processing, it is not entirely clear whose perspective is being presented in (54–55): potential watchers noticing the ‘I’ sitting in the garden (cultural values)? The ‘I’ in the Narrative Domain? The ambiguity is resolved by (56) “she said”, which makes clear that this would have been an utterance by the mother, embedded in the Narrative Domain. This postponed reporting verbal construction (van Duijn and Verhagen, 2019) marks the preceding clauses as indirect speech. Specifically, it characterizes the speech report as an instance of ‘distancing’ indirect speech, i.e., a form of represented speech characterized by an embedded Speaker, in this case ‘mother’, who is not fully acknowledged as such by the narrator, in this case, the itee (Vandelanotte, 2004). The viewpoint of the mother is presented, but it is not attributed to her alone, but rather shared with a more generic viewpoint, both in the narrative past as well as the speech act present (van Duijn and Verhagen, 2019).

186

L. van Schuppen et al.

Fig. 8  Navigation of deictic center between Narrative Domain and Indirect Speech Domain (54–55)

Conclusion and Discussion The aim of the analysis we presented was to explore viewpoint structures in the oral narrative discourse of a person diagnosed with schizophrenia, demonstrating a methodology that might help to bridge the gap between phenomenological theory and empirical research. The results of the analysis lead to several conclusions. The first conclusion is that even relatively short narratives show complex viewpoint structures. Telling a story was demonstrated to be a process in which the Speaker introduces the viewpoints of a range of persons whose origo’s or deictic centers are projected onto a domain distinct from the here-and-now of the speech act, including the Speaker’s own viewpoint from the past as well as the viewpoints of others. By introducing and representing these viewpoints in a Narrative Domain, the Speaker enables the Hearer to build a model of the narrative subjects and events and to project her own viewpoint onto – and into – the story world. The recursivity of viewpoint embedding was demonstrated in cases where the Speaker in the here-and-­now represented persons in the narrative as speaking and thinking subjects. In these instances, a Narrative Speech Act Domain was opened up within the Narrative Domain, multiplying the number of projected deictic centers and hence increasing the complexity of navigating across the evolving network of viewpoints. A second conclusion is that shifts between the various domains occur frequently, requiring Speaker and Hearer to keep track of multiple spatiotemporal viewpoints that are alternately salient and latent. Shifts between the Narrative Domain and the Speech Act Domain were typically marked by shifts from past tense to present tense

Navigating Narrative Subjectivity in Schizophrenia

187

and vice versa. Guided by these linguistic signals, the discourse participants temporarily move out of the narrative into the here-and-now and back. Such shifts are typical for both written and conversational narratives and may fulfill various pragmatic functions, ranging from providing further clarification about the narrative to emphasizing the truthfulness of the narrative (e.g., Dolores Porto and Romano, 2017; Sanders and van Krieken, 2019). The shifts identified in the present analysis resemble the shifts identified in American Sign Language narratives, in which narrators frequently return to the here-and-now of the conversational setting to check in with the addressee or to evaluate the past narrative events from a present viewpoint (Janzen, 2019). The current study also signals an additional function of shifts from the Narrative Domain to the Speech Act Domain, i.e., to announce the start of a new, specific narrative episode. Taken together, these findings illustrate the complexity of navigating viewpoint structures in conversational narrative, both in linguistic and cognitive respect. This study also shows that the Deictic Navigation Network (van Schuppen et al., 2019) is a useful model to analyze the viewpoint structure of conversational stories. The value of the model stretches further, though: a DNN-analysis not only reveals how multiple viewpoints are constructed, embedded, connected and navigated in the unfolding of a narrative, but also has the potential to identify potential issues in the Speaker’s cognitive management of these viewpoints. The theoretical assumptions underlying the Deictic Navigation Network hold that the ability to build complex viewpoint structures in language corresponds to several cognitive requirements, including the requirement to consider oneself and other persons as having a distinctive viewpoint and the requirement to distinguish between these various viewpoints (Sweetser, 2008). Results of our analysis signal several potential deficiencies in these requirements. For example, in embedding the viewpoints of others through speech reports, the Speaker seemed to not fully acknowledge the viewpoints of the embedded Speakers. This was reflected in an unresolved pronominal reference (“they”) and the use of a distancing indirect speech report emphasizing the Speaker’s here-and-now viewpoint rather than the viewpoint of the embedded Speaker. Both cases may signal that the Speaker, at least at these instances, does not linguistically acknowledge the Hearer’s separate and different perspective. This could imply a difficulty to take into account relevant others as viewpointed human beings (Sweetser, 2008; van Schuppen et al., 2019). A second potential confusion is the Speaker’s apparent  – albeit occasional  – inability to separate between viewpoints anchored in different domains and, in general, to maintain viewpoint stability. This was reflected in tense shifts seemingly indicating transfers between domains whereas it could be pragmatically inferred that there was no transfer at all. Similarly, inconsistent use of spatial deictic adverbs (“here”, “there”) signaled a problem in separating between viewpoints anchored in different domains and maintaining the narration from the currently profiled viewpoint. A promising avenue for future research is to study how the ways in which the navigation of different viewpoints may fail can be related to specific intersubjective disruptions such as described in the phenomenological literature. The different

188

L. van Schuppen et al.

intersubjective experiences that are described in the work of Pienkos (2015) and Pienkos and Sass (2016), for example, would be expected to drastically influence the ways in which someone with those experiences linguistically manages viewpoint. (Quantitative) empirical studies could establish whether a person who does not acknowledge the ‘mindedness’ of others and experiences perceptions of devitalization, has indeed linguistically detectable difficulties in acknowledging and demarcating the deictic centers of other people, as well as her own.4 Likewise, this could be studied for people with a pathological sense of empathy and openness. In addition, it could be analyzed whether abnormalities of common sense and conventionality are revealed in a different handling and tracking of the Hearer’s knowledge and in miscommunications caused by a lack of common ground. It is expected that such problems would especially be apparent with regards to the (in)adequate use of deixis and reference; for in optimal conditions, these phenomena require the Speaker to track the Hearer’s knowledge base, to actively depend and build on the knowledge that is currently in the common ground, as well as to navigate the spatiotemporal embedding of all the deictic centers of ‘others’ and that of one’s ‘self’, in the story. Possible future research might take a more quantitative approach in comparing the stories of language users with a schizophrenia diagnosis to life-narratives of a neurotypical population5 to establish systematic differences in perspective taking between groups. To conclude, use of the Deictic Navigation Network is helpful to pinpoint individual perspective taking problems and experiences, and to specify the ways in which people with a schizophrenia diagnosis might struggle with particular viewpoint presentation difficulties. A person who experiences the experiential disturbances as described in the phenomenological literature, would not be expected to be able to meet the cognitive requirements for the navigation of complex viewpoint structures as pointed out by Sweetser (2008), that is, to (i) recognize self’s and other’s viewpoints and (ii) project and blend them; to make constructed viewpoint structures relevant for (iii) social relations and selfhood as well as for (iv) global and local knowledge and finally (v) to meta-navigate this viewpoint system. The findings of the present study show how a cognitive linguistic analysis of viewpoint in schizophrenic narrative has the potential to illuminate the nature of perspectivization issues in patients and, in a broader sense, to strengthen the connection between theoretical philosophical and empirical linguistic research on schizophrenic (inter) subjectivity and perspective-taking. An advantage of such an analysis is that it takes into account the heterogeneity of manifestations of schizophrenia by thoroughly scrutinizing the language use of individual subjects (Joyce and Roiser, 2007). However, identifying general characteristics of viewpoint management in schizophrenic narrative requires large-scaled analyses. The Deictic Navigation Network 4  The exclusion of other perspectives would logically and epistemically result in the merger of one’s own perspective and the external world, possibly explaining self-demarcation problems in self-disorders. Sass indicates this attitude with the term ‘quasi-solipsism’ (Sass, 1994). 5  A possibility might be to use the ‘Corpus Gesproken Nederlands’, a large body of about 900 hours of spoken Dutch and Flemish, as a benchmark.

Navigating Narrative Subjectivity in Schizophrenia

189

offers a model to conduct such analyses and to quantify, for example, the frequency of successful versus unsuccessful shifts between domains. Future research in this direction might further advance our understanding of how self, other and world are experienced and moderated by people diagnosed with schizophrenia and, thus, help us grasp the disorder’s phenomenology. Funding  This work was partially supported by the Netherlands Organization for Scientific Research, NOW, awarded to K. van Krieken [grant number 275–89-038].

 ppendix: Original Transcript in Dutch and its A English Translation

Phrase 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Dutch Ik weet nog goed Wij woonden in [loc x] En mijn school was in [loc y] Dus dat was elke dag veertig kilometer fietsen heen en terug Dus dat vond ik niet echt Leuk Toen ik zestien was kreeg ik een …ah zelf een brommer van mijn geld gekocht En toen reed ik gewoon eh van [loc x] naar eehhh naar [loc y] Dus dan hoefde ik niet meer te fietsen Dus dat was dan wel fijn Ehhh totdat mijn eh zussen en broers Ja hogerop kunnen leren En dat soort dingen Eh ehh heb ik toch maar gewoon eh gedaan wat ze zeiden Ehhhmm was ik van de leerplicht af Toen was ik achttien Ehh ben een reis gaan maken Een wereldreis Maar in elk geval kwam ik in [land x] terecht En ehhh daar heb ik gewerkt Ben ik acht maanden geweest Ehhh kreeg ik vrienden Ik had er een vriendin Maar dat ging toen uit Dus dus maar eh…

English I remember well We lived in [loc x] And my school was in [loc y] So that was a bike ride of forty kilometers every day back and forth So that I did not really like When I was sixteen I got a …ah bought a moped with my own money myself And then I just drove eh from [loc x] to eehhh to [loc y] So then I did not have to bike anymore So that was kind of nice then Ehhh until my eh sisters and brothers Yes are able to learn at a higher level And stuff like that Eh ehh I just did eh what they said Ehhhmmm I got out of public education law Then I was eighteen Ehh went on a journey a journey around the world But any way I wounded up in [country Z] And ehhh there I worked I was for eight months Ehhh [I] made friends I had a girlfriend there But then that ended So so but eh…

190

L. van Schuppen et al.

Phrase Dutch 26 het was trouwens een Franse Franse vrouw 27 Die ik eh met wie ik een leuke tijd heb gehad 28 Dus dat is ook wel vreemd dat ik dat tegenkwam 29 Dat ik Frans wel op school had 30 Maar ehh dat niet heb afgemaakt 31 Dus dus gelukkig kon ze ook Engels 32

43 44 45 46 47 48 49

Dus dan konden we een beetje Frans-­ Engels met elkaar praten Maar het weer en de planten enzo Zo mooi - zo mooi ## Toen dacht ik van Naja dan moet ik maar naar huis Hoe - hoe vervelend het ook is Dus dan ben ik Moest ik thuis wel heel erg wennen Ik weet nog wel Dan was ik bij mijn moeder op bezoek En die had een voor- voor- voortuin en een achtertuin En ehhh Toen ging ik in de voortuin zitten Want ik had daar een barak Waar gewoon zo’n ehm trap is Waar je kunt zitten En ehhh ja Dan – Dan - dan ging je wat eten ofzo

50 51 52 53 54 55 56 57 58

En dan zat je in de in de zon ehhh Zonder dat je veel bekijks had Maar als je dat hier in de stad doet Heb je natuurlijk wel bekijks En dat was niet de bedoeling Dat ik daar ging zitten Zei ze Hah Dus dat is ook wel te begrijpen

33 34 ## 35 36 37 38 39 40 41 42

English It was a French French woman by the way [who] I eh with whom I had a great time So that is kind of strange that I encountered that That I did learn French at high school But ehh didn’t finish it So so fortunately she could also [speak] English So then we could speak a bit of French-­ English with each other But the weather and the plants and stuff So beautiful - so beautiful ## Then I thought like Well [in that case] I must go home As much - as much as I hate it So then I am I really had to adjust back home again I do remember I would visit my mother And she had a front- front- front yard and a back yard And ehh Then I sat in the front yard For I had a barrack there Where there’s steps Where you can sit down And ehh yes Then – Then – Then you’d grab something to eat or something And then you would sit in the sun ehh Without getting a lot of attention But if you do that here in the city You will get attention of course And that was not desirable That I would sit there She said Hah So that is also understandable

Navigating Narrative Subjectivity in Schizophrenia

191

References Allé, M. C., Potheegadoo, J., Köber, C., Schneider, P., Coutelle, R., Habermas, T., et al. (2015). Impaired coherence of life narratives of patients with schizophrenia. Scientific Reports, 5, 12934. https://doi.org/10.1038/srep12934 Bambini, V., Arcara, G., Bechi, M., Buonocore, M., Cavallaro, R., & Bosia, M. (2016). The communicative impairment as a Core feature of schizophrenia: Frequency of pragmatic deficit, cognitive substrates, and relation with quality of life. Comprehensive Psychiatry, 71, 106–120. https://doi.org/10.1016/j.comppsych.2016.08.012 Barch, D. M., & Berenbaum, H. (1996). Language production and thought disorder in schizophrenia. Journal of Abnormal Psychology, 105(1), 81–88. https://doi.org/10.1037/0021-843X.105.1.81 Barrera, A., Mckenna, P. J., & Berrios, G. E. (2005). Formal thought disorder in schizophrenia: An executive or a semantic deficit? Psychological Medicine, 35(1), 121–132. https://doi. org/10.1017/s003329170400279x Bedi, G., Carrillo, F., Cecchi, G. A., Slezak, D. F., Sigman, M., Mota, N. B., et al. (2015). Automated analysis of free speech predicts psychosis onset in high-risk youths. NPJ Schizophrenia, 1, 15030. https://doi.org/10.1038/npjschz.2015.30 Blankenburg, W. (1971). Der Verlust der natürlichen selbstverständlichkeit: ein Beitrag zur Psychopathologie symptomarmer Schizophrenien. Stuttgart, Germany: Enke. Bower, G., & Morrow, D. (May 1990). Mental models in narrative comprehension. Science, 247(4938), 44–48. https://doi.org/10.1126/science.2403694 Brune, M. (2005). ‘Theory of mind’ in schizophrenia: A review of the literature. Schizophrenia Bulletin, 31(1), 21–42. https://doi.org/10.1093/schbul/sbi002 Buck, B., Minor, K. S., & Lysaker, P. H. (2015). Differential lexical correlates of social cognition and metacognition in schizophrenia; a study of spontaneously-generated life narratives. Comprehensive Psychiatry, 58, 138–145. https://doi.org/10.1016/j.comppsych.2014.12.015 Buck, B., & Penn, D.  L. (2015). Lexical characteristics of emotional narratives in schizophrenia. The Journal of Nervous and Mental Disease, 203(9), 702–708. https://doi.org/10.1097/ nmd.0000000000000354 Bühler, K. (1934/1982). The deictic field of language and deictic words. In R.  J. Jarvella & W.  Klein (Eds.), Speech, place, and action: Studies in deixis and related topics (pp.  9–30). New York, NY: Wiley. Cardella, V. (2018). Language and schizophrenia: Perspectives from psychology and philosophy. London, UK: Routledge. Chaika, E., & Lambe, R. A. (1989). Cohesion in schizophrenic narratives, revisited. Journal of Communication Disorders, 22(6), 407–421. https://doi.org/10.1016/0021-9924(89)90034-8 Champagne-Lavau, M., & Stip, E. (2010). Pragmatic and executive dysfunction in schizophrenia. Journal of Neurolinguistics, 23(3), 285–296. https://doi.org/10.1016/j.jneuroling.2009.08.009 Corcoran, R., Cahill, C., & Frith, C. (1997). The appreciation of visual jokes in people with schizophrenia: A study of ‘Mentalizing’ ability. Schizophrenia Research, 24(3), 319–327. https://doi. org/10.1016/s0920-9964(96)00117-x Covington, M.  A., He, C., Brown, C., Naçi, L., McClain, J.  T., Fjordbak, B.  S., et  al. (2005). Schizophrenia and the structure of language: The linguist's view. Schizophrenia Research, 77(1), 85–98. https://doi.org/10.1016/j.schres.2005.01.016 Cummings, L. (2014). Pragmatic disorders and theory of mind. In L.  Cummings (Ed.), Cambridge handbook of communication disorders (pp. 559–577). Cambridge, UK: Cambridge University Press. Dancygier, B., & Sweetser, E. (2005). Mental spaces in grammar: Conditional constructions. Cambridge, UK: Cambridge University Press. DeLisi, L. E. (2001). Speech disorder in schizophrenia: Review of the literature and exploration of its relation to the uniquely human capacity for language. Schizophrenia Bulletin, 27(3), 481–496. https://doi.org/10.1093/oxfordjournals.schbul.a006889

