Negation and Nonveridicality in the History of Greek (Oxford Studies in Diachronic and Historical Linguistics) 9780198712404, 0198712405

Citation preview

Negation and Nonveridicality in the History of Greek

O X F O R D S T U D I E S I N D I A C H R O N I C A ND H I S T O R I C A L L I N G U I S T I CS  : Adam Ledgeway and Ian Roberts, University of Cambridge  : Cynthia Allen, Australian National University; Ricardo Bermúdez-Otero, University of Manchester; Theresa Biberauer, University of Cambridge; Charlotte Galves, University of Campinas; Geoff Horrocks, University of Cambridge; Paul Kiparsky, Stanford University; Anthony Kroch, University of Pennsylvania; David Lightfoot, Georgetown University; Giuseppe Longobardi, University of York; George Walkden, University of Konstanz; David Willis, University of Cambridge       Transitive Nouns and Adjectives Evidence from Early Indo-Aryan John J. Lowe  Quantitative Historical Linguistics A Corpus Framework Gard B. Jenset and Barbara McGillivray  Gender from Latin to Romance History, Geography, Typology Michele Loporcaro  Clause Structure and Word Order in the History of German Edited by Agnes Jäger, Gisella Ferraresi, and Helmut Weiß  Word Order Change Edited by Ana Maria Martins and Adriana Cardoso  Arabic Historical Dialectology Linguistic and Sociolinguistic Approaches Edited by Clive Holes  Grammaticalization from a Typological Perspective Edited by Heiko Narrog and Bernd Heine  Negation and Nonveridicality in the History of Greek Katerina Chatzopoulou For a complete list of titles published and in preparation for the series, see pp. –.

Negation and Nonveridicality in the History of Greek KATERINA CHATZOPOULOU

1

3

Great Clarendon Street, Oxford,  , United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © Katerina Chatzopoulou  The moral rights of the author have been asserted First Edition published in  Impression:  All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press  Madison Avenue, New York, NY , United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number:  ISBN –––– Printed and bound by CPI Group (UK) Ltd, Croydon,   Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

ΣΩΚΡΑΤΗΣ ὑμεῖς μέντοι, ἂν ἐμοὶ πείθησθε, σμικρὸν φροντίσαντες Σωκράτους, τῆς δὲ ἀληθείας πολὺ μᾶλλον, ἐὰν μέν τι ὑμῖν δοκῶ ἀληθὲς λέγειν, συνομολογήσατε, εἰ δὲ μή, παντὶ λόγῳ ἀντιτείνετε, εὐλαβούμενοι ὅπως μὴ ἐγὼ ὑπὸ προθυμίας ἅμα ἐμαυτόν τε καὶ ὑμᾶς ἐξαπατήσας, ὥσπερ μέλιττα τὸ κέντρον ἐγκαταλιπὼν οἰχήσομαι. SOCRATES But you, if you do as I ask, will give little thought to Socrates and much more to the truth; and if you, think what I say is true, agree to it, and if not, oppose me with every argument you can muster, that I may not in my eagerness deceive myself and you alike and go away like a bee, leaving my sting behind. Plato, Phaedo c

Contents Series preface Preface List of abbreviations List of figures and tables . Introduction . . . .

The two-negator system of Greek Theoretical backgrounds: semantics, syntax, and agreement Selection of texts and methodology Periodization .. Classical Greek (fifth to fourth centuries ): the Attic dialect .. Hellenistic and Roman times (third century  to fourth century ): Atticism and the Koine .. Early Medieval Greek (fifth to tenth centuries ) .. Late Medieval Greek (eleventh to fifteenth centuries ) .. Early Modern Greek (sixteenth to nineteenth centuries ) . Previous work .. The Classical scholars .. More recent contributions within linguistics . Markedness, the Nonveridicality projection, and pragmatic grounding . Roadmap . The (non)veridicality theory of polarity and negator selection . Negative polarity: precursors and (non)veridicality (Giannakidou ) . Polarity phenomena in Standard Modern Greek .. Negative polarity items .. The focus particle ούτε /útε/: an instantiation of a negative polarity EVEN .. (Non)veridicality and mood choice .. (Non)veridicality and negator choice . Crosslinguistic analogies . Beginning the journey: the oldest language relatives and the first attestations .. Indo-European origins and affinities

xi xiii xv xvii                         

viii

Contents

.. Homeric Greek negators and traces of the Proto-Indo-European negatives .. The homotics νή /nε:/ and μά /ma/: remnants of (Proto-)Indo-European negators? . Summary . Negation, mood, and (non)veridicality in Classical Greek (fifth to fourth centuries ) . The Attic dialect .. Morphology of verbs and verbals .. Attic Greek word order . The syntactic status of the Attic Greek negators: phrasal . Negation and morphological mood: no consistent relation . Negator selection and (non)veridicality .. The distribution of N oὐ(κ) and N μή in terms of (non)veridicality . The nonnegative uses of N μή /mε:/ .. N μή /mε:/ as a question particle .. N μή /mε:/ as complementizer .. N μή /mε:/ of pathos: N as an attitudinal . The morphologically negative indefinites of Attic Greek: N-words and N-words .. Attic Greek Negative Concord .. The syntactic analysis of Attic Greek Negative Concord .. N-thing and N-thing as negative adverbs . Summary . Developments in Hellenistic-Roman times and the Nonveridicality projection (third century  to fourth century ) . The Greek of the Hellenistic-Roman era . Analyticity linked to phonological change .. Language developments: pro-analyticity .. Phonological changes in the vowel system and repercussions . Changes in the mood and complementation system: the broadening of ἵνα /hina/ .. Loss of consecutio modorum .. The broadening of ἵνα /hina/: reanalysis driven by homophony .. The INA broadening and generative theories of syntactic change .. The Nonveridicality projection (NVP) .. Precursors and differentiation: Cinque (), Culicover (), Laka ()

                              

Contents

. .

. .