192

L. van Schuppen et al.

Deutsch-Link, S. (2016). Language in schizophrenia: What we can learn from quantitative text analysis  (Doctoral dissertation,  Yale University School of Medicine, New Haven, CT). Retrieved from https://elischolar.library.yale.edu/ymtdl/2047/ Elvevåg, B., Foltz, P. W., Weinberger, D. R., & Goldberg, T. E. (2007). Quantifying incoherence in speech: An automated methodology and novel application to schizophrenia. Schizophrenia Research, 93(1–3), 304–316. https://doi.org/10.1016/j.schres.2007.03.001 Fineberg, S. K., Deutsch-Link, S., Ichinose, M., McGuinness, T., Bessette, A. J., Chung, C. K., & Corlett, P. R. (2015). Word use in first-person accounts of schizophrenia. The British Journal of Psychiatry, 206(1), 32–38. https://doi.org/10.1192/bjp.bp.113.140046 Fludernik, M. (2003). Chronology, time, tense and Experientiality in narrative. Language and Literature: International Journal of Stylistics, 12(2), 117–134. https://doi. org/10.1177/0963947003012002295 Fuchs, T. (2015). The intersubjectivity of delusions. World Psychiatry, 14(2), 178. https://doi. org/10.1002/wps.20209 Fuchs, T., & Röhricht, F. (2017). Schizophrenia and intersubjectivity: An embodied and enactive approach to psychopathology and psychotherapy. Philosophy, Psychiatry, & Psychology, 24(2), 127–142. https://doi.org/10.1353/ppp.2017.0018 Gallagher, S., & Zahavi, D. (2008). The phenomenological mind. London: Routledge. Gundel, J. K., Hedberg, N., & Zacharski, R. (2005). Pronouns without NP antecedents: How do we know when a pronoun is referential. In A. Branco, T. McEnery, & and R. Mitkov (Eds.), Anaphora Processing: Linguistic, Cognitive And Computational Modelling (pp. 351–364). Amsterdam, Netherlands/Philadelphia, PA: John Benjamins. Haan, D., & Sanneke, E. (2010). The minimal self is a social self. Comment on Zahavi’s minimal self and narrative self: A distinction in need of refinement. In T.  Fuchs, H.  Sattel, & P. Henningsen (Eds.), The embodied self. Dimensions, coherence, and disorders (pp. 12–17). Stuttgart, Germany: Schattauer Verlag. Haas, M.  H., Chance, S.  A., Cram, D.  F., Crow, T.  J., Luc, A., & Hage, S. (2014). Evidence of pragmatic impairments in speech and proverb interpretation in schizophrenia. Journal of Psycholinguistic Research, 44(4), 469–483. https://doi.org/10.1007/s10936-014-9298-2 Harvey, P. D. (1983). Speech competence in manic and schizophrenic psychoses: The association between clinically rated thought disorder and cohesion and reference performance. Journal of Abnormal Psychology, 92(3), 368–377. https://doi.org/10.1037/0021-843X.92.3.368 Hinzen, W. (2017). Reference across pathologies: A new linguistic lens on disorders of thought. Theoretical Linguistics, 43(3–4), 169–232. https://doi.org/10.1515/tl-2017-0013 Hinzen, W., & Rosselló, J. (2015). The linguistics of schizophrenia: Thought disturbance as language pathology across positive symptoms. Frontiers in Psychology, 6, 971. https://doi. org/10.3389/fpsyg.2015.00971 Holm, T., Thomsen, D. K., & Bliksted, V. (2016). Life story chapters and narrative self-­continuity in patients with schizophrenia. Consciousness and Cognition, 45, 60–74. https://doi. org/10.1016/j.concog.2016.08.009 Hong, K., Nenkova, A., March, M. E., Parker, A. P., Verma, R., & Kohler, C. G. (2015). Lexical use in emotional autobiographical narratives of persons with schizophrenia and healthy controls. Psychiatry Research, 225(1–2), 40–49. https://doi.org/10.1016/j.psychres.2014.10.002 Janzen, T. (2019). Shared spaces, shared mind: Connecting past and present viewpoints in American sign language narratives. Cognitive Linguistics, 30(2), 253–279. https://doi. org/10.1515/cog-2018-0045 Joyce, E. M., & Roiser, J. P. (2007). Cognitive heterogeneity in schizophrenia. Current Opinion in Psychiatry, 20(3), 268–272. https://doi.org/10.1097/YCO.0b013e3280ba4975 Kiang, M., Light, G.  A., Prugh, J., Coulson, S., Braff, D.  L., & Kutas, M. (2007). Cognitive, neurophysiological, and functional correlates of proverb interpretation abnormalities in schizophrenia. Journal of the International Neuropsychological Society, 13(4), 653–663. https://doi. org/10.1017/S1355617707070816 Kraepelin, E. (1919). Dementia praecox and Paraphrenia. Edinburgh, UK: E. & S. Livingstone Ltd. Kuperberg, G. R. (2010). Language in schizophrenia part 1: An introduction. Language and linguistics compass, 4(8), 576–589. https://doi.org/10.1111/j.1749-818X.2010.00216.x

Navigating Narrative Subjectivity in Schizophrenia

193

Kyselo, M. (2016). The enactive approach and disorders of the self-the case of schizophrenia. Phenomenology and the Cognitive Sciences, 15(4), 591–616. https://doi.org/10.1007/ s11097-015-9441-z Langdon, R., Davies, M., & Coltheart, M. (2002). Understanding minds and understanding communicated meanings in schizophrenia. Mind & Language, 17(1–2), 68–104. https://doi. org/10.1111/1468-0017.00189 Langdon, R., & Ward, P. (2009). Taking the perspective of the other contributes to awareness of illness in schizophrenia. Schizophrenia Bulletin, 35(5), 1003–1011. https://doi.org/10.1093/ schbul/sbn039 Lawley, J., & Tompkins, P. (2000). Metaphors in mind: Transformation through symbolic modelling. London, UK: The Developing Company Press. Linscott, R.  J. (2005). Thought disorder, pragmatic language impairment, and generalized cognitive decline in schizophrenia. Schizophrenia Research, 75(2–3), 225–232. https://doi. org/10.1016/j.schres.2004.10.007 Lorenz, M. (1961). Problems posed by schizophrenic language. Archives of General Psychiatry, 4, 603–610. https://doi.org/10.1001/archpsyc.1961.01710120073008 Lysaker, P. H., Clements, C. A., Plascak-Hallberg, C. D., Knipscheer, S. J., & Wright, D. E. (2002). Insight and personal narratives of illness in schizophrenia. Psychiatry, 65(3), 197–206. https:// doi.org/10.1521/psyc.65.3.197.20174 Lysaker, P. H., & Hamm, J. A. (2015). Phenomenological models of delusions: Concerns regarding the neglect of the role of emotional pain and intersubjectivity. World Psychiatry, 14(2), 175–176. https://doi.org/10.1002/wps.20207 Mishara, A.  L., Lysaker, P.  H., & Schwartz, M.  A. (2014). Self-disturbances in schizophrenia: History, phenomenology, and relevant findings from research on metacognition. Schizophrenia Bulletin, 40(1), 5–12. https://doi.org/10.1093/schbul/sbt169 Moro, A., Bambini, V., Bosia, M., Anselmetti, S., Riccaboni, R., Cappa, S.  F., et  al. (2015). Detecting syntactic and semantic anomalies in schizophrenia. Neuropsychologia, 79, 147–157. https://doi.org/10.1016/j.neuropsychologia.2015.10.030 Mossaheb, N., Aschauer, H.  N., Stoettner, S., Schmoeger, M., Pils, N., Raab, M., et  al. (2014). Comprehension of metaphors in patients with schizophrenia-spectrum disorders. Comprehensive Psychiatry, 55(4), 928–937. https://doi.org/10.1016/j.comppsych.2013.12.021 Olsen, M. R., Schredl, M., & Carlsson, I. (2013). Sharing dreams: Frequency, motivations, and relationship intimacy. Dreaming, 23(4), 245–255. https://doi.org/10.1037/a0033392 Parnas, J., & Henriksen, M. G. (2013). Subjectivity and schizophrenia: Another look at incomprehensibility and treatment nonadherence. Psychopathology, 46(5), 320–329. https://doi. org/10.1159/000351837 Perlini, C., Marini, A., Garzitto, M., Isola, M., Cerruti, S., Marinelli, V., ... & Bellani, M. (2012). Linguistic production and syntactic comprehension in schizophrenia and bipolar disorder. Acta Psychiatrica Scandinavica, 126(5), 363–376. https://doi.org/10.1111/j.1600-0447.2012.01864.x Pesciarelli, F., Gamberoni, T., Ferlazzo, F., Lo Russo, L., Pedrazzi, F., Melati, E., et al. (September 2014). Is the comprehension of idiomatic sentences indeed impaired in paranoid schizophrenia? A window into semantic processing deficits. Frontiers in Human Neuroscience, 8, 799. https://doi.org/10.3389/fnhum.2014.00799 Pienkos, E. (2015). Intersubjectivity and its role in schizophrenic experience. The Humanistic Psychologist, 43(2), 194–209. https://doi.org/10.1080/08873267.2014.990459 Pienkos, E., & Sass, L. (2016). Expressions of alienation: Language and interpersonal experience in schizophrenia. Journal of Psychopathology, 22, 62–70. Porto, M.  D., & Romano, M. (2017). ‘You’Ve just got to walk away’: Mixed viewpoints in radio call-in trauma narratives. Journal of Pragmatics, 122, 35–49. https://doi.org/10.1016/j. pragma.2017.06.002 Rapp, A. M., Langohr, K., Mutschler, D. E., & Wild, B. (2014). Irony and proverb comprehension in schizophrenia: Do female patients ‘dislike’ ironic remarks? Schizophrenia Research and Treatment, 2014, 1–10. https://doi.org/10.1155/2014/841086 Ratcliffe, M. (2015). The interpersonal world of psychosis. World Psychiatry, 14(2), 176–178. https://doi.org/10.1002/wps.20208

194

L. van Schuppen et al.

Ratcliffe, M. (2017). Real hallucinations: Psychiatric illness, intentionality, and the interpersonal world. Cambridge, MA: The MIT Press. Rochester, S., & Martin, J. R. (1979). Crazy talk: A study of discourse of schizophrenic speakers. New York, NY: Plenum Press. Saavedra, J. (2010). Quantitative criteria of narrative coherence and complexity in persons with paranoid schizophrenia. The Journal of Nervous and Mental Disease, 198(5), 349–355. https:// doi.org/10.1097/NMD.0b013e3181da47f5 Saavedra, J., Cubero, M., & Crawford, P. (2009). Incomprehensibility in the narratives of individuals with a diagnosis of schizophrenia. Qualitative Health Research, 19(11), 1548–1558. https:// doi.org/10.1177/1049732309351110 Sanders, J., Sanders, T., & Sweetser, E. (2012). Responsible subjects and discourse causality. How mental spaces and perspective help identifying subjectivity in Dutch backward causal connectives. Journal of Pragmatics, 44(2), 191–213. https://doi.org/10.1016/j.pragma.2011.09.013 Sanders, J., & van Krieken, K. (2019). Traveling through narrative time: How tense and temporal deixis guide the representation of time and viewpoint in news narratives. Cognitive Linguistics, 30(2), 281–304. https://doi.org/10.1515/cog-2018-0041 Sass, L. A. (1994). The paradoxes of delusion: Wittgenstein, Schreber, and the schizophrenic mind. Ithaca, NY: Cornell University Press. Sass, L. A. (2014). Self-disturbance and schizophrenia: Structure, specificity, pathogenesis (current issues, new directions). Schizophrenia Research, 152(1), 5–11. https://doi.org/10.1016/j. schres.2013.05.017 Sass, L. A., & Byrom, G. (2015). Phenomenological and neurocognitive perspectives on delusions: A critical overview. World Psychiatry, 14(2), 164–173. https://doi.org/10.1002/wps.20205 Sass, L.  A., & Parnas, J. (2007). Explaining schizophrenia: The relevance of phenomenology. In M. C. Chung, K. W. M. (B.) Fulford, & G. Graham (Eds.), International perspectives in philosophy and psychiatry. Reconceiving schizophrenia (pp. 63–95). Oxford, UK: Oxford University Press. Sass, L. A., & Pienkos, E. (2015). Faces of intersubjectivity: A phenomenological study of interpersonal experience in melancholia, mania, and schizophrenia. Journal of Phenomenological Psychology, 46(1), 1–32. https://doi.org/10.1163/15691624-12341283 Schettino, A., Lauro, L. R., Crippa, F., Anselmetti, S., Cavallaro, R., & Papagno, C. (2010). The comprehension of idiomatic expressions in schizophrenic patients. Neuropsychologia, 48(4), 1032–1040. https://doi.org/10.1016/j.neuropsychologia.2009.11.030 Schreber, D.  P. (2000).  Memoirs of my nervous illness. New  York, NY: New  York Review of Books. (Original work published in 1903). Schredl, M. (2000). Dreams and dreaming: The effect of dreams on waking life. Sleep and Hypnosis, 2(3), 120–124. Sechehaye, M. (1951). Reality lost and regained: Autobiography of a schizophrenic girl. New York, NY: Grune & Stratton. Sweetser, E. (2008). Viewpoint and perspective in language and gesture: ACES distinguished international lecture. Cleveland, Ohio: Case Western Reserve University. 26 March 2008, Retrieved April 27, 2019 at https://www.youtube.com/watch?v=-EGxoAIQ0mM. In press. Sweetser, E. (2012). Introduction: Viewpoint and perspective in language and gesture, from the ground down. In B.  Dancygier & E.  Sweetser (Eds.), Viewpoint in language: A multimodal perspective (pp. 1–23). Cambridge, UK: Cambridge University Press. Tavano, A., Sponda, S., Fabbro, F., Perlini, C., Rambaldelli, G., Ferro, A., et al. (2008). Specific linguistic and pragmatic deficits in Italian patients with schizophrenia. Schizophrenia Research, 102(1–3), 53–62. https://doi.org/10.1016/j.schres.2008.02.008 Thoma, P., Hennecke, M., Mandok, T., Wähner, A., Brüne, M., Juckel, G., et al. (2009). Proverb comprehension impairments in schizophrenia are related to executive dysfunction. Psychiatry Research, 170(2–3), 132–139. https://doi.org/10.1016/j.psychres.2009.01.026 Vandelanotte, L. (2004). Deixis and grounding in speech and thought representation. Journal of pragmatics, 36(3), 489–520. https://doi.org/10.1016/j.pragma.2003.10.003 van Duijn, M., & Verhagen, A. (2019). Recursive embedding of viewpoints, irregularity, and the role for a flexible framework. Pragmatics, 29(2), 198–225. https://doi.org/10.1075/prag.18049

Navigating Narrative Subjectivity in Schizophrenia

195

van Duijn, M. J., Sluiter, I., & Verhagen, A. (2015). When narrative takes over: The representation of embedded Mindstates in Shakespeare’s Othello. Language and Literature: International Journal of Stylistics, 24(2), 148–166. https://doi.org/10.1177/0963947015572274 van Duppen, Z. (2016). Schizophrenia: a disorder of intersubjectivity: a phenomenological analysis (Doctoral dissertation, Heidelberg University, Heidelberg, Germany).  https://doi. org/10.11588/heidok.00022553 van Krieken, K., Hoeken, H., & Sanders, J. (2017). Evoking and measuring identification with narrative characters – A linguistic cues framework. Frontiers in Psychology, 8, 1190. https:// doi.org/10.3389/fpsyg.2017.01190 van Krieken, K., Sanders, J., & Sweetser, E. (2019). Linguistic and cognitive representation of time and viewpoint in narrative discourse. Cognitive Linguistics, 30(2), 243–251. https://doi. org/10.1515/cog-2018-0107 van Schuppen, L., Van Krieken, K., & Sanders, J. (2019). Deictic navigation network: Linguistic viewpoint disturbances in schizophrenia. Frontiers in Psychology, 10. https://doi.org/10.3389/ fpsyg.2019.01616 Verhagen, A. (2005). Constructions of intersubjectivity: Discourse, syntax, and cognition. Oxford, UK: Oxford University Press. Watson, A.  R., Defteralı, Ç., Bak, T.  H., Sorace, A., McIntosh, A.  M., Owens, D.  G. C., et  al. (2012). Use of second-person pronouns and schizophrenia. The British Journal of Psychiatry, 200(4), 342–343. https://doi.org/10.1192/bjp.bp.111.095448 Willits, J. A., Rubin, T., Jones, M. N., Minor, K. S., & Lysaker, P. H. (2018). Evidence of disturbances of deep levels of semantic cohesion within personal narratives in schizophrenia. Schizophrenia Research, 197, 365–369. https://doi.org/10.1016/j.schres.2017.11.014 Zahavi, D. (2014). Self and other: Exploring subjectivity, empathy, and shame. Oxford, UK: Oxford University Press. Zeman, S. (2017). Confronting perspectives: Modeling perspectival complexity in language and cognition. Glossa: A Journal of General Linguistics, 2(1). https://doi.org/10.5334/gjgl.213 Zimmerer, V.  C., Watson, S., Turkington, D., Nicol Ferrier, I., & Hinzen, W. (2017). Deictic and propositional meaning—New perspectives on language in schizophrenia. Frontiers in Psychiatry, 8, 17. https://doi.org/10.3389/fpsyt.2017.00017 Zwaan, R.  A. (2016). Situation models, mental simulations, and abstract concepts in discourse comprehension. Psychonomic Bulletin & Review, 23(4), 1028–1034. https://doi.org/10.3758/ s13423-015-0864-x

Pragmatic Perspective of Literary Texts for Children Alla Tsapiv

Introduction: Narrative Text – Key Issues Children’s literature is viewed as literary texts of different genres (fairy tales, short stories, novels and fantasy) addressed to children. We assume that there are two criteria to understand who a child is. There is juridical and psychological approach to the age and psychological competence of the child. Juridically, basing on the UN Convention on the Rights of the Children (1989), a person under the age of 18 is considered to be a child. Psychologically, children’s age is determined in correspondence with the school education – primary school, secondary school, high school. A person above 13 is a teenager and his/her psychological maturity, critical thinking, encyclopedic knowledge differ from a person under 13. In our research, we assume that literary texts for children are addressed, written and thus dedicated to children. The factor of addressee is principal here. Children believe in magic and realize all the magic transformations (Cinderella), fairy/ unusual chronos and topos, wizards, witches as something real and common. They do not split real and fairy worlds in fairy tales. The main text-typological characteristics of literary texts for children: • Main characters: anthropomorphic characters as a rule are of children’s age or animalistic/floristic characters with anthropomorphic features; • Plot: as a rule dynamic plot about magic adventures. It can be a search for some supernatural object or a miraculous person; • Composition: composition tends to have a chronological structure and all the constituent elements  – exposition, rising action, climax, falling action and resolution; A. Tsapiv (*) Kherson State University, Kherson, Ukraine © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 F. Macagno, A. Capone (eds.), Inquiries in Philosophical Pragmatics, Perspectives in Pragmatics, Philosophy & Psychology 28, https://doi.org/10.1007/978-3-030-56696-8_11

197

198

A. Tsapiv

• Moral aspect: distinct division of good and evil characters. In literary texts for children ambivalence is avoided; • Functions: the main function is regulatory. Such texts tend to influence a child’s consciousness and to regulate his/her behavior, ethic norms, relationship with friend and family. Literary texts for children educate, motivate, advertise, teach what is good and what is evil. Narrative text  In the research, we understand narrative text as a complex structure, which consists of such obligatory elements: narrative (story), narration (the way in which a story is represented to a reader), narrator (a text category who/ which narrates the story), narratee (an implied reader to whom the narrative is addressed and who becomes the main factor in the choice of narrative units which the narrator chooses to create the narrative) the model of narration which is viewed as an inbuilt construal of cognitive and narrative nature. Gerard Genette understands narrative as the way in which characters and events of the story are offered to the reader. The narrative level of the narrative text focuses on (a)chronology organizational principles. Narration is the surface level of the narrative text which comes down to the formulation of the story. Narration refers to the concrete and directly visible way in which a story is told, comprising word choice, sentence length and narrating agent of the narrative text (Genette, 1998: 64–67). What is narrative  We understand narrative as a complex unit which is construed of two layers. First – narrative as a story about some events. In such approach narrative refers to a literary text of any genre and conveys a story with its plot, characters, fabula. Narration refers to the way in which the narrator manipulates with all the narrative units. Narration presupposes a certain chronological (achronological) structure of composition units, the choice of stylistic devices and expressive means, it enables the narrator to furnish the text world with the units he considers the most appropriate. Narrator  In the literal sense, the term “narrator” designates the inner-textual (textually encoded) speech position from which the current narrative discourse originates and from which references to the entities, actions and events that this discourse is about are being made (Margolin, 2009). Narratee (implied child reader) becomes the main factor who influences and determines the choice of the units to furnish the text world of any literary text addressed to children. A child reader is a special reader who perceives the world in a different from adults way. A child reader needs simplicity and humor, interesting plot and original characters to get immersed into the literary text. The model of narration is viewed as linguistic and cognitive construal which is inbuilt into the narrative structure of the fairy tale. It integrates compositional plot structure, compositional meaning structure, linguistic and stylistic means of their

Pragmatic Perspective of Literary Texts for Children

199

actualization in the text. It is actualized on the grammatical, semantic and semiotic text levels. We assume narrative model as the way of construing the story-telling, it includes a number of operations, which enable to develop the plot. A case study of the research  Roald Dahl’s fairy narratives “Charlie and the Chocolate factory” and “Charlie and the Great Glass Elevator”.

Intentions and Images Author’s intention – to make a child-reader believe in himself and make him/her sure, that any cherished dream can come true. Narrator’s intention – “make child’s dreams come true in fairy narratives”. As children tend to comprehend a fairy literary text as a piece of real world (quite possible for them, children) magic transformations, fairy objects with extra powers capture their attention and appear as quite real. The narrator in “Charlie and the Chocolate factory” creates the image of Mr. Willy Wonka who can make child’s dream come true. Willy Wonka creates the most fantastic in the whole world Chacolate factory which produces absolutely unique and delicious chocolates, candies, chewing gums a child can only dream about. For example, Willy Wonka invents a chewing-gum which may have different tastes and can substitute the whole lunch. While chewing, it changes the taste from the first course to the desert: my latest, my greatest, my most fascinating invention! It’s a chewing-gum meal! It’s . . . it’s . . . it’s . . . That tiny little strip of gum lying there is a whole three-course dinner all by itself! (Dahl, 2013a: 73).

The images of good wizards or extraordinary people like Willy Wonka who can make magic things appear in fairy narratives, have certain function – they make a child reader believe in magic, believe in himself/herself. A child can realize that even his most cherished dream can come true. Something that seems absolutely impossible and unreal can be implemented into life just if you want to. If a child reader may feel confused while reading definite terms in fairy narratives, the narrator uses additional comments and explanations to make a child be aware of unknown things for him/her: Grandpa Joe’s rockets turn us to starboard, to the right. Charlie’s turn us to port, to the left

Starboard (right) and port (left) refer to special marine terminology and may be not known for a child, do not exist in his/her experience. The narrator provides a child-reader with additional comments and explains what is starboard and what is port. The narrative strategy simplification is realized by means of narrator’s explanatory comments, simple sentences and primary word order.

200

A. Tsapiv

 airy Text Worlds Furnished with Child-Oriented F Metaphorical Units As it is known, text world is the sets of scenarios and type of reality that the text is about. The text worlds of fiction and literature are cognitive and cultural constructs that are imagined by speakers or writers in text production and by listeners or readers in text comprehension. The text world of fiction and literature are rich, dynamic, “furnished” worlds: they are inhabited by concrete individuals who are endowed with specific properties and involved in specific events unfolding in specific settings (Semino, 2009). The key option – the way in which text worlds are furnished in fairy narratives. The nature of text worlds results from the fact that texts can only explicitly provide a limited amount of information about the worlds they project. A child-reader furnishes the world of a fairy narrative by combining elements from his/her knowledge and experience of the world of the fairy tale with the knowledge of the real world he/she lives in. Differences in the nature and amount of the knowledge available to readers will result in differences in their interpretation and appreciation of the text (Semino, 2009). One of the most important things is that it may happen that a child-reader requires additional knowledge and experience to construct a text word of a fairy tale in his mind. It happens if a child-reader meets in the text some objects, notions, things he cannot reconstruct from his encyclopedia. Thus, the narrator of a fairy narrative, which is addressed to a child-reader must construct fairy text world in a special way so to make it interesting, clear and intelligible for the addressee. It is claimed that expressive means, stylistic devices, images, plot and composition of such texts have a child-oriented tendency. It proves the realization of poetics of simplicity as the key narrative strategy of the author. The author chooses the narrative strategy and the model of narration according to his intention – to educate, advertise, motivate or amuse a child. In fairy narratives for children the harmony of expressive means, stylistic devices, lexical, syntactic, composition and plot text forming units have a child-oriented tendency. Child-oriented metaphors and similes have in their background play/ game, sensitive, gastronomic images which a child can see, feel and taste in his/her everyday life. Such storytelling units create poetics of simplicity in literary texts for children. Gastronomic/sensory images make a child reader immerse in the text and realize unknown for him/her objects with the help of well-known things that already exist in a child’s experience. Conceptual metaphor is viewed as key issue in the language analysis from the perspective of cognitive linguistics. Conceptual metaphor presupposes projection between two mental representations (source and target conceptual domains). The mechanisms that underlie gastronomic metaphors in fairy narratives involve relationships between source and target conceptual domains (definite characteristics/elements of the source domain are projected into characteristics of the target domain).

Pragmatic Perspective of Literary Texts for Children

201

The concept of the Great Chain of Being, suggested by Lakoff & Turner (1989), explains the mechanisms that underlie conceptual mapping of two conceptual domains. The Great Chain of Being suggests hierarchy: “The Great Chain of Being is a cultural model that concerns kinds of beings and their properties and places them on a vertical scale with \“higher\” beings and properties above \“lower\” beings and properties” (Lakoff & Turner, 1989: 166) (More than Cool Reason. A Field Guide to Poetic Metaphor). The Great chain of Being is represented as a scale: • • • •

HUMAN ANIMAL PLANT INANIMATE OBJECT

Humans are though higher-order beings than animals, animals are higher-order beings than plants and so on. Each of the levels has sublevels. i.e. – dogs are higher than insects and trees are higher than algae. • Each unit in this scale of being embodies a scale of properties – generic-level parameters, interior states, cognitive abilities (for humans these are mental, moral, aesthetic parameters), instincts (for animals), for substance – a part-whole functional structure, properties: • HUMANS: Higher-order attributes and behavior (thought, character); • ANIMAL: Instinctual attributes and behavior; • PLANTS: Biological attributes and behavior; • COMPLEX OBJECTS: Structural attributes and functional behavior; • NATURAL PHYSICAL THINGS: Natural physical attributes and natural physical behavior. Child-oriented metaphors and similes aim at furnishing a fairy-text world of a fairy tale with the units that are understandable for a child-reader. Most unknown objects, living beings, emotional states of characters, their feelings are represented by the narrator by means of child-oriented metaphors, similies that have gastronomic nature – all potentially unknown things for a child are understood by means of projecting physical properties (shape, size, temperature, structure, density) of well-known items of food to the properties of unknown or not easily comprehended objects. To illustrate child-oriented metaphors and similes and their pragmatic function in fairy narratives, we will analyze Roald Dahl’s fairy narratives “Charlie and the Chocolate factory” and “Charlie and the great glass elevator” The main characters of a well-known fairy narrative “Charlie and the Chocolate factory” are Charlie Bucket and four children-participants of a quest on a Chocolate factory. Charlie Bucket symbolizes modesty and honesty. He is sincere, kind and compassionate person. Though, four other quest-participants personify simulacrums of modern society (Bodriyar) – greed and gluttony (Augustus Gloop), parent’s permissiveness (Veruca Salt), uncontrolled TV watching (Mike Teavee), vanity (Violet).

202

A. Tsapiv

The narrator as if a “hidden adult” of the fairy tale tells the story from the point of view a didactic adult, who criticizes pseudo values of the characters and supports honesty of the main hero Charlie. The “hidden adult” who exists inside the text, describes the appearance of Augustus Gloop, the boy who adores eating and who is obviously obese, by means of gastronomic metaphors: the picture showed a nine-year-old boy who was so enormously fat he looked as though he had been blown up with a powerful pump. Great flabby folds of fat bulged out from every part of his body, and his face was like a monstrous ball of dough with two small greedy curranty eyes peering out upon the world (Dahl, 2013a: 19)

The combination of two lexical units – monstrous and dough create the image of ugly and shapeless body, as dough, it is sticky and soft but awfully looking. A child reader can imagine a fat Augustus whose great desire to eating too much has led him to such monstrous appearance. Such gastronomic simile (was like a monstrous ball of dough with two small greedy currant eyes) enables the narrator to create a desired perlocutionary effect on a child-reader  – to demonstrate awful consequences of overeating. The conceptual metaphor that underlies HUMAN’S CHARACTERISTICS ARE FOOD ITEM’S PROPERTIES (Fat face = ball of dough; appearance of a child – the view of a ball of dough) For example, the emotional state of fear is represented with gastronomic simile: “We’ll be scrambled like eggs!” said Grandma Georgina (Dahl, 2013b: 5); When Grandma Josephine feels the trouble they are in is too dangerous, she explains her emotions: “We’re in a hot enough stew already” as if it is a boiling hot stew. The President of the USA is the fairy narrative is depicted as absolutely childish person who is afraid of aliens and unknown space ships: “We’ll be mashed like potatoes” (Dahl, 2013b: 48). All the characters, either grown ups and children in the fairy text worlds of Roald Dah’s fairy narratives speak the same language and explain their feelings and emotions like children: Turbulence is realized by the fairy narrative characters as the feeling of fish in a tank: “…Theay are all floating about like fish in a tank!” . In the fairy narrative “Charlie and the great glass elevator”, all the heroes appear in a magic glass elevator that flies high in the sky. Grandma Josephine looks down on earth and the narrator comments: “Through the glass floor she saw the entire continent of North America nearly two hundred miles below and looking no bigger than a bar of chocolate”) (Dahl, 2013b: 8). The whole continent (its size) is compared with the bar of chocolate = THE SIZE OF THE NATURAL PHYSICAL THING (CONTINET) IS THE SIZE OF A GASTRONOMIC ITEM (BAR OF CHOCOLATE). An unusual look of Space Hotel, which flies in the sky, is represented to a child-­ reader via gastronomic epithet: “…its first Space Hotel, a gigantic sausage-shaped capsule…” (Dahl, 2013b: 12). A space hotel is a magic object the knowledge of which does not exist in a child-reader’s experience. Such epithet enables the reader to imagine its physical configuration and at the same time makes the hotel funny and curious. A space hotel is not something ugly and strange, it is nothing more than a big flying sausage.

Pragmatic Perspective of Literary Texts for Children

203

The alien in the fairy narrative has such appearance: “It was as tall as a big boy and wider than the fattest man. The greenish-brown skin had a shiny wettish apeearance and there were wrinkles in it. About three-quarters of the way up, in the widest part, there were two large round eyes as big as tea-cups…the entire egg-­ shaped body was itself moving…” (Dahl, 2013b: 56–57). LIVING BEING APPEARANCE (SHAPE AND SIZE) IS THE APPEARANCE OF OBJECTS (FOOD ITEM (EGG) AND DISHES (TEA-CUPS)). The distance that is cover in a fairy narrative is the distance between daily meals …travelling a million miles between lunch and supper, and then another million before breakfast the next day. How else could they travel between the planet Vermes and other stars?

A strange noise becomes familiar when compared with the sound of the frying bacon: Inside the Elevator they could actually hear it sizzling. It made a noise like bacon frying.

Conclusions The key narrative strategy – simplification is actualized in literary texts for children by means of realization poetics of simplicity as the harmony of expressive means, stylistic devices, lexical, syntactic, composition and plot text forming units which have a child-oriented tendency. The main factor, taken into account while creating literary texts for children (fairy narratives) is the factor of addressee. The addressee of fairy narratives is a child and his/her linguistic competence, life experience, emotional and psychological maturity require a special way of storytelling to achieve a necessary fiction communication. The narrator creates magic characters and characters with extra power (Willy Wonka) to make a child reader believe in magic. Child-­ oriented metaphors and similes enable to furnish a fairy text world with original images and explain to a child-reader unknown for him/her things by means of gastronomic and sensory images that exist in child-reader’s experience. The realization of the narrative strategy of simplification enables to achieve a desired perlocutionary effect and to realize author’s intention to make a child believe in magic, magic transformation, to teach a child what is good and what is evil.

References Genette, G. (1998). Figury [Figures]. (S. Zenkin, Trans). Moscow, Russian Federation: Izd-vo im. Sabashnikovyh [in Russian]. Lakoff, G., & Turner, M. (1989). More than cool reason: A field guide to poetic metaphor. Chicago, IL: The University of Chicago Press. Margolin, U. (2009). Narrator. In F. Jannidis, M. Martınez, J. Pier, & W. Schmid (Eds.), Handbook of narratology (pp. 351–365). Berlin, Germany/New York, NY: De Gruyter.

204

A. Tsapiv

Semino, E. (2009). Text worlds // Cognitive poetics: Goals, gains and gaps. Berlin, Germany/New York, NY: Mouton de Gruyter. UN Convention on the Rights of the Children. (1989). Retrieved from: http://zakon5.rada.gov.ua/ laws/show/995_021

Sources Dahl, R. (2013a). Charlie and the chocolate factory. London, UK: Puffin. Dahl, R. (2013b). Charlie and the great glass elevator. London, UK: Puffin.

When Both Utterances and Appearances are Deceptive: Deception in Multimodal Film Narrative Marta Dynel

Introduction Despite the prevalence of deception in film,1 there is very little rigorous research on this topic in the fields of philosophy and pragmatics. Paying little heed to the characteristics of deception in film discourse per se, some philosophical scholarship on deception is illustrated with verbal examples taken from fiction, presented as specimens of scripted but natural discourse (Dynel, 2016, 2018; see also Vincent Marrelli, 2004). Apart from a few mentions and/or very brief discussions of famous deceptive scenes in film studies (Bordwell, 1985; Chatman, 1978, 1990; Kozloff, 1988), and a range of stylistic content analyses of deceptive characters in fiction, mainly in literature but also in film (e.g. DePaulo, 2010; Ferenz, 2008; Sorlin, 2016), the precious few works in film studies offer more extensive analyses of the phenomenon at hand, albeit variously labelled (e.g. Anderson, 2010; Elsaesser, 2009; Klecker, 2013; Laass, 2008). The topic of deception in literary and film fiction is also tacitly related to the notion of unreliable narration examined in narrative studies (e.g. Anderson, 2010; Burgoyne, 1990; Koch, 2011; Stühring, 2011; Zipfel, 2011). All of these works will be critically addressed in the course of this paper. The prime aim of this article is to intertwine the relevant research threads and address the gap in philosophical pragmatics by giving a comprehensive theoretical

 “Film” is here used as a technical term for both feature films, and television series/serials (see Dynel, 2011a). It is also used as shorthand for fictional productions (as opposed to documentaries).

1

M. Dynel (*) University of Łódź, Łódź, Poland Vilnius Gediminas Technical University, Vilnius, Lithuania e-mail: [email protected]; https://orcid.org/0000-0003-4647-946X © The Author(s) 2021 F. Macagno, A. Capone (eds.), Inquiries in Philosophical Pragmatics, Perspectives in Pragmatics, Philosophy & Psychology 28, https://doi.org/10.1007/978-3-030-56696-8_12

205

206

M. Dynel

account of deception in multimodal film narratives, focusing on its general characteristics and workings, as well as its specific types. In order to meet this goal, the present paper brings together a few disciplines and approaches: the pragmatics of film discourse, the cognitive philosophy of film, multimodal analysis, narrative studies and – last but not least – the philosophy of deception. Thereby, it will be shown that, similar to utterances and simple non-verbal (e.g. gesture-based) messages, complex multimodal communications (dependent on various cinematic strategies) are amenable to pragma-philosophical analysis as vehicles for deception in film. This paper is organised into eight sections. Following this introduction, the section entitled “Film as a Multimodal Narrative Constructed by the Cinematic Narrator” gives an introduction to film as a multimodal narrative, focusing on the notion of the cinematic narrator, and on multimodal transcription and analysis. “Film Cognition” presents a brief overview of film cognition, taking into account the pragmatics of film construction (notably, the two communicative levels). A range of philosophical observations on viewers’ cognition of fictional worlds are summarised, with special attention being paid to the notion of make beliefs. The section “Narrative Unreliability” concerns the topical issue, which – as is argued – is intimately connected with deception in fiction. This is the topic of “Accounting for Types of Deception and Deception in Film”, which critically addresses previous discussions of select types of film deception and, prior to this, gives the gist of deception and its main types investigated in pragma-philosophical literature. In “Types of Multimodal Deception in Film”, three main types of film deception are proposed in reference to the two levels of communication on which it materialises, the characters’ level and the recipient’s level, as well as the narrating performer of the deception, the intradiegetic and/or the extradiegetic narrator. This discussion is illustrated with multimodally transcribed examples of deception extracted from the American television series House. The section “Can the Cinematic Narrator Lie or Just Deceive Otherwise?” attempts to answer the thorny question of whether or not the extradiegetic (cinematic) narrator can lie or only perform other forms of deception. The paper closes with “Conclusions and Final Comments”.

 ilm as a Multimodal Narrative Constructed by the Cinematic F Narrator Films can be conceptualised as multi-modal narratives (e.g. Bateman & Schmidt, 2011; Wildfeuer, 2014). Examined originally in literary studies, narration has been a research topic in film studies for a few decades now (e.g. Bordwell, 1985; Branigan, 1984, 2013; Chatman, 1978, 1990; Forceville, 2002; Lothe, 2000; Phillips, 2000; Wildfeuer, 2014). Film scholars frequently focus on the narrative, paying little attention to the narrator. This is because the persona of the narrator, i.e. a perceptible personal/figural narrator (cf. Köppe & Kindt, 2011), is only intermittent in films (see e.g. Stam, Burgoyne, & Flitterman-Lewis, 1992). Only sometimes