.. Feature economy: what happens to the finite verb after the reanalysis of INA Change in word order: from OV to VO Negator distribution in Koine Greek .. The distribution of N οὐ(κ) /u(k)/ in Koine Greek .. The distribution of N μή /mi/ in Koine Greek .. Statistics of N and N attestations and comparison with Attic Greek The negative indefinites of Koine Greek: N-thing and N-thing Summary

. Late Medieval and beyond: the renewal of N . Negation in Late Medieval Greek: basic developments .. How many negators? .. Oὐδέν /udhén/ and μηδέν /midhén/: diagnosing plain propositional negation . Change of syntactic status: from phrases to heads .. Strict preverbal position .. Unattested as DP internal or constituent/metalinguistic negation .. Unavailability of True Negative Imperatives: a parameter reset .. The van Gelderen () Head Preference Principle . The ban of N from the conditional protasis .. N fossilized in exceptive conditionals .. An explanation: upward reanalysis of N on Cinque’s () hierarchy .. Comparison to N’s persistency in purpose clauses: possible support for upward micromovement . Evidence from Early Modern Greek texts: the end of the journey . Summary . Renewal and stability: one full Jespersen’s Cycle and one persistent polarity item . Traditional accounts of Jespersen’s Cycle . Atypical negative cycles in Greek and beyond .. The atypicality of Greek N and N developments: no doubling stage .. Other atypical Jespersen’s Cycle manifestations . Jespersens unite: a new definition for Jespersen’s Cycle .. Jespersen bleaching as scalar endpoint lexicalization .. The representation of the Greek Jespersen’s Cycle and word-order considerations

ix

                              

x

Contents . The diachronically stable uses of N .. Negative directives .. The diachronic nonnegative uses of N . Summary

   

Conclusion



Appendix of texts References Index

  

Series preface Modern diachronic linguistics has important contacts with other subdisciplines, notably first-language acquisition, learnability theory, computational linguistics, sociolinguistics, and the traditional philological study of texts. It is now recognized in the wider field that diachronic linguistics can make a novel contribution to linguistic theory, to historical linguistics, and arguably to cognitive science more widely. This series provides a forum for work in both diachronic and historical linguistics, including work on change in grammar, sound, and meaning within and across languages; synchronic studies of languages in the past; and descriptive histories of one or more languages. It is intended to reflect and encourage the links between these subjects and fields such as those mentioned above. The goal of the series is to publish high-quality monographs and collections of papers in diachronic linguistics generally, i.e. studies focusing on change in linguistic structure, and/or change in grammars, which are also intended to make a contribution to linguistic theory, by developing and adopting a current theoretical model, by raising wider questions concerning the nature of language change, or by developing theoretical connections with other areas of linguistics and cognitive science as listed above. There is no bias towards a particular language or language family, or towards a particular theoretical framework; work in all theoretical frameworks, and work based on the descriptive tradition of language typology, as well as quantitatively based work using theoretical ideas, also feature in the series. Adam Ledgeway and Ian Roberts University of Cambridge

Preface This book is an updated version of my doctoral dissertation, which was submitted to the Department of Linguistics, University of Chicago, in August . A few changes were made in structure and content, in order to include more recent references and research insights. I would like to thank my dissertation advisor Professor Anastasia Giannakidou for her guidance throughout this project and also for her encouragement, stable yet discreet, to finish the present book. I am grateful to her above all else for her research and scientific insight, particularly regarding the theory of (non) veridicality, without which my contribution would not have been possible. I would furthermore like to thank for the same reasons professors Brian Joseph, Jason Merchant, Elly van Gelderen, and Johan van der Auwera, as it was a combination of their lifelong study and their devotion to the field of Linguistics that enabled me to understand and make some sense of the particular data examined in the present study. I am furthermore grateful to Jo Willmot and Hedde Zeijlstra for key suggestions and detailed advice regarding the restructuring of some parts of the dissertation, as well as providing more analytical discussion on some of the axioms of the analyses of the data. All deviations from a desirable outcome in this work are my own. I have tried to include most relevant studies that were published after  and hope to have done justice to them and the proposed new analyses. Yet, although one could claim that the history of scientific and academic writing is a history of misunderstanding, I think that, nevertheless, we have done well so far. As a linguist and as a researcher I am proud of all the advances in Linguistics and all related fields. A particularly rewarding outcome of this endeavor for me was to witness the attempts in quite diverse fields, starting with the classical scholars of the eighteenth and nineteenth centuries and continuing up to the generative grammar theorists, typologists, and field linguists, to account for the same or similar language phenomena, in the present case the history of negation, using different terminology. It is the quest for regularity and the desire to make sense that unites all fields synchronically and diachronically and may reveal something like a team spirit. Whether it is the Classical scholars, more knowledgeable than anyone might credit, and profoundly insightful with a bare minimum of crosslinguistic data, or the brilliant, inventive, and determined semanticists and syntacticians, or the very respectable, devoted, and brave field linguists and typologists, it is always us against the data, and we will win.

List of abbreviations P

second position clitic (Wackernagel clitic)



accusative



aorist

CP

complementizer phrase



dative

DP

determiner phrase



enclitic (postpositive)



feminine



future



gerundive (verbal adjective with deontic meaning)



genitive



imperative



imperfective



indicative



infinitive



imperfective non-past



masculine



mediopassive voice



neuter



nominative



optative



participle



plural



perfective non-past



perfective past



present



proclitic (prepositive)

QP

question particle



singular



subjunctive

TNI

true negative imperative

List of figures and tables Figures .

The syntactic status of the Modern Greek negators



.

The Modern Greek N in complementizer position



.

Geographical distribution of languages with a distinct negator for prohibition (types  and )



.

The syntactic status of the negators in Standard French



.

The phrasal status of N oὐ(κ) /u:(k)/ and N μή /mέ:/



.

Greenberg’s realis–irrealis continuum



.

N and N DP internal in Attic Greek



.

The μή οὐ (N N) construction



.

The οὐ μή construction (N N)



.

Distribution of N and N in Attic Greek ( tokens)



.

Syntactic distribution of n-thing in NC and non-NC structures



.

The constellation of Attic N μή /mε:/



.

The vowel system of Attic Greek



.

Developments in the vowel system from Attic to Koine



.

The vowel system of Late Koine Greek



.

The ἵνα /hina/ complementizer in Attic Greek



.

The nonveridical space and the broadening of ἵνα /hina/



.