When Both Utterances and Appearances are Deceptive

207

are personal narrators present in films (a) on the (intra)diegetic (cf. Genette 1980 [1972]) level as on-screen character narrators, or (b) on the extradiegetic (cf. Genette 1980 [1972]) level as voiceover narrators, whether characters or not. The latter case coincides with what Bordwell (2008) refers to as the Voice of God narrator. There is an ongoing debate on whether film, or narrative fiction in general,2 must involve the extradiegetic/heterodiegetic narrator “as the illocutionary source or instance of emission of the narrative discourse” (Burgoyne, 1990: 4). According to one view, film intrinsically involves an impersonal multimodal narrator (cf. Burgoyne, 1990) that is “external to (not part of) any diegesis” (Prince, 1987: 29) and entirely responsible for the presentation of the diegetic world. This extradiegetic narrator in film narrative has been labelled previously: le grande imagier (Metz, 1974), the camera-narrator (Kozloff, 1988), the intrinsic narrator (Black, 1986), the fundamental narrator (Gaudreault, 1987) or the cinematic narrator (Anderson, 2010; Burgoyne, 1990; Chatman, 1990; Stam et  al., 1992). This last term is preferred here, serving as a shorthand metaphor. A view is endorsed here that if there is a narrative, there must be a type of impersonal heterodiegetic narrator involved as a matter of logical necessity (Chatman, 1990; Kozloff, 1988; Ryan, 1981). In addition, the heterogeneous impersonal narrator facilitates establishing the “hierarchy of narrative voices structuring the narrative film” (Burgoyne, 1990: 6). Bordwell (1985: 62) famously questions this kind of a narrator in film, based on the fact that the viewer is “seldom aware of being told something by an entity resembling a human being,” and states that film narration is “the organization of a set of cues for the construction of a story.” Although Bordwell (2008: 121–122) later does acknowledge that the narrator in literary works may be non-human/impersonal, which is a widely recognised fact (e.g. Ryan, 1981), he is adamant that this narrative voice is an otiose “personification of the narrative dynamics in film” as it is rarely explicit and can seldom be identified, while the relevant recognised agent is the filmmakers. Bordwell (2008) also presents the notion of a narrator in film as not appealing to any “psychological activity”. Whilst this last critical observation seems correct, little support can be given to his explanation that viewers attribute the narrative or some effect to the film itself or the filmmakers, unlike readers who may identify the speaking voice in the narrator or the novel writer. In any case, these claims are nothing but speculative, to put it mildly, and they are not compatible with what many cognitive philosophers of fiction have proposed (see “Film Cognition”). Most importantly, Bordwell’s (1985, 2008) argument concerning narrator-less film seems to be based on the assumption that a narrator is part of the viewer’s cog-

 Many (Hamburger, 1957; Banfield, 1978; Branigan, 1984; Köppe & Kindt, 2011; see Sternberg & Yacobi, 2015 and references therein) claim that the narrator may be absent in literary fiction altogether and the reader is invited to imagine that something is the case without any intermediary telling the story. However, others (e.g. Doležel, 1980; Ryan, 1981) argue this intermediary, albeit not necessarily anthropomorphic, is a must, tacitly representing the real author in a given work. On some accounts, the presence of this narrator is what makes extradiegetic deception possible, for it prevents the author from shouldering the blame for it (see “Can the Cinematic Narrator Lie or Just Deceive Otherwise?”).

2

208

M. Dynel

nitive reality, i.e. that the viewer needs to be aware of the narrator (see also Branigan, 1984). However, as Ryan convincingly puts it, “the concept of ­narrator is a logical necessity of all fictions, but it has no psychological foundation in the impersonal case. This means that there is no need for the reader of impersonal narrations to seek an answer to the question ‘who speaks?’” (1981: 519). The viewer will not typically ask this question also because their focus is strictly on the fictional world, rather than the medium or the actual author, i.e. production crew, contrary to what Bordwell (2008) contends (see “Film Cognition”). Hence, the cinematic narrator should be seen as an empirically validated construct; rather, it is a theoretical and abstract notion that facilitates philosophical discussion. Burgoyne neatly summarises this issue: “In creating the fictional world, the impersonal narrator produces a type of discourse that is read directly as the facts of the ‘real world’ of the fictional universe. The impersonal narrator’s lack of human personality allows the viewer to imagine that he or she is confronting the fictional universe directly, putting aside any reflection on the form of the narrative discourse” (1990: 7). This kind of narrator bears some resemblance to the impersonal narrator in literary fiction regarded as being, at least partly, “covert”, “backgrounded” or “non-perceptible” (see e.g. Chatman, 1978; Ryan, 1981; Toolan 2001 [1988]; Bal 1997 [1992]). Even though this narrator is usually not to be consciously recognised by the recipient, i.e. the model viewer (Dynel, 2011a, b), it does constitute the lens through which the recipient perceives select elements of the fictional world.3 Importantly, the cinematic narrator is, in fact, the product of the collective sender (see Dynel, 2011a, b), a technical term for the film production crew responsible for the creation of the multimodal narrative. Multimodality is also a heterogeneous notion that escapes easy definitions. Despite numerous attempts (e.g. Elleström, 2010; Jewitt, 2009; Kress, 2010), there is no unanimous agreement on how modes or modalities should be defined (Forceville, 2010). It is, nonetheless, commonly held that multimodality involves multiple integration of semiotic resources (Baldry & Thibault, 2006; Kress, 2010). Suffice it to say that films are multimodal products insofar as they encompass characters’ utterances and non-verbal messages (facial expressions and body language), as well as actions; written language; sounds, music and lyrics; and – last but not least  – visuals, which depend on various cinematic components (e.g. continuity editing). This multimodality encompasses both film discourse, i.e. characters’ verbal interactions (see Dynel, 2011a, b, for discussion and alternative terms), and cinematic discourse, which encompasses the various forms of “communication not in film but through it (cf. Metz, 1974)”, such as “mise-en-scène, cinematography, montage, and sound editing used in narrating cinematic stories to viewers” (Janney, 2012: 85). Both conventionalised and novel cinematic strategies (for an excellent overview, see Bordwell and Thompson 2013 [1979]) are carriers of explicit and implicit meanings relayed to viewers.

3  Presumably, the proponents of the fiction-without-narrator view would still argue that this film narrator is actually absent since we are not asked “to imagine anything about a fictional narrator” (Köppe & Kindt, 2011: 84).

When Both Utterances and Appearances are Deceptive

209

Operating across different modes, both film discourse and the whole gamut of cinematic components perform the narrative function (cf. Kozloff, 2000) and affect viewers’ interpretation of the goings-on on screen (Wildfeuer, 2014). Hence, the recipient reconstructs the story in the fictional world, based on the multimodal narrative, i.e. the story’s material representation which “consists of material signs, the discourse, which convey a certain meaning (or content)” (Ryan, 2007: 24). According to Bordwell, film narration is “the process by which the film prompts the viewer to construct the ongoing fabula [i.e. the events, as represented, cf. Genette 1980 (1972)] on the basis of syuzhet organization and stylistic patterning” (2008: 98). This necessitates viewers’ active participation, inferences and sense-making. For instance, “[w]hen information is missing, perceivers infer it or make guesses about it. When events are arranged out of temporal order, perceivers try to put those events in sequence” (Bordwell, 1985: 33f.).

Multimodal Analysis The joint use and interplay of different elements from various semiotic resources across modalities require that they be investigated jointly, as stipulated by Baldry and Thibault’s (2006) resource integration principle. The different semiotic resources “are combined and integrated to form a complex whole which cannot be reduced to or explained in terms of the mere sum of its separate parts” (Baldry & Thibault, 2006: 18). In “Types of Multimodal Deception in Film”, select scenes from the American televisions series House will be transcribed in order to exemplify film communication processes and to present what the recipient is exposed to when an act of film deception is performed. An adapted version of Baldry and Thibault’s (2006) multimodal transcription toolkit will facilitate the description of the meaning-making resources across modes, and hence promote viewer’s understanding and (make) belief making about the fictional world (see “Film Cognition”). As Forceville (2007) aptly observes, Baldry and Thibault’s (2006) toolkit can benefit from film terminology known from various sources, such as Bordwell and Thomson’s (2013 [1979]) oft-quoted and continually revised magnum opus. In line with what Forceville (2007) suggests, clusters of variables within and across modes are included in the transcriptions. For the sake of clarity, the transcription is restricted to the most salient elements that are central to the recipient’s meaning-making (Baldry & Thibault, 2006: 183) in order to best depict the multimodal construction of various types of deception operating within and across scenes. A scene, here involving an act of deception, can be divided into the key multimodal analytic units, namely phases. These are sets of “copatterned semiotic selections that are codeployed in a consistent way over a given stretch” of discourse (Baldry & Thibault, 2006: 47; Martinec, 1998). A phase is a carrier of the central meaning and often comprises more than one shot, i.e. a filmed sequence without the camera’s spatial displacement. As Baldry and Thibault observe, a multimodal text analyst needs “to specify both which selections are selected (sic) from which

210

M. Dynel

s­ emiotic modalities and how they are combined to produce a given, phase-specific meaning” (2006: 47). The transcription tables in the analyses conducted here are organised as follows: • Column 1 number and name/meaning of the phase • Column 2 the lead-in time starting from the beginning of the scene (Example 1) or the episode (Examples 2 and 3) • Column 3 select pivotal frames, i.e. stills in chronological order • Column 4 elements of cinematography (e.g. distance or camera position) and mise-­en-­scène (e.g. location or characters’ kinetic action) • Column 5 diegetic and non-diegetic soundtrack (e.g. dialogues, voice-over or music score) As also pointed out by Baldry and Thibault (2006), the transcription and analysis of the multimodal discourse must be performed from an etic perspective. The aim is to depict the (ideal) model of the interpretation process and to account for the deceptive meanings the collective sender will intend the recipient to glean about the fictional world.

Film Cognition In linguistic terms, film operates on two levels of communication, which have been captured by two spatial metaphors: “The first one understands fictional interaction as embedded in communication between authors/producers and audiences; the second one places inter-character talk on a layer on top of the primary layer between the producers and recipients of the artefact” (Messerli, 2017: 33). Essentially, many authors (e.g. Bubel, 2008; Dynel, 2011a; Kozloff, 2000; Piazza, Bednarek, & Rossi, 2011 and references therein) underscore the simple fact that the characters’ interactions (here called the characters’ level of communication), as shown on screen, depend on what the film production crew (i.e. the collective sender) wants to communicate to the audience on the recipient’s level of communication with the use of a wide range of multimodal (cinematic and discursive) strategies. The characters’ level of communication constitutes what is known in narratology and the philosophy of fiction as the fictional world,4 i.e. the diegesis (Genette 1980 [1972], see e.g. Daugherty, 2007), which is constructed for the recipient to follow.5 The collective sender can “organize the film so as to solicit a range of effects”, but “the viewer has a freedom to seize upon certain cues and not others” and use 4  This world shows many overlaps with the real world and is the point of departure for the recipient engaged in sense-making (see e.g. Doležel, 1998; Lamarque, 1996; Lewis, 1978; Margolin, 1992; Ronen, 1994; Ryan, 1980). 5  If the production crew should make mistakes (for instance, in the script, camera work or editing), the viewers may be (inadvertently) misled, but not (purposefully) deceived about the fictional world, developing some false beliefs about it.

When Both Utterances and Appearances are Deceptive

211

them “in ways that couldn’t be foreseen by the filmmakers. (…) Cinematic ­traditions, however, secure a considerable amount of convergence between what filmmakers know can affect viewers and what viewers do experience” (Bordwell 2008: 123). The preferred inferential path and understandings projected by the production crew for the model viewer (here, the recipient), may then differ from actual viewers’ actual interpretations (see Bordwell, 1985; Persson, 2003; Stafford, 2007) depending on their life experiences, values, and beliefs social background, preferences, cognitive abilities, etc. (e.g. Morley, 1980; Phillips, 2000; Wharton & Grant, 2005). This problematising aside, the focus of this paper is film deception as designed by the collective sender for the recipient, the model viewer who does follow the projected inferential paths and is indeed deceived. Deception is then one of the effects in film that rely on “convergent inference making. The filmmaker has gotten us to walk down the path she planned” (Bordwell 2008: 124). Prototypically, the recipient of a fiction film focuses on the characters’ level of communication, which means basically the fictional world depicted in the diegesis, without consciously appreciating the production crew’s work,6 and even forgetting that he/she is engaged in the act of watching a film (see Dynel, 2011a, b). Films encourage the Coleridgean willing suspension of disbelief (Bordwell, 1985; Ryan, 2001), and “sustain the illusion that the viewer is observing the action as a fly on the wall” (Kozloff, 2000: 47), being judged based on how “well they transport us into the worlds of the stories. If an adventure story is good, we imagine its world so vividly that it is like a movie running off in our heads” (Clark, 1996: 366). Hence, while it is the collective sender that communicates meanings to the recipient, the latter does not consciously see the meanings as being produced by the former unless purposefully choosing to do so (as film critics and scholars do) or unless the collective sender invites the recipient to consider their intentions (e.g. in complex metaphors) or manifestly shatters the diegesis with their cinematic methods (e.g. breaking the fourth wall). This line of reasoning corresponds to a number of postulates made in the philosophy of fiction about film watching. It has been proposed that this is an act of joint pretence (see Clark, 1996; Clark & Van Der Wege, 2001; Walton, 1990). Viewers thus forget that they are interpreting fictional interactions and events as if they are actually taking place in front of them, and they let themselves be captivated by the fictional world. This is known, among other things, as transportation (e.g. Gerrig, 1993; Green & Brock, 2000, 2002) or immersion (Ryan, 2001). As Ryan (2001: 14) observes, “immersion is the experience through which a fictional world acquires the presence of an autonomous, language-independent reality populated with live human beings.” A pending query is what mental states recipients have with regard to these fictional worlds. It may be claimed that one of the primary goals of films “is to make it compelling for us to believe in their implied [fictional] worlds for the duration of any screening

6  This does not seem to hold for non-mainstream, highbrow films rich in metaphors or symbols, or genres such as sitcom comedies, where the “code of realism” is broken.

212

M. Dynel

(and, in our private reflections, for any time thereafter)” (Murphet, 2005: 52, ­emphasis added). However, it is most often stated that recipients only imagine (Currie, 1995; Currie & Ichino, 2013) or make believe (for different conceptualisations, see Lewis, 1978; Currie, 1990; Walton, 1990; Lamarque, 1996; Ryan, 2001; Sainsbury, 2009) that something is the case in a fictional world. Therefore, rather than having genuine beliefs per se, film viewers can only have pretend beliefs (Nichols and Stich 2003). Similarly, employing formal belief revision theory, Badura and Berto (2018) conceptualise beliefs based on fiction as make believed beliefs. According to Nichols and Stich (2003), genuine and pretend beliefs are very much alike in practice but differ mainly in their functional roles. Also, cognitive processes can take input from the “pretence box and from the belief box”, forming “parallel representations much the same way” (Nichols & Stich, 2003: 131). Despite the striking similarities between the two belief categories, viewers’ understandings of the fictional world cannot be technically labelled “beliefs” per se, as many authors assert (but see Stühring, 2011 on beliefs about the fictional world)7; these are rather pretend beliefs, make believed beliefs or simply make beliefs. This last label is preferred here. Thus, recipients develop make beliefs about the fictional world, i.e. what they make believe to be true in/about the fictional world. As will be shown here, these make beliefs may be false, as intended by the collective sender. In other words, the recipient of fiction, notably a model film viewer, can develop false make beliefs about the fictional world at hand, hence being deceived, or technically make believe deceived, about its make believe/fictional truth. The truth and, by analogy, falsity, as well as make beliefs about both are relative to the fictional world considered, and the “fictional facts” therein (see Badura & Berto, 2018; Lewis, 1978; Sainsbury, 2009; Stühring, 2011). Within a fictional universe “truth is not determined relative to an extratextual universe, but relative to a fictional world” (Ronen, 1994: 40). Fictional worlds, however, involve not only fictional facts but also nonfactual elements, such as a character’s beliefs, desires or predictions (Doležel, 1980; Ronen, 1994). Thus, the objective world of the text, comprised of facts in fiction, does not exclude characters’ private, subjective domains defined as pretended worlds the characters create to deceive other ones, as well as their minds’ fabrications, such as hallucinations, fantasies and dreams (Ryan, 1991). The “truth of the text,” and “the reliability of characters purporting to speak that truth, can be measured only against the authentic facts of the fictional universe” (Burgoyne, 1990: 10). Overall, just like the real world has its objective – but frequently, non-­verifiable – truth, each fictional world has its own fictional truth (Walton, 1990; Margolin, 1992; Zipfel, 2011; cf. Lewis, 1978), based on the narrator’s make believe truthfulness. The make believe truth of the fictional world is what “readers [or viewers] can reasonably assume to be an existing state of affairs in a fictional world or, in other words, what readers [or viewers] are authorized to believe to be an existing state of affairs in a fictional world according to the [multimodal]

7  Make beliefs about fictional worlds should not be mistaken for genuine beliefs that films can also inspire in viewers (e.g. regarding moral choices or political opinions).

When Both Utterances and Appearances are Deceptive

213

text” (Zipfel, 2011: 110), based on the presumption of truthfulness for fiction (Lamarque, 1996), a ­mirror reflection of the same principle holding for real-life communication (e.g. Bok, 1978; Kupfer, 1982). It is noteworthy that the relevant aspects of the make believe truth of the fictional word are often easier to determine without a shadow of a doubt than the truth of the real world. Participating in all interactions and events shown on screen, the viewer usually enjoys the necessary knowledge about all pertinent fictional facts. This is because the collective sender needs to guarantee that the recipient has access to the relevant aspects of the fictional truth and develops “correct” make beliefs about the fictional world, unless make beliefs are meant to be “wrong”, precisely false. The latter situation means that the collective sender aims to deceive the recipient about some aspect of the fictional truth of the fictional world, as designed by the former. As will be argued here, this deception may be performed with or without the use of deceptive fictional characters, who have their fictional intentions and beliefs about their fictional world. Albeit carried out in the real world by the collective sender, deception of the recipient about the fictional world is anchored in the latter’s developing, not false beliefs, but rather false make beliefs, namely the specific beliefs about the fictional truth (e.g. Zipfel, 2011), guided by the cinematic narrator. This is the central topic of the remainder of this article, and it is connected with the notion of narrative unreliability.

Narrative Unreliability Narrative unreliability (or unreliable narration) is a literary notion (for an extensive overview, see Sternberg & Yacobi, 2015). Although it has been most frequently addressed with reference to literary works, it “can be found in a wide range of narratives across the genres, the media, and different disciplines” (Nünning, 2005: 90, cf. the contributions in Nünning, 2015). The concept of narrative unreliability is credited to Booth (1983 [1961], see Olson, 2003; Shen, 2011), who conceived it as a rather eclectic construct. An unreliable narrator is one that “does not act in accordance with the norms of the work” (1983 [1961]: 158). Also, Booth uses the labels “unreliable”, “untrustworthy” and “fallible” interchangeably (1983 [1961]: 158). Most importantly, unreliability need not always concern dishonesty and purposeful untruthfulness. There may be various causes of it (see Shen, 2011 for an overview). Thus, only some (but not all) unreliability qualifies as the narrator’s deceptive unreliability (Stühring, 2011). Olson (2003) rightly distinguishes between “fallible” and “untrustworthy” narration, which she presents as gradable notions. Whilst the former can be attributed to the narrator’s naiveté, the latter concerns a narrator’s being “dispositionally unreliable”, being driven by “ingrained behavioral traits or some current self-interest” (Olson, 2003: 102, emphasis in original). This dispositionally unreliable narrator can then be responsible for various forms of deception, including “lies” (Fludernik, 1999). Essentially, unreliable narrators can purposefully report something that is

214

M. Dynel

false in the fictional world (Badura & Berto, 2018). However, if the receivers of fiction are not aware of a narrator’s (unintended) fallibility, it is also the fallible narrator that may unwittingly invite false make beliefs in readers (and viewers alike), and hence mislead, but not (purposefully) deceive them. It is, however, also in this case that the receiver of a fictional narrative develops false make beliefs about the fictional reality (see “Can the Cinematic Narrator Lie or Just Deceive Otherwise?”). Therefore, all unreliability, as long as the receiver of fiction is not cognisant of it, may serve deceptive purposes, as envisaged by the authors (here, the collective sender). For their part, Köppe and Kindt (2011: 90) propose that a fictional work is “mimetically unreliable” if it “authorizes imagining that the narrator does not provide completely accurate information” or “authorizes imagining states of affairs that are not completely accurate.” Reporting on this definition, Zipfel states that unreliable narration concerns situations when “the narrator does not give completely accurate information, or that the narration is misleading as to what counts as fictional truth” (2011: 117). This lack of (complete) accuracy and the concept of misleading are rather vague notions. The underlying ideas are better captured by “deception”, which is the technical blanket term for various ways of inviting false (here, make) beliefs. Zipfel also addresses narrative unreliability in the context of Grice’s Cooperative Principle, especially “with its quality and quantity maxims”, stating that it is “not the author who does not organise his communicative contribution according to the Cooperative Principle but the narrator” (2011: 119). Similarly, Heyd (2011: 7) claims that a “narrative is unreliable if it violates the CP without the intention of an implicature”. These statements give rise to several misgivings given the thrust of Grice’s (1989) philosophy of communication. Essentially, the Cooperative Principle, the principle of rationality, cannot be violated: implicatures are generated based on the joint assumption that the principle holds, and deception succeeds when both the deceiver and the target of deception operate on the same presumption, the mismatch in their perception of the ongoing communication regardless (see Dynel, 2018 for detailed discussion, cf. Meibauer, 2014). Heyd (2011), however, is right in suggesting that unreliable narratives exploit maxim violations (to which the target of deception is oblivious), which are indeed conducive to various forms of deception, and it always involves the violation of Grice’s first maxim of Quality (see Dynel, 2018; Vincent Marrelli, 2004). The pending query is who performs this maxim violation, the narrator or the author; this is the query that the section “Can the Cinematic Narrator Lie or Just Deceive Otherwise?” will address. Narrative unreliability has been discussed also in film studies (e.g. Chatman, 1978; Burgoyne, 1990; Laass, 2008; Bordwell 2008; Anderson, 2010; Koch, 2011). The proposals therein are of immediate relevance to the present account of film deception which needs to be done in the light of the types of deception discerned by (language) philosophers.

When Both Utterances and Appearances are Deceptive

215

Accounting for Types of Deception and Deception in Film Based on previous definitions (e.g. Bok, 1978; Carson, 2010; Chisholm & Feehan, 1977; Mahon, 2007), deception (in real-life interactions) may be defined as an intentional communicative act of attempting (whether or not successfully)8 to cause the target, i.e. targeted receiver of the believed false message, to (continue to) have a false belief, that is to believe to be true something the deceiver believes to be false (for a detailed explanation, see Dynel, 2018). Deception is then conceptualised with regard to the communicator’s beliefs, and more specifically, intentional untruthfulness rather than (objective) falsity (e.g. Aquinas, 1972; Augustine, 1952; Bok, 1978; Dynel, 2018; Fallis, 2010; Mahon, 2015; Meibauer, 2005, 2014; Vincent Marrelli, 2004, 2006). In the present context, the make beliefs about the truth/falsity are relative to the fictional world/reality (cf. Zipfel, 2011) depicted in a given film. Deception can take a number of forms, the most important of which for the present purposes are: (a) lying, i.e. roughly, making a believed-false covertly untruthful assertion, (b) deceptively implicating defined as saying something truthful but implicating something covertly untruthful through flouting the Gricean maxims, (c) deceptively withholding information, which involves communicating nothing or something truthful and keeping covert an all/part of the believed-true meaning in order to deceive the hearer, (d) covert ambiguity, which relies on two alternative interpretations, with the salient interpretation being the covertly untruthful one, and the hidden meaning – which is not to be discovered by the target – being the truthful one, and (e) covert irrelevance, which rests on providing covertly irrelevant information as if it is relevant to the question under discussion (for a detailed overview and references, see Dynel, 2018, 2019). These (and other) types of deception can be seen through the lens of a neo-Gricean framework of communication. Essentially, all involve (covert) violation of the maxim of truthfulness (the first maxim of Quality), which may be the consequence of violating or flouting another maxim (see Dynel, 2018). Additionally, deception may be performed in different modes and through different channels: in written or spoken discourse, and verbally and non-verbally, notably via gestures, actions and artefacts (e.g. Bok, 1978; Chisholm & Feehan, 1977; Ekman, 1985; Linsky, 1963; Mahon, 2007, 2015; Meibauer, 2005; Siegler, 1966; Simpson, 1992; Smith, 2004; Vrij, 2008). Importantly, lies need to involve asserting, but asserting is not restricted to using words in speaking or writing; lies can be told non-verbally as long as a non-verbal signal carries a conventionalised or previously established assertoric meaning (see e.g. Bok, 1978; Chisholm & Feehan, 1977; Green, 2001; Mahon, 2015), which can be paraphrased verbally. This also

8  It needs to be repeated again that here the focus is on model film deception of the recipient, who is successfully deceived, as envisaged by the collective sender.

216

M. Dynel

pertains to multimodal lies in films. What is submitted here is that all deception may also be performed through multimodal means, transcending the level of non-verbal messages in everyday interpersonal interactions. Films are a case in point. There is no novelty in the observation that films may involve deception, even though this issue has been hidden under various labels and in various conceptualisations in film studies (see also section “Can the Cinematic Narrator Lie or Just Deceive Otherwise?”). For instance, Bordwell (1985: 39) states that “a film may contain cues and structures that encourage the viewer to make errors of comprehension” conducive to his/her “misunderstanding”. However, these alleged inadvertent “errors” or “misunderstandings” are epistemic effects that are purposefully exerted on the recipient by the collective sender with the help of the unreliable cinematic narrator. For her part, Laass (2008: 43–79) distinguishes between a number of forms of unreliability, and hence deception, on several narrative levels. The first level coincides with what is called here the characters’ level of communication, involving deception performed by characters, also narrating previous events. On the second level, Laass (2008) distinguishes unreliable “explicit” voice-over narration, and on the third level, she identifies “implicit” narrative communication, which concerns non-verbal communication (e.g. focalisation or perspectivisation through presenting a character’s point of view). Finally, the fourth narrative level concerns textexternal communication. The problem with this approach is that the levels, and the types of unreliability they entail, are not distinct, and the rationale and criteria for the divisions, especially the one between the second and third levels, are rather obscure and confusing (notice also the label “implicit” used in reference to “nonverbal”, which are two independent dimensions in linguistics). It is proposed here that film deception may rely strictly on elaborate multi-modal cinematographic ploys orientated towards deceiving the recipient or it may manifest itself in character’s actions and utterances aimed at deceiving other characters. This dichotomy is, presumably, what Klecker (2013: 134) means when she (rather vaguely) states that either “the film itself deceives the audience, or one or more characters do.” More specifically, deception in film may arise on the characters’ communicative level, where the characters deceive one another in the fictional world, with the recipient being simultaneously deceived too or not (when privy to the act of deception at hand); or solely on the recipient’s level, with the only target of deception being the recipient (cf. Dynel, 2013). Interestingly, in her discussion of mind-tricking narratives, Klecker (2013) claims that deception performed by characters can be divided into two types. She states that either “a character simply lies” (Klecker, 2013: 134), a claim that should be generalised as: a character deceives (by lying or performing any other type of deception); or “characters lie to themselves” (2013: 135), and more broadly “deceive themselves unknowingly” (2013: 136). However, the latter strategy, which is the consequence of characters’ mental incapacity (e.g. memory loss or s­ chizophrenia)

When Both Utterances and Appearances are Deceptive

217

rather than, at least partly, intentional acts can hardly be seen as self-­deception in a technical sense (deception is purposeful, and being wrong/deluded has little to do with it). What is crucial is the way the character’s mental incapacity is presented by the cinematic narrator; the (only) deception that does arise is targeted solely at the recipient when the cinematic narrator (in accordance with the collective sender’s plan) presents the fictional world the way the misguided or incapacitated character perceives it (see also Anderson, 2010; Ferenz, 2008; Koch, 2011), a fact about which the recipient learns only with the benefit of hindsight.9 Usually, this perspectivisation is not done through point-of-view shots (see Bordwell and Thompson 2013 [1979]), whereby the camera takes the position where the (deluded) character’s eyes would be. The character’s “subjectively distorted” perceptions are “presented without external framings or internal markers to distinguish them from shots of fictionally real events” (Koch, 2011: 73). The cinematic narrator deceives by adopting “an unreliable focalizer’s point of view as if it were accurate” (Anderson, 2010: 89). Ultimately, the deceived recipient is allowed to recognise that the previous scenes, sometimes amounting to the whole film, have presented the character’s warped view of the fictional world. The thrust of this critique of Klecker’s (2013) work is that a character’s deception of another character can only be juxtaposed with the cinematic narrator’s deception of the recipient, even though the former may also entail the deception of the recipient and does rely on the cinematic narrator too (e.g. in how the deception is shown to the recipient). Additionally, film deception may be discussed in the context of its time span. Sometimes, film deception may be specific to a single scene and even only part of it. Alternatively, it may span an entire film, echoing across many scenes until the denouement, in various forms and guises, as is the case with films such as Primal Fear, The Machinist or Fight Club. Albeit showing striking differences, these are the films that Elsaesser (2009) calls “mind game films”10 and Klecker (2013) dubs “mind-tricking narratives”. This temporal distinction corresponds to “temporary”

 Interestingly, Stühring (2011) vindicates literary narratives that involve “a narrative-persona whose report is wrong or lacking relevant information”, stating that they are not deceptive insofar as the work per se “does not justify any wrong beliefs (…) about the fictional facts.” This conceptualisation, however, misses the crucial fact that the intradiegetic narrator is actually the product of the author whose aim is to deceive the reader. 10  However, this seems to be a broad notion that goes beyond deception and covers films that involve “playing games” with characters and/or audiences. Many of the examples of films that Elsaesser (2009) provides do not really involve “disorienting or misleading spectators”, which is presented as “one overriding common feature of mind-game films” (Elsaesser, 2009: 15). Nor do they even involve characters’ deceiving other characters. Overall, the mind-game film category is rather vague. 9

218

M. Dynel

vs “sustained” (Currie, 2004; cf. Zipfel, 2011) or “local” vs “global” (Koch, 2011) narrative unreliability.

Types of Multimodal Deception in Film The present account of the types of film deception is dictated primarily by the communicative level on which it materialises, namely the characters’ level and/or the recipient’s level of communication. The former concerns deception performed by characters in the diegetic world, while the latter concerns the manner in which the diegetic world, free from deception per se, is presented to the recipient. These two ways in which deception may be deployed in films (by the collective sender) can be intersected with two types of unreliable narrators, intradiegetic and extradiegetic. The three forms or film deception thus discerned are illustrated with examples of deceptive scenes taken from House and discussed in terms of specific categories of deception examined in philosophical studies.

 xtradiegetic Deception Performed Only by the Cinematic E Narrator The first broad category of film deception targets only the recipient, being performed solely by the cinematic narrator. This deception does not present itself on the characters’ level of communication and is, therefore, external to the fictional world per se, which is why it can be conceptualised as extradiegetic deception. It is the collective sender’s various multimodal strategies that cause the recipient to develop false make beliefs about the diegetic goings-on in the fictional world. This form of multimodal film deception can be interpreted as a special case of narrative unreliability performed solely and directly by the unreliable cinematic narrator, who – at the service of the collective sender – invites false make beliefs in the recipient. This form of deception underlies the salient category of film that is anchored in the presentation of the key protagonist’s warped perception of the fictional reality (e.g. Anderson, 2010; Ferenz, 2008; Klecker, 2013; Koch, 2011; Laass, 2008). However, this focalisation through a misguided or incapacitated character is by no means the only application of this type of deception, as Example 1 illustrates (Fig. 1). Example 1: Episode 16, Season 7 [Dr Lisa Cuddy split up with Dr Greg House after he had failed her by missing her cancer tests and by starting to abuse Vicodin again after seemingly successful rehab. Devastated, he engages in frivolous distractions in a hotel room where he is staying, with Dr Wilson, his only friend, checking up on him from time to time.]

When Both Utterances and Appearances are Deceptive

219

220

M. Dynel

When Both Utterances and Appearances are Deceptive

221

222

M. Dynel

When Both Utterances and Appearances are Deceptive

223

224

M. Dynel

When Both Utterances and Appearances are Deceptive

225

226

M. Dynel

Fig. 1  Example 1. Episode 16, Season 7

In this series of consecutive phases, the multimodal abundance of clues leads the recipient to develop a false belief that House indents to (Phases 3, 4, 6 and 7), and does (Phase 8), commit suicide only to realise that this is not the case (Phase 9). The initial interaction between House and the bartender (Phase 1) indicates that the former is in a bad mood and just wants to get inebriated and dull his senses, given that Cuddy has terminated their intimate relationship. The next two Phases (2 and 3) seem to picture House as pining for the olden days that are never coming back. Mesmerised by the young people celebrating (their voices resonating in his head), he appears to be coming to a realisation that he is advanced in age, that he cannot have such fun, and that he cannot ever be happy in general. These dark thoughts are metaphorically represented by the troubling extradiegetic sounds. The close shots of House are aimed at inspiring in the recipient a feeling of intimacy and empathy with him. As the story unfolds, with the benefit of hindsight, the recipient can also conjecture that it is already in the pub that House may have come up with his suicidal plan. This plan becomes evident when House, whilst sitting in the messy room, takes the Vicodin he has left (Phase 4) and contemplates his jump from the balcony (Phases 6 and 7), which is shown to be very high thanks to a shot from inside the room. This interpretation is reinforced by Peter Gabriel’s gothic-like orchestral rendition of Arcade Fire’s My Body Is a Cage, whose instrumental part comes to a climax together with the visual representation of the alleged suicide. The lyric alone

When Both Utterances and Appearances are Deceptive

227

(about mental and physical entrapment) appears relevant to House’s plight (his physical handicap and the constant pain in the mutilated thigh muscle, his failure to sustain the relationship with Cuddy, and his general sense of loneliness and isolation). Another component contributing to this interpretation is the expression of fear and anxiety that draws on the face of Wilson – who has arrived to check how House is handling the post-break-up crisis (Phase 5) – when he sees his friend on the balcony railing high above (Phase 7), and his verbal reaction when House has jumped down (Phase 8). House’s flight downwards is dramatically presented in various shots from different angles (Phase 8). The recipient is thus invited to nurture as long as possible the make belief that House has decided to commit suicide and has jumped to his death. It needs to be stressed that none of the characters performs any deception. Most importantly, Wilson’s reactions of shock and angst are sincere given that House’s dive is indeed quixotic and extremely dangerous and can easily lead to severe injuries, if not death. The entire act of extradiegetic deception is duly revealed to the sole target, i.e. the recipient, who learns about House’s genuine intentions and the nature of his actions, this time corresponding to the fictional truth. This happens when the camera narrator shows House tucking up his knees, shouting “cannonball” and landing safely in the swimming pool (Phase 9). Consequently, the recipient needs to revise the previously developed make beliefs about House’s “epiphany” in the pub, and about his intentions as he was standing on the balcony railing and as he took a leap. The recipient may then infer that, regardless of how he was feeling in the aftermath of the break-up with Cuddy, House must have decided to mingle with the crowd of the partying students and to impress them with his diving stunt, thereby proving the bartender wrong. This appears to be the prototypical inferential path the collective sender has designed for the model viewer, the recipient, to follow in order to be successfully deceived through the camera narrator’s strategies. However, as is the case with any deception, not all viewers need to be taken in. Resisting to get enthralled in the fictional world, a viewer may rationalise that this is only episode 16  in the current season of House (and each season comprises from 22 to 24) and that Season 8 is planned (from the perspective of a viewer watching the televised premiere of the episode in 2013) or has already been released, and that the series cannot exist with the eponymous character, who has not shown any suicidal streak and who is usually impervious to any social influence. There are also on-screen deception-­indicating cues, such as that the people surrounding Wilson (Phase 7) look amused rather than mortified when gazing up at House. The pending query is what specific strategies of deception the cinematic narrator employs in Example 1. Four general strategies come across as being the most important. Firstly, one component of deception resides in the covert ambiguity of House’s nonverbal reactions (whose significance is boosted by the close-ups, the slow motion and sounds) to the celebrating students and the inspiration he gets based on that sight. The salient and contextually relevant interpretation of his thoughtfulness (and the fact that he is devastated and even suicidal) is later disconfirmed and ousted by another one, viz. House’s decision to impress the celebrating students. Secondly, seeking the relation between the seemingly unrelated consecutive scenes featuring

228

M. Dynel

the main character: the one in the pub and the one in the hotel room (taking the remaining Vicodin pills), as well as on the balcony (preparing for the jump), involves recognising the deceptive multimodal implicature that House is resigned to being drugged and committing suicide, having realised that his best time is gone. Thirdly, contrary to what purports to be manifest and what is consonant with the suicide scenario, the dramatic lyric and tune of the extradiegetic song is covertly irrelevant to the fictional reality, notably the protagonist’s plans and actions. Fourthly, the camera work (its positioning and angle) performs deception via withholding information by not showing the swimming pool below House’s balcony until the very moment he lands in it. This withholding information makes for another covert ambiguity underlying House’s action, namely a jump to his death vs an awe-­inspiring stunt. Overall, this example proves that cinematic strategies can function as pragmatic acts (Janney, 2012) and that the cinematic narrator can communicate extradiegetic deceptive meanings multimodally, with no character being responsible for or aware of them.