The ἵνα /hina/+Subjunctive mood reanalysis from Classical Greek to Koine



.

The Modern Greek Subjunctive: a product of reanalysis



.

Cross-clausal grammaticalization: INA from Attic to Koine



.

N-thing reanalysis from Classical Greek to Koine Greek



. The distribution of N and N in Koine Greek



. Comparison of N and N distribution between Koine Greek and Attic Greek



. Rates of Negative Concord in Koine Greek (Negative Spreading included)



. Rates of Negative Concord in Attic Greek (Negative Spreading included)



List of figures and tables

xviii .

Negator distribution and diversity in Late Medieval Greek



.

The functions of oὐδέν /udhén/ in three major stages of spoken Greek



.

The phrasal status of N and N in Classical and Koine Greek



.

The head status of negators in Late Medieval Greek



.

Negator distribution in the conditional protasis in Attic, Koine, and Late Medieval



N and N percentages in eighteenth- to nineteenth-century folk songs ( tokens)



. .

Negator distribution in Early Modern Greek (folk songs)



.

Upward lexical micromovement of the Greek N indefinite



.

Upward lexical micromovement of the Greek N indefinite



.

Syntactic distribution of N- and N-thing in Attic Greek (Stage I)



.

Syntactic distribution of N- and N-thing in Koine Greek



.

N-thing reanalysis from Attic Greek to Koine and Medieval Greek



.

Leftward lexical micromovement in the French Jespersen’s Cycle



.

Leftward lexical micromovement in the Greek Jespersen’s Cycle



.

The distribution of N μη /mi/ in Attic Greek, in Koine, and in Late Medieval Greek



Tables .

The two-negator contrast from Proto-Indo-European to Standard Modern Greek



.

Periodization of the Greek language



.

Prototypical nonveridical environments



.

Tense and aspect combinations in Standard Modern Greek



.

The distribution of N and N in Standard Modern Greek



.

The seven negative markers of Romeyka



.

Negators in some of the oldest Indo-European languages



.

Negative particle distribution in Hittite



.

Negator distribution in Vedic Sanskrit



.

Environments of negative particles in Old Persian



.

Negative particles in Classical Armenian



. Environments of negative particles in Albanian



. Environments of negative particles in Homeric Greek



.

Paradigm of active third person singular of λύω (/ly:o:/, ‘set free’) in Attic Greek



List of figures and tables

xix

. The mood sequence of Attic Greek (consecutio modorum)



. Prototypical environments of the Attic Greek moods



. The distribution of N oὐ(κ) /u:(k)/ and N μή /mέ:/ in Attic Greek



. The n-words of Attic Greek



. Zeijlstra’s typology of (non)inherent negativity of negative markers and n-words



. Homophony in the mood system of Koine Greek (active voice)



. Homophony in the mood system of Koine Greek (mediopassive voice)



. The loss of consecutio modorum from Attic Greek to Koine



. The two-negator contrast from Proto-Indo-European to Standard Modern Greek



1 Introduction . The two-negator system of Greek The primary goal of this study is to provide a detailed investigation of the expression of sentential negation in the history of the Greek language. Greek preserves a contrast between two complementary negators in all its attested history, from Homeric Greek to Standard Modern. By tracking down the environments in which each of the two negators is licensed in a sequence of twenty-five centuries, the regularity behind negator choice can be identified: the second negator, μη /mε:/ in Classical Greek and /mi/ after Koine Greek, is a polarity item in all stages of the Greek language; an item licensed by nonveridicality, in the sense of Giannakidou ( et seq.). The history of the Greek negator system also bears witness to an interesting asymmetry regarding the individual courses of the complementary negators through time (table .). The first negator is renewed, in that the functions of the Attic Greek u:(k[h]) are in Modern Greek taken over by the etymologically unrelated item dhen. The second negator, however, is not eventually renewed, but remains stable and can be traced back to the reconstructed Proto-Indo-European negative *meH. The second negator of Classical Greek, the polarity item μή /me:/, undergoes only the phonological alterations pervasive in each synchronic linguistic stage (me: > me > mi > mi(n)) and basically remains stable in form and, from its exclusive presence in nonveridical contexts, most prominently in its uses (i) as the negator of prohibitions,

TABLE . The two-negator contrast from Proto-Indo-European to Standard Modern Greek N Proto-Indo-European

*ne

N vs

*meH

........................................................................................................................................

Homeric Greek Classical Greek Koine Late Medieval Modern Greek

[h]

u:(k ) u:(k[h]) u(k) u(k) (and udhén) dhe(n)

vs vs vs vs vs

me: me: mi mi (and midhén) mi(n)

Negation and Nonveridicality in the History of Greek. First edition. Katerina Chatzopoulou. © Katerina Chatzopoulou . First published  by Oxford University Press.



Introduction

and (ii) as a nonnegative complementizer, as a sort of expletive or paratactic (strictly, redundant) negation (in the sense of van der Wouden , ). Through qualitative and quantitative data from representative texts from three major stages of vernacular Greek (Attic Greek, Koine Greek, Late Medieval Greek), this study depicts the asymmetry in the diachronic development of the Greek negator system: the replacement of the first negator, N, and the preservation of the second one, N (Classical Greek me:, Modern Greek mi(n)). The explanation provided relates to the particulars of the uses of N, specifically the inertial forces drawn by the nonnegative uses of N, which, being nonnegative, did not experience the renewal pressures predicted by the Jespersen’s cycle: its use in introducing (i) yes/no questions, and (ii) verba timendi complements. The exact developments of N and N, however, do not properly fall under the Jespersen’s Cycle phenomenon (Jespersen , ; term coined in Dahl ), as already observed in Willmott , although studies of grammaticalization and language typology have included Greek in the discussion of Jespersen’s Cycle (cf. Roberts and Roussou , Roussou ). I propose that the history of the Greek negator system provides motivation for a refinement of the traditional understanding of Jespersen’s Cycle, in a fashion that abstracts away from the morphosyntactic and phonological particulars of the phenomenon and explicitly places its regularities in the semantics. This is an intuition that is found in the Jespersen’s Cycle literature (Horn , van Kemenade , Roberts and Roussou , Kiparsky and Condoravdi , van Gelderen , de Cuypere , van der Auwera , a). Our task is to state it explicitly, accommodating not only Greek, but a number of other languages that deviate in different ways from the traditional description of Jespersen’s Cycle (cf. de Swart : ). The viewpoint I introduce for Jespersen’s Cycle is one that treats it as a phenomenon which targets intensified predicate negation and with time transforms it into propositional negation. In the history of Greek, N participates in such a development: the Classical and Hellenistic Greek indefinite udhén (‘nothing’, morphologically: ‘not.even.one’, see also Roussou ) leads to the Late Medieval negator of propositions and by pervasive phonological alterations (loss of initial unstressed vowels, see Horrocks /) develops into the Standard Modern Greek N dhen. The surviving N experienced a similar stage, in that the former N indefinite μηδέν /midhén/ (‘nothing’, morphologically: ‘not.even.one’) had bleached to plain sentential negation by the Late Medieval stage. This form, however, did not persist to the Standard Modern Greek stage. Overall, this study aims to offer more insight regarding the regularities that pertain to sentential negation both synchronically and diachronically with data primarily from the history of Greek, but the claims made are supported by a significant amount of crosslinguistic evidence. Synchronically, because it is shown that sentential negation itself, although a nonveridical operator, can be a polarity item licensed in

Theoretical backgrounds



nonveridical semantic contexts, as predicted by the (non)veridicality theory, which places no categorial restrictions on the elements that exhibit polarity behavior (Giannakidou : –, –). Diachronically, because the developments in the history of Greek are in agreement with current theories of diachronic chage, in that major tendencies for up-the-tree movement (described also as loss of covert movement in Roberts and Roussou , Roberts ), and anticipated syntactic status shifts (from Specifier to Head, see van Gelderen , ) are found.

. Theoretical backgrounds: semantics, syntax, and agreement This study is located at the interface of historical semantics and historical syntax, and comes to assess how the theoretical claims and tools developed based on the synchronic analyses of negation and polarity bear on the description and explanation of the relevant evidence from the diachrony of Greek. Owing to the multidimensionality of the phenomena examined, as well as the still nascent stage of our understanding of historical syntactic and semantic change, distinct yet not contradictory theories are employed in order to state the proposed analyses and accommodate the evidence. Central is the (non)veridicality theory of polarity developed in Giannakidou ( et seq.), which accounts for the distribution of the Greek negators in all the attested history of the language: the second negator, N, is licensed only in nonveridical environments, such as modal, intensional, generic, and nonassertive environments in general (e.g. imperatives, optatives, questions, and the protases of conditionals, see Chatzopoulou and Giannakidou ). There is variation from language stage to language stage regarding which of these environments license the Greek N. The environments are not always identical, but they are always nonveridical. Crucially, no pragmatic manipulations can synchronically affect the licensing of N. It is purely a grammatical phenomenon, as predicted by the (non)veridicality theory: this theory relies on entailments available and not just discourse relations. Although the ultimate goal of this study is the unveiling of the semantic regularities that relate to the expression of sentential negation (both in the case of negator choice and in the semantic definition proposed for Jespersen’s Cycle), these regularities have syntactic repercussions as well. For the syntactic analyses, the representational frameworks developed within generative grammar have been employed (Chomsky ), as an enriched system for mapping sentential meaning synchronically. It is through the generative perspective regarding the abstractness and the hierarchy of syntactic structures that the developments in the Greek negator system are accounted for. The Agree model since Chomsky () has been of particular use regarding not only the syntactic treatment of negative concord structures in terms of agreement in the sense of Zeijlstra (, ), but also the syntactic representation of nonveridical marking with a goal and a probe, assuming a covert or overt (in the case of Koine Greek) nonveridical head that carries [iNonVer] and agrees with



Introduction

elements such as nonveridical moods, evidentials, or in this study in particular with the Greek N and N-words that are taken to carry a [uNonVer] feature. There can be different paths in the analysis and compositional treatment of multiple exponents in languages regarding e.g. locality considerations (see among others Haegeman and Lohndal  for a binary agreement of negative concord structures in West Flemish), as well as regarding the challenge of determining which element is the goal and which is the probe. The present study is rather conservative in this sense, as flexibility of word order in Classical Greek and Koine does not add anything new to this discussion. However, our findings are in support of the historical syntactic models of Roberts and Roussou (), and van Gelderen (), as the descriptions of syntactic change as ‘upward reanalysis’ and the notion of ‘Late Merge’ readily account for the transformation of predicate negation to propositional. The Head Preference Principle of van Gelderen () also explains the change in syntactic status of the Greek negators from phrases in Classical Greek to heads in Late Medieval and Modern Greek, as well as the attestations of N in nonnegative complementizer positions (in its use as a particle introducing yes/no questions and as a complementizer selected by timendi predicates), where it has stayed inert, however, for more than twenty-five centuries. The pretheoretical representations of language change that are included in this study1 capture the notions of up-the-tree movement and reflect the independence between the lexical/morphological and the syntactic/semantic components, providing a better visualization of the diachronic developments as instances of upward (and in the case of Greek leftward) lexical micromovement, assuming the principle of syntactic inertia of Keenan () and Longobardi (), which has been posited for diachronic change. However, as will be shown, what is actually inert is the semantics, since communicational needs among humans have not changed, at least in historical times. In particular the preservation of nonveridical marking appears to be a language priority (see Chatzopoulou ), which if jeopardized (as was the case during the Koine Greek stage, owing to the homophony in the mood system), other mechanisms will be employed in order for it to resurface. Regarding the broad notion of grammaticalization, our understanding takes into consideration the criticism it has received (Newmeyer , , Joseph a, , , Campbell , Campbell and Janda , Janda , Norde , Joseph and Janda ) as a putative theory or process distinct from general mechanisms of language change. In this study, grammaticalization is viewed neither as a process, nor as a theory, but as a phenomenon: as a tendency in endogenous

1 Inspired by the philosophy of Autolexical Grammar (Sadock , Sadock & Schiller , Sadock ) regarding the notion of mismatch across components as something which is diachronically verified.