Intradiegetic Deception Performed by Characters The deception depicted in “Extradiegetic Deception Performed Only by the Cinematic Narrator” is multimodally performed by the cinematic narrator beyond the characters’ level of communication only to be revealed at the end of the same sequence of scenes. This can be juxtaposed with a case of deception that is performed by characters, with the cinematic narrator merely facilitating this by “objectively” reporting the deception-driven interactions to the recipient, who is not omniscient. Also, unlike in Example 1, the scenes involving deception are scattered across the episode, the very end of which discloses the relevant truth of the fictional world. For reasons of space, only two scenes are transcribed here (Fig. 2 and Fig. 3), while the other relevant ones are only briefly summarised. Example 2: Season 5, Episode 10 [Dr Kutner, a member of House’s team, set up an online second opinion clinic in the name of Dr House. Kutner reveals this fact to Dr Taub when asking his opinion on whether a leaking breast implant can cause joint pain. Kutner involves Taub in the online business so as to prevent him from sharing the news with House.] [Kutner tells Taub about another email from the “boob lady”, who threatens to sue him if he fails to diagnose her. House is close to overhearing this conversation.] [Taub and Kutner chance upon the patient by the name of DeeDee, a heavily tattooed blonde, in an elevator as she is on her way to see the Dr House with whom she has been exchanging emails regarding her joint paint and fatigue, as well as hair falling out in clumps. The real House, who is waiting for the two doctors vis-à-vis the elevator, catches a glimpse of DeeDee, Taub and Kutner on their way to the ER.] [During the team’s meeting, Kutner’s and Taub’s beepers go off. House reacts, “So who’s paging you? Your wife? Does it worry you that she paged Kutner first?”. After the meeting, Kutner and Taub rush off, knowing they were paged about DeeDee.]

When Both Utterances and Appearances are Deceptive

Fig. 2  Example 2, Part a

229

230

M. Dynel

[Kutner duly meets Dr Cameron and Dr Chase, House’s former team members, in the cafeteria. Chase agrees to scan DeeDee for biliary tumour in exchange for 25% of Kutner’s income.] [Kutner and Taub reach the ER area only to find DeeDee gone from her bed. A nurse tells them that the patient had a respiratory arrest and did not survive. Kutner and Taub stare at each other in horror.]

When Both Utterances and Appearances are Deceptive

231

232

M. Dynel

When Both Utterances and Appearances are Deceptive

233

Fig. 3  Example 2. Season 5, Episode 10 (part 2)

The above strand of the multimodal narrative (interweaving with a few others in this episode) concerns an elaborate act of deception targeted at Kutner and Taub (to whose mischief the recipient is privy from the beginning, being aware that they are deceptively withholding information from House) performed by House and his accomplices, namely DeeDee (an escort), and two former members of his team, Cameron and Chase. Throughout this act, the recipient is given the perspective of the intradiegetic target, and is thus deceived, similar to the two characters, only to discover this at the end of the episode thanks to the deceivers’ revelation. Essentially, House turns out to have found out about the illegitimate online business set up in his name and to have hired the prostitute to covertly pretend to be an online patient, who arrives at the hospital to feign not only strange bodily symptoms (e.g. the uncontrollable singing of Harry Nilsen’s Coconut) but also ultimately death and rebirth. Additionally, as House ultimately admits (Phase 6b), he invited Cameron and Chase to falsely corroborate the symptoms. All the deceiving characters perform verbal deception, mainly through lies, i.e. covertly untruthful assertions (e.g. Cameron’s “She could be having a partial seizure right now” in Phase 2a, and House’s “What

234

M. Dynel

she had was easily treatable”, “She could’ve been fine” and “Damn treatment’s so simple” in Phase 3b), as well as non-verbal deception, such as covert pretending to be dead (Phase 1b). In principle, intradiegetic deception performed on the characters’ level of communication can display the same forms as deception in real-life interactions (Dynel, 2018), being facilitated by cinematic strategies so that the recipient can be deceived too. The cinematic narrator carefully presents the relevant deceptive meanings (e.g. close-ups of the bleeding ear in 2a), whilst deceptively withholding crucial information (e.g. no scenes of House inviting his accomplices are present, and nor are those involving DeeDee’s preparation for the acts of deception). Also, the success of the deception of the recipient is co-determined by the genuine reactions of the diegetic targets manifest in the close-ups, as in Phases 3a, 1b and 4b). The same holds for the pleasure-giving surprise when House’s prank is revealed (Phase 6b), the effect being boosted by the quick shots of DeeDee inhaling and the two petrified doctors jumping back, in tandem with the tension-boosting extra-diegetic music. These strategies help deception succeed despite some cues for the deception that observant viewers may notice (e.g. Cameron notices the blood coming from DeeDee’s ear even though it does not really show until the former pulls the hair back in Phase 1a). As this example indicates, film deception may be performed strictly by characters interacting in the fictional/diegetic world, which, thanks to complex cinematic strategies, imitate real-life interactions that one eavesdrops on, with the interlocutors’ being oblivious to this fact. The receiver is deceived in tandem with the characters until being allowed to discover the fictional truth with the benefit of hindsight. The revelation may sometimes come at the very end of the film or in the same interaction. (On the other hand, deceptive acts rendered by a character towards another/others may be highly transparent to the recipient, who is never meant to be deceived, in which case there is no deceptive unreliability on the cinematic narrator’s part.) When a character commits an act of deception targeted at another character, he/she cannot ever be considered to have intended to deceive the recipient. A character is a fictional inhabitant of a fictional world oblivious to the recipient, who has access to the fictional world thanks to the cinematic narrator, not deceptive per se. It is, however, the collective sender who aims to deceive the recipient, using the characters as their mouthpieces and the cinematic narrator as a seemingly “unbiased” reporter that withholds some information to facilitate the deception. Overall, the different forms of deception that arise in a fictional world inhabited by characters are amenable to the same theoretical considerations as deception in real-life interactions, which fictional interactions aim to imitate (see Dynel, 2016, 2018). The recipient is offered, albeit frequently only in retrospect, insight into the interactants’ intentions and beliefs and hence their (un)truthfulness, with various discursive and cinematic strategies facilitating their understanding.

When Both Utterances and Appearances are Deceptive

235

 eception Performed Jointly by the Intradiegetic and Cinematic D Narrators The third type of deception, which is very often mentioned in the literature on the narrator’s unreliability in film (see “Can the Cinematic Narrator Lie or Just Deceive Otherwise?”), comes into being when the two categories of film deception presented in the previous two sub-sections co-occur. This happens when an interactant on the characters’ level becomes an intradiegetic narrator and deceptively recounts a story which is presented on screen in the form of a flashback, with both the narrator’s utterances and the multimodal representation being covertly untruthful relative to the fictional world and its truth. An intradiegetic narrator is also known as an invoking narrator (Black, 1986), for it is a character whose words invoke a multimodal illustration of the verbal narrative. It is widely acknowledged that a character-narrator on the diegetic level can lie in literary fiction (Ryan, 1981) and in film (Burgoyne, 1990: 7). An intradiegetic character-narrator “can lie (...) or distort the facts of the fictional world” (Stam et al., 1992: 102) in order to deceive the interlocutor (cf. deception at the characters’ level of communication in “Intradiegetic Deception Performed by Characters”), which can be coupled with an “untruthful account, in flashback”,11 with the camera being “at the service of the narrator” and conspiring with him/her (Chatman, 1978: 237). In this vein, Kozloff talks about visual “presentation” that “colludes with the narrator’s false [-believed] account of events” (1988: 115).12 On the other hand, as Chatman (1978: 235) states, “a voiceover depicting events and existents in the story may be belied by what we see so clearly for ourselves.” Alternatively, an intradiegetic narrator may only introduce the multimodal cinematic narrator’s presentation of a previous verbal interaction in its entire form, as is the case with Example 3 (see Fig. 4 and Fig. 5). Example 3: Season 2, Episode 8 [Stacy, the hospital lawyer, is investigating the case of Dr Chase’s alleged malpractice, as a result of which a patient died. She has just learnt from Chase that the patient had liver transplant, her brother donating the organ, and that it was House who had arranged for a surgeon to do this live donor transplant on very short notice. This scene of the interaction between Stacy and Chase cuts to what follows.]  If no such flashback is present, the deception is restricted to the characters’ level.  Kozloff (1988) also claims that only the “camera narrator” may deceive the viewer, while the narrating character is truthful. Such a case, if at all possible, would qualify as the type of deception presented in “Extradiegetic Deception Performed Only by the Cinematic Narrator”. On the other hand, one may also envisage the opposite scenario; the camera narrator may give a truthful account and so reveal the fact that a character is lying, or generally deceiving, through a narrative. This qualifies as intradiegetic deception (“Intradiegetic Deception Performed by Characters”), to which the recipient is privy. This is the case with a scene from Ang Lee’s Brokeback Mountain, when the wife mendaciously reports on the circumstances of her husband’s death to his male ex-lover, while the cinematic narrator presents different events, which are to be interpreted as corresponding to the fictional world’s truth.

11 12

236

M. Dynel

When Both Utterances and Appearances are Deceptive

237

Fig. 4  Example 2, Part a

[Stacy is adamant that House should tell her the truth as, in her capacity as his attorney, she cannot tell anybody. Another flashback follows.]

238

Fig. 5  Example 3, Part b

M. Dynel

When Both Utterances and Appearances are Deceptive

239

This example presents an act of verbal deception performed by the intradiegetic narrator, a character who is simultaneously one of the two interactants in conversation in the flashback presented by the cinematic narrator. The cinematic narrator takes the position of the intradiegetic narrator’s mouthpiece, reporting to the recipient what the latter is telling Stacy in the current interaction in a multimodal manner (Phases 1a and 2a). Stacy sees through House’s deception (Phase 3a), which prompts him to give a truthful, as it seems, account of his interaction with the surgeon: the unsuccessful attempt at bribery (Phase 1b) and on the successful blackmail (Phase 3b). This s­ econd account sheds new light on the preceding narrative, revealing its deceptive nature also to the recipient. Therefore, the first flashback (Phases 1a–b) turns out to have been riddled with lies. The only truthful aspects, the ones true in the fictional world, seem to have been that House did go to meet the surgeon in his office so that the latter would do the transplant, and the surgeon did examine the patient’s X-ray. Nonetheless, all the utterances and non-verbal messages exchanged by the interlocutors turn out to have been House’s lies. It can be surmised that House was not friendly, House did not politely ask or cajole the surgeon into doing the transplant through flattery, the surgeon did not give a smug smile or agree to do the risky procedure thanks to House’s ingratiation, the two doctors did not shake hands, etc. All the mendacious messages in House’s intradiegetic narrative (Fig. 4), which would take numerous utterances to be conveyed in the form of a verbal account, are neatly communicated on screen in a multimodal manner by the cinematic narrator. The latter thus presents the interaction that never took place in this form and is rife with deception, specifically multimodal lies (see “Can the Cinematic Narrator Lie or Just Deceive Otherwise?”). Needless to say, cinematic techniques collude with the character narrator’s deceptive account and enhance the effect of the deceptive utterances and interactant’s non-verbal messages presented as an i­nteraction. For instance, in Phases 1a and 2a, the camera frames the surgeon as looming over House, as if he is superior towards the latter, which is covertly untruthful, given House’s contemptuous attitude towards him. Two interesting points are worth addressing with reference to this example. The revelation flashback contains a nested act of momentary deception targeted at the surgeon and performed non-verbally by House, which is non-deceptively reported on by the cinematic narrator and the intradiegetic narrator. In Phase 2b, House purports to be retreating (as he takes the envelope and heads for the door) only to attack again and manipulate the surgeon through blackmail. On the other hand, the cinematic narrator seems to be responsible for another manifestation of deception that is independent of House’s mendacious account. A viewer may interpret the flashback as being a representation of Chase’s account in the preceding scene. Based on a dialogue hook and discontinuity editing (see Bordwell and Thompson 2013 [1979]), the recipient can rightly infer that the grey-haired man on screen must be the surgeon just mentioned by Chase (a multimodal implicature generated through the Gricean Relation maxim flouting), and that the interaction on the screen is another flashback of Chase’s. However, it is only in Phase 3a that the recipient realises that the narrator is House, rather than Chase. The recipient is thus deceived about who the intradiegetic narrator is through the cinematic narrator’s withholding information (editing out) of the scene that shows a change in Stacy’s interlocutor.

240

M. Dynel

Can the Cinematic Narrator Lie or Just Deceive Otherwise? Previously, deception in film has been most often addressed with reference to what was presented in the section “Deception Performed Jointly by the Intradiegetic and Cinematic Narrators”: the intradiegetic narrator (Genette (1980 [1972]) gives a deceptive account of prior events, with the cinematic narrator presenting it in a flashback (Chatman, 1978; Kozloff, 1988; Stam et  al., 1992). Claims have been made that in such a situation the camera narrator seems to be able to “lie”, contrary to the well-established convention (Chatman, 1978: 237, 1990: 132; Kozloff, 1988: 115; Anderson, 2010). This case (cf. Example 3 here) is typically illustrated with the (in)famous example of Alfred Hitchcock’s Stage Fright, at the beginning of which the main male character, Johnny, is talking to Eve and recounts a falsebelieved story of his meeting with Charlotte. A crossfade to Jonathan in his kitchen indicates the beginning of a “lying flashback”, in which the multimodal presentation “colludes with the narrator’s false account of events” (Kozloff, 1988: 115). This scene is known to have caused an angry outcry from critics (Kozloff, 1988). Bordwell (1985: 61) concludes that the narration is “duplicitous (…) by appearing to be highly communicative – not just reporting what the liar said but showing it as if it were indeed objectively true.” In this vein, Anderson (2010: 84) observes that by “invoking the conventions for a flashback, but presenting instead a dramatization of Johnny’s lie, the film misreports diegetic events—the film lies to us.” Interestingly, Hitchcock himself considered this a mistake with the benefit of hindsight, “I did one thing in [Stage Fright] that I never should have done; I put in a flashback that was a lie” (Truffaut, 1983: 189). This kind of lie is also considered a violation of the basic principle of extradiegetic narration (Thompson 1977). However, as evidenced by Example 3, Hitchcock’s lying flashback is by no means an isolated incident, and this narrative strategy is indeed employed in contemporary cinematography. Many philosophers have tried to vindicate Hitchcock’s cinematic strategy, or at least, account for its workings in the light of narrative theory. Casetti (1986), proposes that the “false” images (and also utterances, which he ignores) represent what the narrating character’s addressee pictures in her head while listening to the homodiegetic narration. However, it is intuitive to assume that “the deceptive images and their juxtaposition must be thought of as representations of Johnny’s account, though we begin by taking them also to be representations of what is real with the fiction itself” (Currie, 1995: 27, emphasis in original). Following the same premise, Chatman (1986: 141) goes a step further and states that, when the flashback is used, Hitchcock’s Johnny is the only narrator and “everything we see and hear originates with him (...) since every cinematic tool – editing, lighting, commentative music, etc. – contributes to his intention.” Chatman (1990: 132) maintains that a narrating character like this “is ‘responsible’ for the lying images and sounds that we see and hear”, and he invokes the role of the invariably reliable implied author, who does allow the facts to emerge ultimately. In a similar vein, Burgoyne suggests that the intradiegetic narrator in Stage Fright is responsible for the multimodal lying flashback which relies on using both images and characters’ utterances and that the cinematic narrator “controls the entire cinematic apparatus” (1990: 11).

When Both Utterances and Appearances are Deceptive

241

Burgoyne (1990: 7) thus claims that a personal character-narrator (on the intradiegetic level) can utilise not only words but also the multimodal tools available to the cinematic narrator in order to salvage his central premise that the cinematic narrator cannot lie but can retroactively invalidate the intradiegetic narrator’s false account. This explanation results in conflating the narrative of the intradiegetic character narrator with the narrative that only the cinematic narrator can produce. The views endorsed by Chatman (1986, 1990) and Burgoyne (1990) are not entirely appealing. Surely, although the mendacious account originates with a character like Johnny, he cannot lie to us, the viewers, for “he is unaware not only of us but of the world we inhabit” (Anderson, 2010: 88). Additionally, as Currie (1995: 27) rightly observes, this narrating character “exists within the story, and it is no part of that story that he produced and edited cinematic images in order to convince his fictional fellows (and us?)”. It is then impossible for a character to be in charge of various cinematic techniques, which would involve his/her ontological superiority over the story in which he/she partakes (Anderson, 2010). The tools that the two narrators, intradiegetic and cinematic, have at their disposal are fundamentally different. The cinematic narrator can tell a false-believed story in a flashback showing a distorted interaction or just images while the intradiegetic narrator can produce only an oral narrative. The pending question is then who deceives and who discloses the deception to the recipient. For his part, Currie (1995) proposes that it is an implied author who reveals the extradiegetic narrator’s deception to the viewer. Nonetheless, the “implied author” seems to be an otiose concept proposed, besides real authors and narrators, by many scholars (e.g. Booth 1983 [1961]; Chatman, 1978, 1990; Currie, 1995; Stam et  al., 1992; cf. Sternberg & Yacobi, 2015 and references therein). Arguing against Chatman’s (1978) implied author (necessitating an implied reader), as the creator of fiction (besides the narrator, i.e. the presenter), Bordwell (2008) rightly states that the agents in the process of fiction reception are real authors and receivers, rather than the unreal agents that Chatman proposes. However, Bordwell’s (2008: 128) metaphorical roller-coaster explanation  – “I don’t have to imagine a ghostly intelligence standing between the engineer and me, shaping the thrills and the nausea I feel” – misses the basic fact about film watching: it is based on make believing about the fictional narrative, rather than experiencing non-narrated real stimuli, such as a roller-coaster ride. Also, immersed in the fictional world, the recipient does not take cognisance of the film production crew (or literary authors), the real but hidden (rather than “implied” for the recipient to infer) authors, who stand behind extradiegetic narrators telling the story. In any case, the notion of “implied author” is not endorsed here, based on Ockham’s razor, the principle of theoretical parsimony. Here, preference is shown for this simple explanation: the cinematic narrator (representing the collective sender) legitimately deceives, specifically lies to, the recipient (while the narrating character deceives the addressee) only to disclose this later. Thus, the multimodal narrative shown in a flashback is not the product of the intradiegetic narrator at all (who can only be held accountable for a running verbal commentary),13 but rather the cinematic narrator, the entity responsible for the 13

 Sometimes such a commentary may run off-screen and accompany the silent flashback images.