Theoretical backgrounds



language change that is usually described as loss of semantic features, based on frequently attested crosslinguistic patterns. One such pattern is Jespersen’s Cycle. Thus, although several conspiratorial forces from different linguistic levels can be involved, and the ‘processes’ of grammaticalization can indeed be accounted for through other mechanisms of language change (e.g., reanalysis, phonological erosion), in this study grammaticalization is viewed as a phenomenon, in that it happens. A phenomenon whose end result (or side effect) is the upward reanalysis of lexical elements in the syntactic domain (Roberts and Roussou ). This may not be the heart or the most defining property of grammaticalization, but it is a secure diagnostic, with robust crosslinguistic support. In the case of propositional negation the endogenous change is represented here as upward lexical micromovement, where instances of predicate negation are reanalyzed as propositional, and therefore higher in the hierarchical structure. The micro- part of our ‘micromovement’ terminology refers to fine shifts, especially on the Cinque expanded CP approach, that account for changes in the distribution of the Greek N, and captures furthermore the gradualness of the changes described, which is in agreement with outlooks on grammaticalization and language change that highlight the gradual nature of the phenomenon (Lichtenberk , Haspelmath , Hopper and Traugott , Lehmann , Roberts , Traugott and Trousdale ; but see Lightfoot  for the notion of ‘catastrophic’ change); but most importantly it is in agreement with our historical data.2 This diachronic movement correlates also with changes in word order from SOV, which is claimed to be the tendency of unmarked word order in Classical Greek (cf. Ebeling , Devine and Stephens , Taylor , Dik , ) to VSO in Hellenistic Greek and later stages (Haug , Horrocks , Deligianni a), right up to Standard Modern Greek (Philippaki-Warburton , , Tsimpli , Alexiadou and Anagnostopoulou ), where we see that the verb itself has been elevated in a direction which observes vertical structure: the verb has moved to a position higher in the structural domain. Greek is a left-branching language. The prediction of Roberts and Roussou (), as well as that of van Gelderen () from a similar perspective, is that in right-branching languages grammaticalization processes will follow the opposite direction in their linear representation, as instances of rightward lexical micromovement, according to the terms proposed here.

2 See in the same spirit but from a syntactic point of view the dissertation of De Clercq (), building on Starke (), regarding the nanosyntax of negation crosslinguistically, with the identification of subatomic features into which negation can be decomposed and then a detailed examination of the role of syncretism among features that gives rise to typological diversity. Such micro- and nano- terminology reflects the quest for primitives and indivisible syntactic or semantic features, which so far seems to be a promising direction for capturing crosslinguistic regularities and providing plausible accounts for empirical evidence.



Introduction

. Selection of texts and methodology In this section I present the stages of the Greek language examined, along with the texts and resources used for each stage. A thorough overview of the entire history of Greek is given by Horrocks (/) in his book Greek: A History of the Language and its Speakers, earlier in the historical grammar of Jannaris () and more recently by Christidis (). The purpose of my investigation is to study the expression of negation in the ancestral lineage of Standard Modern Greek. The criteria used for the selection of the stages of the language and texts included in this study are three: (i) The selected stage must be included in the ancestral lineage of Standard Modern Greek with some degree of certainty. (ii) The stage must be documented in a sufficient number of texts that we can draw secure conclusions. (iii) The selected texts must represent or be close to the vernacular of the time. Therefore, although surviving texts in Greek can be found as early the Linear B syllabary (fourteenth century ) from the Mycenaean civilization—where N is in fact attested as an o-u- verbal prefix (see Rijksbaron )—these texts are not included in this study, as it cannot be claimed with any certainty that they hold an ancestral relation to Attic Greek, which is among the precursors of Standard Modern Greek. Homeric Greek (eighth century ) is discussed to a narrower extent (although written in an artificial language particular to epic songs), because Homeric Greek was intelligible to the speakers of the Classic era and most importantly it contains the first occurrences of the two negative markers of the Greek language, N u:(k[h]) and N me:. Furthermore, Homeric Greek negators have been extensively studied most recently in Willmott , which provides a taxonomy of their uses that is in agreement with what is proposed here. The above criteria regarding text selection exclude dialectal variation in Ancient Greece, (i.e. dialects other than Attic) and the vast atticizing literature of the Hellenistic and medieval periods, as well as postmedieval, modern Greek dialects such as Pontic and Cypriot (with the exception of Romeyka Greek, discussed briefly in section . for its typological relevance and uniqueness). Regarding the methodology, lexical statistics were calculated on samples of at least , instances of negation per synchronic linguistic stage: from Classical Greek, Koine Greek, and Late Medieval. The goal was to examine a potential correlation between the frequency of each negator in each use and its diachronic stability. No such correlation was found, in that there appears to be no relation between diachronic stability and the frequency of each negator or each individual use of N and N. This finding may be surprising for theories of language change that rely on frequency, but it is consonant with Chomsky’s () observation that

Periodization



‘probabilistic considerations have nothing to do with grammar’ (Chomsky : , see also van Gelderen : –). They are merely the reflection of preservation or change, not the driving force behind either.

. Periodization The periodization of the Greek language that I follow here (table .) is that of Markopoulos (), who conducted a similar diachronic study regarding the expression of the future tense in Greek, much broader in terms of examination of textual evidence. Markopoulos (), however, ends at the Late Medieval stage, while here sentential negation in Early and Standard Modern Greek is examined as well. By contrast, negation in Early Medieval is mostly excluded, though I do present examples from this stage for the sake of completeness. The evidence we have for the Early Medieval stage is neither securely chronologized, nor close to the vernacular (see also Willmott ) and is thus ruled out according to the principles of this investigation. Some basic facts regarding each chronological stage and the texts from which qualitative and quantitative data are drawn are presented in the following sections. .. Classical Greek (fifth to fourth centuries ): the Attic dialect Attic Greek was the most influential of the Ancient Greek dialects. It was spoken mainly in the region of Athens during the fifth and fourth centuries  (and for some texts written in the third century ) and it was intelligible to speakers of other dialects. It is the language in which the works of Demosthenes, Isocrates, Lysias, Plato, Aristoteles, Aristophanes, Euripides, Sophocles, and other well-known authors were written. The surviving texts of all these writers and of many of their contemporaries are included in the Thesaurus Linguae Graecae (TLG) database, a digitized database of literary texts written in Greek which covers the phases of the Greek language until the fifteenth century. This is the database that is used for this phase of Greek.

TABLE . Periodization of the Greek language Language stage

Time span

Classical Greek: The Attic dialect Hellenistic Greek: Koine Early Medieval Greek (only some examples) Late Medieval Greek Post-Medieval/Early Modern Greek

Fifth to fourth centuries  Third century –fourth century  Fifth to tenth centuries  Eleventh to fifteenth centuries  Sixteenth to nineteenth centuries 



Introduction

.. Hellenistic and Roman times (third century  to fourth century ): Atticism and the Koine Although the TLG database contains surviving linguistic material with no substantial gap until the fifteenth century  and therefore covers the Hellenistic-Roman stage as well, this material is treated with more caution, owing to the rise of Atticism during the first century . ‘Atticism’ is a term applied to the intellectual movement which considered the classical Attic language of greater value than the koine, the lingua franca of the time that depicted, or was at least very close to, the vernacular. Atticism encouraged the mimicking of Classical Greek, especially in writing, resulting in a diglossia between the learned writers and the writers whose texts were closer to the vernacular. This diglossia in various forms was to follow the Greek language into the twentieth century. For these reasons, non-atticizing texts were selected for this study, such as Strabo’s Geographica, the Greek New Testament, and Epictetus’ Dissertationes ad Arriano, along with some examples from papyri. Regarding the language of the Greek New Testament, I consider the discussion on the presence of Aramaisms and their extent in the Greek text (Winer /, Viteau , Cremer , Torrey ) resolved toward the conclusion that ‘the Greek of the New Testament is basically the vernacular Greek of the Hellenistic world’ in the phrasing of Voelz (: ), a conclusion supported as early as Deissmann (), through a comparison of the Greek New Testament with the language of the papyri of the time. Although the Greek New Testament is a translation from Hebrew to some extent, it depicts a non-atticizing variety that can be identified with the spoken language and it is included in the ancestral lineage of Standard Modern Greek also for its impact on posterity, being in addition the language of the Greek Orthodox church. .. Early Medieval Greek (fifth to tenth centuries ) This period of the Greek language is the most scarcely documented in comparison to the stages that precede and follow. A number of political, religious, and social factors contributed to this situation, so that the rarity of texts, especially from the seventh until the ninth centuries led to the labeling of this period the “Dark Ages” of Byzantium (see Markopoulos : ). From this stage only a small sample of examples are included, for the sake of completeness, from the Chronicle of Johannes Malalas (sixth century) and the Chronicon Paschalae (seventh century), while the Suda dictionary (tenth century) is used for metalinguistic evidence in chapter . .. Late Medieval Greek (eleventh to fifteenth centuries ) Although this too was a period of cultural and political turbulence, marked in particular with the two captures of Constantinople, one by the Franks in , one by the Turks in , there is a variety of textual evidence of the vernacular of the

Previous work



time. This includes the beginning of the folk-song tradition, with the epic tale of Digenis Akritis (twelfth century) in the Escorial manuscript, the Ptohoprodromika (twelfth century) satire and the Late Byzantine romances, written in a less literary style, such as Imperios and Margarona, Velthandros and Hrysantza, Livistros and Rodamne (fourteenth century), along with the Chronicle of Morea (fourteenth century), a long poem on the conquest of the Peloponnese by the Franks. The Chronicle of Morea, although written most likely by a nonnative speaker of Greek, is highly representative of the lower register of Greek at the time (see Markopoulos : ) and of comparable usefulness to the rest of the texts of this period for the purposes of the present study. .. Early Modern Greek (sixteenth to nineteenth centuries ) This period of the history of Greek covers the years under Ottoman rule, the years of the Greek revolution, and the first decades of the independent modern state. Although the diglossia continues and many of the sources represent an educated written form of the language, there is one important source from this period that is of major importance for this study: the folk songs and Kleftic ballads preserved through oral tradition and published as early as Fauriel (–), followed by Politis (). Written in the vernacular of the time and in verse, which facilitates the preservation of earlier forms, folk songs are important for the study and development of ‘demotic’ Greek (Horrocks : ), which is the immediate ancestor of Standard Modern Greek.

. Previous work There are a number of studies that stand in close approximation in one way or another to the topic of this dissertation: the expression of negation in the Greek language. In this section an overview of these studies is presented with emphasis on their positive impact and affinity with the outcome of this project. The aspects in which this study departs from the outlooks of previous studies are discussed in detail in the appropriate sections throughout the dissertation and will be evident by the conclusion of chapter . The valuable yet segmented insight that previous work on Greek negation provides makes the need for a unified and novel approach more prominent and, it is to be hoped, possible. .. The Classical scholars The task of accounting for negator choice in Greek has in part been preceded by a series of erudite classical scholars, starting from the late eighteenth century, throughout the nineteenth and twentieth, either in the writing of Greek grammars (Kühner , Crosby , Goodwin , Jannaris , Kühner and Gerth –, Smyth , Tzartzanos /, Schwyzer and Debrunner /, Rijksbaron /)