242

M. Dynel

whole film communication (indirectly, also the intradiegetic narrator’s verbal account), as constructed by the collective sender. Incidentally, the recognition of the fact that narrative deception is not an immoral but rather pleasure-giving strategy purposefully employed by filmmakers might have pre-empted the need to account for who the guilty party is. The film production crew behind the unreliable cinematic narrator does deceive the recipient with regard to some aspect of the fictional world, leading the latter to develop false make beliefs. However, all film deception (regardless of how it is performed) is ultimately revealed, giving the recipient a pleasurable surprise (see Dynel, 2013). Overall, the prevalent view about the illegitimacy of multimodal lies, and hence by extension, all deception in fiction may be consequent upon the traditional well-­ entrenched view that heterodiegetic narration, especially when impersonal (cf. Ryan, 1981),14 rules out unreliability. Many authors have followed Doležel’s (1998: 149) claim that the heterodiegetic narrator “cannot lie or err” given the performative nature of his heterodiegetic narration. This narrator, who has authentication authority, is responsible for the creation of the fictional world, being unable to “distort” or “falsify” it (Cohn, 2000: 311), unlike a homodiegetic personal narrator, who merely reports on the events. The reason why the narrator might be regarded as inherently reliable, or here truthful, is that “in fictional communication, the text constitutes the reader’s sole source of information about the represented state of affairs”, which involves the recipient’s “taking the fictional world to exist independently of the narrator’s declarations, while using these declarations as material for construing this world” (Ryan, 1981: 530). Consequently, “[e]verything the impersonal [extradiegetic] narrator says yields a fact for the fictional world” and his “lack of personality protects him from any kind of human fallibility” (Ryan, 1981: 534). Nevertheless, unreliable narration, and thus deception, are actually commonplace not only in homodiegetic narration but also in heterodiegetic narration (see also Zipfel 2011). With reference to unreliable literary narration, Stühring (2011: 96) states that a reader can be deceived about fictional facts in two ways, either by being given “wrong” information via “what is said” about what is the case in fiction, or by having relevant information withheld from them “at the point where it would have to be given”. Essentially, these two cases seem to correspond to two forms of deception: lying and withholding information respectively. Similarly, claiming that narrators perform three main roles (reporting, interpreting, and evaluating), Phelan (2005) proposes that each shows a “mis-” category and an “under-” category, which reflects a contrast between being “wrong”, i.e. false-believed, and being “insufficient”, which again appear to be consonant with lying and withholding information respectively. A weaker view seems to hold that the extradiegetic impersonal narrator, i.e. the cinematic narrator, cannot lie but can be deceptive otherwise, without making mendacious multimodal assertions. Many authors seem to have advocated this view, even if not labelling technically the forms of deception (other than lying) involved.

 It should be noted that homodiegetic narrators are always personal, while heterodiegetic narrators can be personal or impersonal, with the two being often addressed collectively. The notion of heterodiegetic impersonal narrator is the most important here.

14

When Both Utterances and Appearances are Deceptive

243

Although philosophers of fiction rarely discuss formally the categories of deception known in philosophy (see Dynel, 2018 for an overview), some relevant observations have been made regarding unreliable narrators in literary works (cf. the paragraph above). These may be seen as pertinent to the camera narrator’s deceptive multimodal messages in films. As Burgoyne puts it, “[b]ecause the narrator produces the discourse through which the viewer reconstructs the fictional world, this discourse comprises the facts of the fictional universe, which always carry the value of authenticity. Consequently, the discourse of the impersonal narrator in film is always reliable in the most basic sense: this type of narrator cannot lie about the fictional world, although the narrator can withhold information and cause the spectator to make incorrect inferences” (1990: 7). Withholding information is indeed frequently addressed as a strategy of the cinematic narrator’s deception. For instance, according to Fink (1982: 24), there is a “silently accepted convention” that on-screen presentations “may omit something but never distort the truth”15 and so they can never be “false”. Kozloff (1988: 115) also allows for the fact that film images can involve a partial presentation of what is true in fiction. Referring to a voice-over flashback in a screen adaptation of Agatha Christie’s Evil under the Sun, Kozloff (1988: 115) reports that the “two murderers, naturally, lie” as they are recounting their alibi to Poirot, while the flashbacks shown on screen are “not false” but are “just partial”.16 Regardless of whether the narrating characters are telling lies and/or performing other forms of verbal deception in practice, technically, the “partial” shots do deceptively withhold information (Dynel, 2018, 2019) through elliptical editing, as well as camera distance and angle, which allow some crucial information to remain covert from the recipient’s perspective. For his part, Koch (2011: 70) addresses “mimetically deceptive” unreliable narration of films, such as A Beautiful Mind, that “play out a delusive strategy by way of the presentation of distorted fictitious sense-data that are not marked accordingly.” Thus, the key narratorial characteristic of such films is “not misrepresenting the fictitious world but presenting someone who misperceives it,” or more adequately presenting this character’s distorted view of the fictional world (Koch, 2011: 70). This explanation is relevant to some of Elsaesser’s (2009) “mind game films” and to some of Klecker’s (2013) “mind-tricking narratives”. In these cases, the cinematic narrator offers multimodal representations of the focalising characters’ (Anderson, 2010) genuine lived mental experience that actually diverts from the fictional reality, about which the recipient is retroactively informed. What the cinematic narrator initially presents as the diegesis later turns out to have been, in total or in part, something untruthful relative to the fictional world, being the focalising character’s figment of imagination, hallucination, delusion, dream, etc. Thus, the recipient needs to backtrack on the meanings generated based on the previous scene(s) or even the entire film and reframe them as being the character’s figment of  Technically, this distortion of truth can be performed by various means, not only through lying, which Kozloff (1988) seems to have in mind. 16  The same shots, albeit re-edited are used later at the end of the film to illustrate Poirot’s account of what he believes to have happened (his believed truth – and given Poirot’s infallibility, objective truth – in the fictional world). 15

244

M. Dynel

imagination, hallucination, delusion or dream, which does not correspond to the fictional world and its truth. This broad category of film deception originates from the primary source of the representation: it is not the omniscient and infallible cinematic narrator’s perspective but the focalising character’s perspective affected by his/her mental state that modifies the fictional reality or is completely divorced from it, which the cinematic narrator merely reports. The challenge of the production crews collective sender in such films is “to avoid shots that amount to forthright false statements or, in other words, to provide a twist that allows, resorting to poetic licensing, for a consistent ex post facto reconstruction of the complete discourse” (Koch, 2011: 74). However, it may be contended that multimodal lies are not ruled out, with the cinematic narrator presenting interactions or events that are the focalising character’s mental creations from start to finish but not the truth of the fictional world, as is typically the case with dreams shown on screen, unannounced; or a character’s personified figment of imagination representing the character and interacting with others in the fictional world. Such untruthful (relative to the fictional world) assertoric messages are not considered lies given the retroactive reframing. This kind of focalising/filtering through characters, who may be adjudged tacit, deceptively covert intradiegetic (unintentionally) unreliable narrators,17 is usually restricted to the representations of the diegetic world determined by the characters’ experienced mental states, whether involuntary (dreams or hallucinations) or voluntary (imaginings). However, focalisation may also be done through a character’s mendacious narrative, which has been depicted as a markedly different form of the cinematic narrator’s deception. As Anderson puts it, by “focalizing through a character who, intentionally or otherwise, mischaracterizes diegetic reality, the cinematic narrator can present a false version of the story” (2010: 89, emphasis added). Comparing The Sixth Sense and Stage Fright, both of which focalise through unreliable characters, Anderson (2010: 87), states that the former “misleads us—underreports events—while the latter lies to us—misreports events.” Anderson (2010) maintains that in presenting the story through the skewed perspective of a character who misinterprets his/her experiences, as in The Sixth Sense, the multimodal narrator does not lie but deceives the viewer also through ambiguity. Similarly, Klecker claims that some mind-­tricking films are based on “withholding information” but do not “actually lie in the sense of deliberately conveying information that is untrue. They leave strategically placed gaps that offer two possible interpretations—one during the first viewing and another (the final one) in retrospect, upon the disclosure of the surprise gap” (2013: 136, emphasis added). It needs to be underscored that covert ambiguity and deceptively withholding information are two different forms of deception. Indeed, “underreporting”, or rather withholding information through not revealing a basic fact about the protagonist in The Sixth Sense (i.e. that he is dead) invites covert ambiguity that underlies  Anderson (2010: 89) claims that “in films that lie to or mislead the viewer, one character is almost always the explicit focalizer.” However, deception performed by the cinematic narrator, which Anderson (2010) seems to have in mind, need not always be done though character focalisers, as evidenced by Example 1.

17

When Both Utterances and Appearances are Deceptive

245

the whole film in a garden-path manner (see Dynel, 2009). Essentially, the first part of a (here, multimodal) text must entail covert ambiguity, with only one meaning being effortlessly accessible to the hearer/recipient, and the second part of the text (the climax) must surprisingly invalidate the recipient’s previous inference and prompt him/her to backtrack and reprocess the previous part of the text to appreciate an alternative, hitherto unobserved, meaning congruent with the meaning revealed in the climax. However, specific scenes may involve other deceptive strategies (e.g covert irrelevance, when the late man is talking to his wife, who is only doing self-talk). This pattern based on withholding information and covert ambiguity underlies many films pivoted on multimodal deception that holds until the climax, not always involving character focalisation (cf. Examples 1 and 2). Whether or not filtering through one character, the tacit unreliable narrator, is deployed, the climax reveals the preceding ambiguity and prioritises an alternative, truthful reading on the entire film (e.g. a character is schizophrenic and has no friend, or a character wants to impress the crowd rather than committing suicide, etc.). What is important is that the specific scenes and phases throughout the film may deploy other forms of multimodal deception and cannot be reduced to the broadly understood covert ambiguity, as corroborated in the analyses in “Types of Multimodal Deception in Film”. Different forms of deception addressed in language philosophy may be deployed jointly across modalities by the cinematic narrator, multimodal lies included (e.g. when a character is shown conversing with a person that later turns out to have been his hallucination rather than a real person inhabiting the fictional world, as in Fincher’s Fight Club), to serve an overarching deceptive purpose that stems from the covert focalisation, which typically prevents the cinematic narrator from being held accountable for lying. A remaining query is whether the flashback in Stage Fright does involve cinema narrator’s lying when focalising through Johnny and his mendacious account, and whether this practice can be considered legitimate at all (see also Example 3). Indeed, this multimodal flashback is one of the scenes that “show us things that never occurred as if they had occurred; they can manifestly lie to the viewer about the diegetic world” (Anderson, 2010: 84). In this case, the cinematic narrator presents the diegetic world as filtered not through a confused or mentally incapacitated character (together with the lived experience, this seems to work as extenuating circumstances for the deceptive cinematic narrator, thus typically not considered to be lying) but, instead, through a character who is purposefully mendacious in his interaction with another character, offering a multimodal representation of his lies, and hence telling multimodal lies, which can be vindicated – it is argued here – by the embedded narration. The cinematic narrator’s lying flashback gives the mendacious perspective of the character who acts as an intradiegetic narrator of the subordinate story. It is not the case that the intradiegetic narrator uses the cinematic narrator (Burgoyne, 1990) but vice versa, it is the cinematic narrator that uses the intradiegetic narrator and lends the latter support by rendering his narrative through a multimodal narrative in a flashback only to disclose the intradiegetic narrator’s

246

M. Dynel

deception at the end of the film.18 This strategy seems to be considered controversial inasmuch as the ultimate reframing of the account as only a focalised fabricated version of events does not cancel or suspend the lies relative to the fictional world, which the other forms of reframing (e.g. as dreams or hallucinations) do, possibly even denying the status of assertions to the previous untruthful messages.19 Nonetheless, even if, intuitively, more controversial than the other forms of focalisation involving the character’s genuine lived mental experience that make for untruthful multimodal statements (apart from using other forms of deception), the focalisation through a mendacious speaker who produces a false-believed narrative of “non-lived” experience can be regarded as being equally legitimate in cinematic narration. All focalising situations may involve what later turns out to have been a sequence of untruthful multimodal messages and even assertions, and hence potential lies about the fictional world and its truth, retroactively reframed as belonging in a dream, illusion, delusion or, as is the case with Hitchcock’s famous scene, as part of an intradiegetic narrator’s mendacious account dramatised multimodally for the recipient’s sake. Unlike in the case of deceptive focalisations through character minds’ fabrications, the recipient is aware of the focalisation through a narrating character who performs verbal lies; it is the narrating character’s mendacity that is covert, and so the retroactive reframing does not concern the very aspect of focalisation. In this case, the cinematic narrator reports the character’s lies (and other deceptive utterances) to the recipient as if oblivious to the intradiegetic narrator’s (un) truthfulness. In practice, through the cinematic narrator’s focalising and later reframing of the multimodal lies, the collective sender aims to give the recipient a cognitive s­urprise, like through any other form of film deception targeted at the recipient (i.e. not revealed to the recipient at the moment of production). It is hoped that this discussion disperses the doubt that critics instilled into Alfred Hitchcock: “Strangely enough, in movies, people never object if a man is shown telling a lie. And it’s also acceptable, when a character tells a story about the past, for the flashback to show it as if it were taking place in the present. So why is it that we can’t tell a lie through a flashback?” (Hitchcock in Truffaut, 1983: 189). As has been argued here, multimodal lies can be told through the cinematic narrator when (and only when) some form of character focalisation is used, whether the character is purposefully deceptive and his/her verbal lies are cinematically narrated as a multimodal lying flashback (the controversial case of non-lived experience) or has a warped view of the fictional world, is dreaming or purposefully imagining things (the widely accepted forms of the cinematic narrator’s mendacity involving a tacit intradiegetic narrator and the representation of their lived experience). These are fine specimens of multimodal film deception targeted at the recipient.  A statement may also be ventured that Hitchcock’s film would not have stirred up so much controversy if the multimodal lies had been revealed to the recipient much sooner (cf. Example 3). 19  The deception based on unannounced presenting a dream or hallucination through multimodal statements is similar to producing an utterance that is meant to be taken as a truthful assertion but is duly reframed as autotelic humour, whereby the non-assertoric nature of the statement is revealed (see Dynel, 2017, 2018). On an alternative account, this deception may then be considered to involve covert pretending to make an assertion rather than a mendacious assertion. 18

When Both Utterances and Appearances are Deceptive

247

Conclusions and Final Comments This paper has given new insight into film deception in the context of relevant postulates made within the fields of film pragmatics and multimodality, the philosophy of fiction, including narrative theory, and the philosophy of deception. Deception in film can materialise through not only characters’ verbalisations and non-verbal messages (i.e. signals, actions and artefacts), but also complex cinematic strategies. It has been shown that film is an elaborate multimodal artefact which can involve three main forms of film deception amenable to analysis also in the light of categories of deception (including lies) teased out by language philosophers, which can easily co-occur and be interdependent. The fictional world constructed on screen can invite false make beliefs in viewers, as intended by the film production crew who use the cinematic narrator that reports on characters’ interactions. The recipient is deceived by the production crew when he/she entertains false beliefs about what is the case in the fictional world at hand, typically without explicitly crediting the production crew (let alone the extradiegetic cinema narrator) for this. Film deception may manifest itself on the characters’ level of communication, being consequent upon fictional interactants’ intradiegetic deception of one another (with the recipient being deceived or not) as shown on screen; and only on the recipient’s level, which is when the recipient, but no fictional interactant, is deceived by the extradiegetic camera narrator, to whose presence the recipient remains oblivious when immersed in the fictional world. These two general forms of deception may co-­occur, which is when the extradiegetic narrator offers a multimodal representation of an intradiegetic narrator’s deceptive account. Each of the three types of film deception may materialise through any of the various types of (real-life) deception examined by philosophers (e.g. lying, deceptive implicature, withholding information, covert ambiguity or covert irrelevance). Contrary to popular opinion, multimodal lies are a legitimate, and by no means intermittent, cinematic strategy necessarily facilitated by focalising characters, who  – typically  – have a warped perception of the fictional world or who may be dreaming/imagining things (cf. their lived experience), but who may  –  more controversially  –  be narrating their mendacious accounts to other characters. Cinematic narrators thus reveal the deception by reframing the multimodal lies (and other deception) as being only a matter of individual mental experience rather than objective facts. Importantly, the recipient must recognise the presence of film deception at some point for its designed effects to hold. Either viewers are privy to deception being performed by a character, or they must recognise a character’s or the narrator’s, and hence the film crew’s, deception with the benefit of hindsight, whether in the same interaction, in a later one, or even at the very end of the film. Having been deluded, the recipient is invited to learn the truth in the fictional world, or more precisely what he/she make believes to be the truth in the fictional world (which, ideally, corresponds to what the production crew seem to regard as the fictional truth to be discovered by the recipient). The recipient’s recognition of film deception necessitates world-repair or word-replacement in the incremental process of interpretation based on new incoming information (cf. Gavins, 2007).

248

M. Dynel

It is important to note that film deception is essentially performed for the sake of the recipient and should not be considered morally reprehensible. Viewers engaged in the act of watching a film develop false make beliefs, which do not carry any repercussions in the real world. This is why film deception and recognition thereof give the recipient cognitive pleasure rather than a sensation of being immorally fooled or taken advantage of, as is frequently the case with real-life deception.