Introduction

or in the form of more focused studies (Hoogeveen /, Hermann , Anton , Franke , Hartung , Kühner , , Bäumlein , , Madvi , Aken , Delboeuf , Gildersleeve , Cook-Wilson –, Gallaway , Postgate , Stahl , Laird , Humbert , Koppers , Moorhouse , , Basset , Gerö ). Discussing the negative particles mainly in Ancient Greek (Homeric Greek, Classical, or Hellenistic Greek), they have provided detailed and accurate taxonomies of the functions of the Greek negators and they made key observations and insightful claims in their attempts to identify the regularity in the distribution of the Greek negative particles. At the heart of these claims lies the conviction that the regulating factors in the distribution of N and N are semantic.3 One dominant assumption in the works of the earliest scholars is that the use of N and N follows from the distinction ‘objective’ versus, ‘subjective’ (Hermann , Bäumlein , , Kühner and Gerth , Anton , Kühner ). Some of their exact phrasings are worth mentioning. Hoogeveen () claims that the primary property of N is the cancelation of actus/rei certitudinem, ‘the certainty of actions or things’ and is in this sense partial, while N has no such effect and negates a totality: sed οὐ [negat] totum, et μὴ partem (Hoogeveen ). Hermann () distinguishes the uses of N and N through the contrast between res ipsa and cogitatio rei, the contrast between ‘things themselves’ and ‘thinking about them’: οὐ negat rem ipsam; μὴ cogitationem rei (Hermann , see Viger : ). Others are skeptical of this division (Hartung , Prüfer , Gayler ) and underline the regulating role of other semantic factors, such as those of conditionality, purpose, and causality, all of which are pertinent to the uses of N. Gayler (), for example, describes the environments of N through the concepts of ‘transmission of a thought’ and ‘will’, while N has a simple canceling result. He furthermore observes the notion of emphasis in some of the uses of N, what he calls confirmatio logica, which he uses to refer to volitionality, a notion that correlates with directive force. Although it is less than often that these scholars appear to agree with one another in their formulations, in spirit they are in agreement with what is proposed here and their taxonomies are supported by our qualitative and quantitative findings. Our endeavor is not merely a change of terminology, but the application of current formal tools in order to provide a unified account of negator selection in all environments and all language stages that stand in the ancestral lineage of Standard Modern Greek, ultimately in a way that may have crosslinguistic value in sheding more light on the typological possibilities and the universal tendencies that pertain to the expression of negation. 3

Although for some environments, like the attestations of the negatives in nominal phrases, there have been attempts to reduce their distribution to phonological factors, e.g. the avoidance of hiatus, cf. Jannaris (), Postgate ().

Previous work



.. More recent contributions within linguistics A number of more recent linguistic studies are relevant in this regard. PhilippakiWarburton and Spyropoulos () have argued that the regulating factor in negator choice is explained through the contrast between epistemic and deontic modality in all the stages of Greek. Their claim is that N is linked to epistemic modality and N to deontic. However, this distinction only partially accounts for the data, as has been shown in detail in Willmott (). Among other inadequacies, the presence of N in the conditional protasis until Early Medieval Greek can hardly be seen as an instance of a deontic use. Nevertheless Philippaki-Warburton and Spyropoulos’ () paper is a significant attempt to identify a single semantic parameter that can account for the distribution of negative particles in all of Greek. Another contribution to this topic that is very close in its intuitions to our central claims is Willmott (), where the focus is on Homeric Greek, followed by Willmott (, ), in which a broader overview of the negators is undertaken with data from different stages of Greek and to some extent different dialects (Tsakonian, Pontic). One crucial observation she makes is of the independence of negator choice from mood choice, already present in Homeric Greek and throughout Hellenistic, which is, however, gradually obscured in later stages of the language and has resulted in an epiphenomenal partition of the uses of the two negators: N as the negator of the indicative, and N of the nonindicative (Joseph and PhilippakiWarburton , Giannakidou ). Another important contribution of Willmott’s work is the introduction of the general notion of irrealis in the discussion of Greek negator choice and its correlation with N. The theory of (non)veridicality (Giannakidou ) that is adopted in this study is a detailed formal account of the pretheoretical notion of irrealis, in a way that singles out the unifying property in various elements with limited distribution crosslinguistically, described as negative polarity items. One such item is the Greek N in all the history of the language. Regarding the diachronic development of the Greek N (Classical Greek, u:(k[h]) > Medieval Greek, udhén > Modern Greek, dhen), Roberts and Roussou () include N in their discussion of Jespersen’s Cycle, as one more case in which the function of clausal negation was assumed at some point by a former indefinite, which for Greek was the negative quantifier udhén (‘nothing’). Its change of function is accommodated within their historical syntax model through the notion of ‘upward reanalysis’ viewed as a diachronic tendency: the indefinite udhén is with time reanalyzed as part of a higher projection, resulting in clausal negation. Thus, Greek in this model provides an instance both of a Move > Merge reanalysis and of a specifier > head reanalysis, because the transition from udhén to dhen additionally involved a change of syntactic status, from specifier to head (Roberts and Roussou : –). I adopt here their general syntactic analysis regarding the transformations of Greek N. However, the applicability of the characterization of Jespersen’s Cycle for the case of Greek is problematic, as has been noted in Willmott (). Willmott points out that although the course of Greek udhén is



Introduction

viewed in Roberts and Roussou (), as well as Roussou (), on a par with that of the colloquial French negator pas, in Greek, plain sentential negation never came to be expressed in a syntactically discontinuous way as it did in the French ne . . . pas (traditional Jespersen’s Cycle stage II). Willmott’s () objections to the applicability of the term ‘Jespersen’s Cycle’ for the case of Greek are justified, not because Greek is not a Jespersen’s Cycle language, but because the handbook definition of Jespersen’s Cycle is morphosyntactic in its description and typically identified with the developments of negation in French and other languages that followed this particular path in negator renewal: the addition of a second element (doubling stage) that was at some point reanalyzed as part of plain propositional negation and gradually assumed the sentential negator function on its own. This study aims to fix this situation by providing a semantic account of Jespersen’s Cycle that can be applied to both typical and atypical Jespersen’s Cycle languages, with Greek placed among the latter. Last but not least in the Greek Jespersen Cycle discussion is Kiparsky and Condoravdi’s () contribution, which provides a diachronic overview of Greek negation and related indefinites and minimizers, with emphasis on their attestations in Greek dialects (Cretan, Cypriot, among others). They identify several cases of transformation of minimizers to degree adverbs, which originate in more specialized functions and then become generalized (as was the case with the French ne . . . pas, see Schwegler , , de Cuypere ). The example below is from Kiparsky and Condoravdi . ()

Δεν πονώ Dhen ponó N hurt... ‘I don’t feel pain at all.’ Condoravdi : )