References Anderson, E. (2010). Telling stories: Unreliable discourse, Fight Club, and the cinematic narrator. Journal of Narrative Theory, 40(1), 80–107. https://doi.org/10.1353/jnt.0.0042. Aquinas, T. (1972). Question 110: Lying. In Summa theologiae, Volume 41: Virtues of justice in the human community. New York, NY: McGraw-Hill. Augustine. (1952). Lying. In R.  J. Deferrari (Ed.), Treatises on various subjects (pp.  53–120). New York, NY: Catholic University of America. Badura, C., & Berto, F. (2018). Truth in fiction, impossible worlds, and belief revision. Australasian Journal of Philosophy, 97(1), 178–193. https://doi.org/10.1080/00048402.2018.1435698. Bal, M. (1997 [1992]). Narratology. In Introduction to the theory of narrative. Toronto, Canada: University of Toronto Press. Baldry, A., & Thibault, P.  J. (2006). Multimodal transcription and text analysis. London, UK: Equinox. Banfield, A. (1978). Where epistemology, style and grammar meet: The development of represented speech and thought. New Literary History, 9(3), 415–445. https://doi. org/10.2307/468449. Bateman, J., & Schmidt, K.-H. (2011). Multimodal film analysis: How films mean. London, UK: Routledge. Black, D. A. (1986). Genette and film: Narrative level in the fiction cinema. Wide Angle, 8(3/4), 19–26. Bok, S. (1978). Lying: Moral choice in public and private life. New York, NY: Random House. Booth, W. C. (1983 [1961]). The rhetoric of fiction. Chicago, IL: University of Chicago Press. Bordwell, D. (1985). Narration in the fiction film. Madison, WI: University of Wisconsin Press. Bordwell, D. (1989). Making meaning. Inference and rhetoric in the interpretation of cinema. Cambridge, MA: Harvard University Press. Bordwell, D. (2008). Poetics of cinema. New York, NY: Routledge. Bordwell, D., & Thompson, K. (2013 [1979]). Film art: An introduction. New  York, NY: McGraw-Hill. Branigan, E. (1984). Point of view in the cinema: A theory of narration and subjectivity in classical film. Berlin, Germany/New York, NY: Mouton. Branigan, E. (2013). Narrative comprehension and film. London, UK: Routledge. Bubel, C. (2008). Film audiences as overhearers. Journal of Pragmatics, 40, 55–71. https://doi. org/10.1016/j.pragma.2007.10.001. Burgoyne, R. (1990). The cinematic narrator: The logic and pragmatics of impersonal narration. Journal of Film and Video, 42, 3–16. Carson, T. L. (2010). Lying and deception: Theory and practice. Oxford, UK: Oxford University Press. Casetti, F. (1986). Antonioni and Hitchcock: Two strategies of narrative investment. SubStance, 51, 69–86. https://doi.org/10.2307/3684715. Chatman. (1986). Review of narration in the fiction film. Wide Angle, 8(3), 139–141. Chatman, S. (1978). Story and discourse. Ithaca, NY: Cornell University Press. Chatman, S. (1990). Coming to terms: The rhetoric of narrative in fiction and film. Ithaca, NY: Cornell University Press.

When Both Utterances and Appearances are Deceptive

249

Chisholm, R., & Feehan, T. (1977). The intent to deceive. The Journal of Philosophy, 74, 143–159. https://doi.org/10.2307/2025605. Clark, H. (1996). Using language. Cambridge, UK: Cambridge University Press. Clark, H., & Van Der Wege, M. (2001). Imagination in discourse. In D.  Schiffrin, D.  Tannen, & H.  E. Hamilton (Eds.), The handbook of discourse analysis (pp.  772–786). Oxford, UK: Blackwell. Cohn, D. (2000). Discordant narration. Style, 34, 307–316. Currie, G. (1990). The nature of fiction. Cambridge, UK/New York, NY: Cambridge University Press. Currie, G. (1995). Unreliability refigured: Narrative in literature and film. The Journal of Aesthetics and Art Criticism, 53, 19–29. https://doi.org/10.2307/431733. Currie, G. (2004). Arts and minds. Oxford, UK: Oxford University Press. Currie, G., & Ichino, A. (2013). Imagination and make-believe. In B. Gaut (Ed.), The Routledge companion to aesthetics (pp. 320–329). London, UK: Routledge. Daugherty, J. (2007). Narrative discourse revisited: A synopsis of Gérard Genette’s Nouveau Discours du Récit. Journal of Literary Studies, 1, 59–66. https://doi. org/10.1080/02564718508529753. DePaulo, B. (2010). Deception: It’s what Dexter does best (well, second best). In B. DePaulo (Ed.), The psychology of Dexter (pp. 65–78). Dallas, TX: BenBella Books. Doležel, L. (1980). Truth and authenticity in narrative. Poetics Today, 1, 7–26. https://doi. org/10.2307/1772407. Doležel, L. (1998). Heterocosmica. Fiction and possible worlds. Baltimore, MD: Johns Hopkins University Press. Dynel, M. (2019). To say the least: Where deceptively withholding information ends and lying begins. Topics in Cognitive Science, 1–28. https://doi.org/10.1111/tops.12379 Dynel, M. (2009). Humorous garden-paths: A pragmatic-cognitive study. Newcastle, UK: Cambridge Scholars Publishing. Dynel, M. (2011a). “You talking to me?” The viewer as a ratified listener to film discourse. Journal of Pragmatics, 43, 1628–1644. https://doi.org/10.1016/j.pragma.2010.11.016. Dynel, M. (2011b). Stranger than fiction. A few methodological notes on linguistic research in film discourse. Brno Studies in English, 37(1), 41–61. Dynel, M. (2013). Humorous phenomena in dramatic discourse. The European Journal of Humor Research, 1, 22–60. Dynel, M. (2016). Killing two birds with one deceit: Deception in multi-party interactions. International Review of Pragmatics, 8, 179–218. https://doi.org/10.1163/18773109-00802002. Dynel, M. (2017). But seriously: On conversational humour and (un)truthfulness. Lingua, 197, 83–102. https://doi.org/10.1016/j.lingua.2017.05.004. Dynel, M. (2018). Irony, deception and humour: Seeking the truth about overt and covert untruthfulness. Mouton series in pragmatics. Berlin, Germany: Mouton de Gruyter. Ekman, P. (1985). Telling lies: Clues to deceit in the marketplace, politics, and marriage. New York, NY: Norton & Company. Elleström, L. (Ed.). (2010). Media borders, multimodality and intermediality. Basingstoke, UK: Palgrave Macmillan. Elsaesser, T. (2009). The mind-game film. In W. Buckland (Ed.), Puzzle films. Complex storytelling in contemporary cinema (pp. 13–41). London, UK: Wiley-Blackwell. Fallis, D. (2010). Lying and deception. Philosophers’ Imprint, 10, 1–22. Ferenz, V. (2008). Don’t believe his lies: The unreliable narrator in contemporary American cinema. Trier, Germany: Wissenschaftlicher Verlag Trier. Fink, G. (1982). From showing to telling: Off-screen narration in the American cinema. Littérature d’América, 3(12), 5–37. Fludernik, M. (1999). Defining (in)sanity: The narrator of The Yellow Wallpaper and the question of unreliability. In W. Grunzweig & A. Solbach (Eds.), Grenzüberschreitungen: Narratologie im Kontext/Transcending boundaries: Narratology in context (pp. 75–95). Tübingen, Germany: Narr.

250

M. Dynel

Forceville, C. (2010). The Routledge Handbook of Multimodal Analysis: Carey Jewitt (Ed.), Routledge, London, 2009, 340 pages, ISBN 13: 978-0-415-43437-9 (hbk). Journal of Pragmatics, 42:2604–2608. https://doi.org/10.1016/j.pragma.2010.03.003. Forceville, C. (2002). The conspiracy in The comfort of strangers: Narration in the novel and the film. Language and Literature, 11, 131–147. https://doi.org/10.1177/0963947002011002883. Forceville, C. (2007). Multimodal Transcription and Text Analysis: A Multimedia Toolkit and Coursebook: Anthony Baldry, Paul J. Thibault, Equinox, London/Oakville, 2006, 270 pp., $40/£24.99 (paperback), ISBN 1-904768-07-5. Journal of Pragmatics, 39(6), 1235–1238. https://doi.org/10.1016/j.pragma.2007.02.007. Friend, S. (2017). The real foundation of fictional worlds. Australasian Journal of Philosophy, 95, 29–42. https://doi.org/10.1080/00048402.2016.1149736. Gaudreault, A. (1987). Narration and monstration in the cinema. Journal of Film and Video, 39, 29–36. Gavins, J. (2007). Text world theory: An introduction. Edinburgh, UK: Edinburgh University Press. Genette, G. (1980 [1972]). Narrative discourse: An essay in method. Ithaca, NY: Cornell University Press. Gerrig, R. (1993). Experiencing narrative worlds: On the psychological activities of reading. New Heaven, CT: Yale University Press. Green, M., & Brock, T. (2000). The role of transportation in the persuasiveness of public narratives. Journal of Personality and Social Psychology, 79, 701–721. https://doi. org/10.1037/0022-3514.79.5.701. Green, M., & Brock, T. (2002). In the mind’s eye: Transportation-imagery model of narrative persuasion. In M. C. Green, J. J. Strange, & T. C. Brock (Eds.), Narrative impact: Social and cognitive foundations (pp. 315–341). Mahwah, NJ: Lawrence Erlbaum. Green, S. P. (2001). Lying, misleading, and falsely denying: How moral concepts inform the law of perjury, fraud, and false statements. Hastings Law Journal, 53, 157–212. Grice, H. P. (1989). Studies in the way of words. Cambridge, MA: Harvard University Press. Hamburger, K. (1957). Die Logik der Dichtung. In M. J. Rose (Ed.), The logic of literature, trans. Bloomington, IN: Indiana University Press. Heyd, T. (2011). Unreliability. The pragmatic perspective revisited. Journal of Literary Theory, 5(1), 3–17. https://doi.org/10.1515/jlt.2011.003. Janney, R. (2012). Pragmatics and cinematic discourse. Lodz Papers in Pragmatics, 8, 85–113. https://doi.org/10.1515/lpp-2012-0006. Jewitt, C. (Ed.). (2009). The routledge handbook of multimodal analysis. London, UK: Routledge. Klecker, C. (2013). Mind-tricking narratives: Between classical and art-cinema narration. Poetics Today, 34, 119–146. https://doi.org/10.1215/03335372-1894469. Koch, J. (2011). Unreliable and discordant film narration. Journal of Literary Theory, 5(1), 57–80. https://doi.org/10.1515/jlt.2011.006. Köppe, T., & Kindt, T. (2011). Unreliable narration with a narrator and without. Journal of Literary Theory, 5, 81–94. https://doi.org/10.1515/jlt.2011.007. Kozloff, S. (1988). Invisible storytellers. Berkeley, CA: University of California Press. Kozloff, S. (2000). Overhearing film dialogue. Berkeley, CA: University of California Press. Kress, G. (2010). Multimodality: A social semiotic approach to contemporary communication. London, UK: Routledge. Kupfer, J. (1982). The moral presumption against lying. Review of Metaphysics, 36, 103–126. Laass, E. (2008). Broken taboos, subjective truths: Forms and functions of unreliable narration in contemporary American cinema. A contribution to film narratology. Trier, Germany: Wissenschaftlicher Verlag Trier. Lamarque, P. (1996). Fictional points of view. Ithaca, NY/London, UK: Cornell University Press. Lewis, D. (1978). Truth in fiction. American Philosophical Quarterly, 15(1), 37–46. Linsky, L. (1963). Deception. Inquiry, 6, 157–169. https://doi.org/10.1080/00201746308601371. Lothe, J. (2000). Narrative in fiction and film: An introduction. Oxford, UK: Oxford University Press. Mahon, J. E. (2015). The definition of lying and deception. In E. N. Zalta (Ed.), The Stanford encyclopedia of philosophy (Fall 2015 Edition). http://plato.stanford.edu/archives/fall2015/entries/ lying-definition/.

When Both Utterances and Appearances are Deceptive

251

Mahon, J. E. (2007). A definition of deceiving. International Journal of Applied Philosophy, 21, 181–194. https://doi.org/10.5840/ijap20072124. Margolin, U. (1992). The nature and functioning of fiction: Some recent views. Canadian Review of Comparative Literature, 19, 101–117. Martinec, R. (1998). Cohesion in action. Semiotica, 120(1–2), 161–180. https://doi.org/10.1515/ semi.1998.120.1-2.161. Meibauer, J. (2005). Lying and falsely implicating. Journal of Pragmatics, 37, 1373–1399. https:// doi.org/10.1016/j.pragma.2004.12.007. Meibauer, J. (2014). Lying and the semantics-pragmatics interface. Boston, MA/Berlin, Germany: Mouton de Gruyter. Messerli, T. C. (2017). Participation structure in fictional discourse: Authors, scriptwriters, audiences and characters. In M. A. Locher & A. H. Jucker (Eds.), Pragmatics of fiction. Handbooks of pragmatics (pp. 25–54). Berlin, Germany: Mouton de Gruyter. Metz, C. (1974). Film language: A semiotics of the cinema. New York, NY: Oxford University Press. Morley, D. (1980). The nationwide audience: Structure and decoding. London, UK: British Film Institute. Murphet, J. (2005). Stories and plots. In H. Fulton, R. Huisman, J. Murphet & A. Dunn (Eds.,) Narrative and media (pp. 47–59). New York, NY: Cambridge University Press. Nichols, S., & Stich, S. (2003). Mindreading. An integrated account of pretence, self-awareness, and understanding other minds. Oxford, UK: Oxford University Press. Nünning, A. (2005). Reconceptualizing unreliable narration: Synthesizing cognitive and rhetorical approaches. In J. Phelan & P. J. Rabinowitz (Eds.), A companion to narrative theory (pp. 89–107). Oxford, UK: Blackwell. Nünning, V. (2015). Unreliable narration and trustworthiness. Intermedial and interdisciplinary perspectives. Berlin, Germany: De Gruyter. Olson, G. (2003). Reconsidering unreliability: Fallible and untrustworthy narrators. Narrative, 11, 93–109. Persson, P. (2003). Understanding cinema: A psychological theory of moving imagery. Cambridge, UK: Cambridge University Press. Phelan, J. (2005). Living to tell about it. Ithaca, NY: Cornell UP. Phillips, P. (2000). Understanding film text: Meaning and experience. London, UK: Bloomsbury. Piazza, R., Bednarek, M., & Rossi, F. (2011). Introduction: Analysing telecinematic discourse. In R.  Piazza, M.  Bednarek, & F.  Rossi (Eds.), Telecinematic discourse: Approaches to the language of films and television series (pp. 1–17). Amsterdam, Netherlands/Philadelphia, PA: John Benjamins. Prince, G. (1987). A dictionary of narratology. Lincoln, NE: University of Nebraska Press. Ronen, R. (1994). Possible worlds in literary theory. Cambridge, UK: Cambridge University Press. Ryan, M.-L. (1980). Fiction, non-factuals, and the principle of minimal departure. Poetics, 9, 403–422. Ryan, M.-L. (1981). The pragmatics of personal and impersonal narration. Poetics, 10, 517–539. Ryan, M.-L. (1991). Possible worlds, artificial intelligence and narrative theory. Bloomington, IN: Indiana University Press. Ryan, M.-L. (2001). Narrative as virtual reality: Immersion and interactivity in literature and electronic media. Baltimore, MD: Johns Hopkins University Press. Ryan, M.-L. (2007). Toward a definition of narrative. In D. Herman (Ed.), The Cambridge companion to narrative (pp. 22–35). Cambridge, UK: Cambridge University Press. Sainsbury, M. (2009). Fiction and fictionalism. London, UK: Routledge. Shen, D. (2011). Unreliability. In P.  Hühn et  al. (Eds.), The living handbook of narratology. Hamburg, Germany: Hamburg University Press. Siegler, F. (1966). Lying. American Philosophical Quarterly, 3, 128–136. Simpson, D. (1992). Lying, liars and language. Philosophy and Phenomenological Research, 52, 623–639. https://doi.org/10.2307/2108211. Smith, D. L. (2004). Why we lie: The evolutionary roots of deception and the unconscious mind. New York, NY: St. Martin’s Press.

252

M. Dynel

Sorlin, S. (2016). Language and manipulation in House of Cards: A pragma-stylistic perspective. London, UK: Palgrave Macmillan. Stafford, R. (2007). Understanding audiences and the film industry. London, UK: British Film Institute. Stam, R., Burgoyne, R., & Flitterman-Lewis, S. (1992). New vocabularies in film semiotics: Structuralism, post-structuralism and beyond. London, UK/New York, NY: Routledge. Sternberg, M., & Yacobi, T. (2015). (Un)reliability in narrative discourse: A comprehensive overview. Poetics Today, 36(4), 327–498. https://doi.org/10.1215/03335372-3455114. Stühring, J. (2011). Unreliability, deception, and fictional facts. Journal of Literary Theory, 5, 95–107. https://doi.org/10.1515/jlt.2011.008. Thompson, K. (1977). The duplicitous text: An analysis of Stage Fright. Film Reader, 2, 52–64. Toolan, M. (2001[1988]). Narrative. A critical linguistic introduction. London, UK/New York, NY: Routledge. Truffaut, F. (1983). Hitchcock/Truffaut. New York, NY: Touchstone. Vincent Marrelli, J. (2004). Words in the way of truth. In Truthfulness, deception, lying across cultures and disciplines. Napoli, Italy: Edizione Scientifiche Italiane. Vincent Marrelli, J. (2006). Truthfulness. In J.  Verschueren, J.-O. Östman, J.  Blommaert, & C. Bulcaen (Eds.), Handbook of pragmatics (pp. 1–72). Amsterdam, Netherlands/Philadelphia, PA: John Benjamins. Vrij, A. (2008). Detecting lies and deceit. Pitfalls and opportunities (2nd ed.). Chichester, UK: John Wiley. Walton, Kendall. (1990). Mimesis as make-believe: On the foundations of the representational arts. Cambridge, MA: Harvard University Press. Wharton, D., & Grant, J. (2005). Teaching analysis of film language. London, UK: British Film Institute. Wildfeuer, J. (2014). Film discourse interpretation. Towards a new paradigm for multimodal film analysis. London, UK/New York, NY: Routledge. Zipfel, F. (2011). Unreliable narration and fictional truth. Journal of Literary Theory, 5(1), 109– 130. https://doi.org/10.1515/jlt.2011.009.

Open Access  This chapter is licensed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made. The images or other third party material in this chapter are included in the chapter's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the chapter's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.