ψίχαλο. psíhalo crumb. (Macedonia, Hatzidakis , cf. Kiparsky and

Although it is not the case that any of these indefinites and minimizers took over the role of propositional negation in the respective dialect (cf. also Willmott ), although they are found as negative responses (anaphoric negation), Kiparsky and Condoravdi’s () paper is important and relevant to the present study in that it views Jespersen’s Cycle as a diachronic development—characterized as a semantic chain shift—whose motivation is semantic: the preservation of the contrast between ‘emphatic negation and plain negation’. We end up with the following view of its [Jespersen’s Cycle] nature and motivation. The contrast that the chain shift maintains is not that between affirmation and negation, as Jespersen assumes, but the contrast between plain and emphatic negation. And the weakening that undermines the contrast is not phonetic weakening of plain negation, but semantic weakening of emphatic negation. (Kiparsky and Condoravdi : , [italics in the original])

Thus, although sentential negation too is crosslinguistically marked compared to affirmation (Horn , Speranza and Horn ), it is the markedness of emphatic

Markedness, Nonveridicality, grounding



negation compared to nonemphatic that the reinforcement part of Jespersen’s Cycle preserves. This is an important insight—that in a way goes back to Meillet ()— whose influence is evident in the account and formalization of Jespersen’s Cycle developed in this book (Chapter ), which attempts to be descriptively inclusive also to negative existential cycles or Croft cycles (after Croft ), although such structures are not attested in the history of Greek.4

. Markedness, the Nonveridicality projection, and pragmatic grounding The notion of markedness in a broader understanding was multiply useful for the description of linguistic facts addressed and analyzed in this study, regarding emphatic and nonemphatic negation, and markedness of features and its relevance in language change most importantly as a property that differentiates between the two negators of Greek in all historical stages, as well as other languages that manifest a negator particular to nonveridical environments. This section spells out more clearly the sense in which markedness is used in this study and argues for the applicability of the term regarding nonveridical semantic environments, linked in particular to the nonveridicality projection introduced in chapter  (see Chatzopoulou a, c, ). We know markedness appears in the linguistic community since the School of Prague (Trubetzkoy , Jakobson ), first in Phonology and it subsequently expands. It plays a crucial role in the research of Greenberg (, ), Chomsky and Halle (), and Kean (), receiving more focused attention regarding all levels of linguistic analysis in the works of Andrews () and Battistella (), while evidence for its relevance in language acquisition, as already proposed by Greenberg, appears in Bardovi-Harlig (), Gass and Selinker () among others, further linked to language change (Smith , Elšik and Matras ). Focusing on the notion of nonveridicality (Giannakidou  et seq.), the implications of this study support a broader understanding of the notion of markedness in natural languages, one that includes the veridical and nonveridical (or realis and irrealis) hierarchical pair, with nonveridical or irrealis environments being crosslinguistically marked (cf. Uljas , Chatzopoulou a,b, , Michael ). Negative polarity items are by definition limited distribution elements, licensed in nonveridical environments only. Thus we explicitly claim that such environments are marked, not only semantically, but as we present in what follows, syntactically as well. This study identifies both representational and explanatory benefits for a syntactic treatment of polarity phenomena and semantic licensing in terms of syntactic agreement (Chomsky ) with a goal and a probe by positing a nonveridicality syntactic projection (Chatzopoulou a, c). Similarly to the syntactic treatment of See van Gelderen  and Larrivée  for further reading; for a more recent overview also van der Auwera et al. . 4



Introduction

negative concord phenomena crosslinguistically in the analysis of Zeijlstra (, ), according to which negative markers and n-words can vary in terms of whether they carry an interpretable [iN] or uninterpretable [uN] feature, the Greek data give motivation for the positing of a nonveridicality syntactic projection, with lexical elements carrying either an interpretable [iNV] or uninterpretable [uNV] nonveridical feature (section ..). The nonveridicality syntactic projection (NVP) is akin to but distinct from the Sigma phrase (ΣP) introduced in Laka () and from the Polarity phrase (PolP), introduced in Culicover () (more recently de Clercq ,  in a more refined syntactic analysis of the NP). As opposed to precursors (section ..), the NVP does not host affirmation or negation of the clause, but exclusively projects in nonveridical environments and allows for syntactic agreement relations with n-words, polarity items, among which is also sentential negation in early IndoEuropean languages and the majority of modern attested languages (van der Auwera and Lejeune , van der Auwera ). This approach may in fact be taken to originate in Haegeman’s () AFFECT-criterion, where a unifying feature is identified in both negative and interrogative propositions. This study broadens this claim to all nonveridical (traditionally known as affective) environments. Evidence from the history of Greek provides also lexical motivation for positing such a projection: in the transition from Attic Greek to Koine phonological developments, namely that homophony in the mood system (indicative, subjunctive, optative), jeopardized the (non)veridicality-based mood distinction, a fact which subsequently triggered the syntactic emergence of nonveridicality in an isomorphic way—lexically represented by the INA particle—as a projection related to, but in Koine Greek still distinct from (i) morphological mood, (ii) C. ()

ᾔτησέν μοι πρᾶξαι αὐτῇ ἵνα πέμψαι ítisén mi prákse autí ína pémpse ask... me. make.. her. INA send.. αὐτοὺς . . . autús them. ‘He asked me to make her (so as to) send them . . . ’ (P.Oxy. . (– c) f) (Translation by Joseph ( [])) (NVP lexicalized in ína)

()

οὐ μὴ παύσωμαι φιλοσοφῶν u: mε: pauso:mai philosopho:n N N cease.... pursue-knowledge..... ‘I shall never give up philosophy’ (emphatic) (NVP lexicalized in mε:) (Plato, Apologia Socratis d –.)

Markedness, Nonveridicality, grounding



As the argument has been made in Horn () and Speranza and Horn () for sentential negation to be crosslinguistically marked, this study, among other things, supports nonveridicality’s being described as crosslinguistically marked, with pragmatic grounding on the cooperative principle, descriptive of human verbal interaction, and particularly on the Gricean maxim of quality (Grice ): ‘do not say that for which you do not have sufficient evidence’, with the addition here that ‘if you do say that for which you don’t have sufficient evidence, make sure it is differentiated from that for which you do have’ (see Chatzopoulou c). In this sense, veridicality is the unmarked default and nonveridicality needs to stand out in some way, as a grammaticalization of the maxim of quality. In prototypical nonveridical environments, with (non)veridicality as a reflection of speaker commitment to the truth of the utterance or the degree of certainty, we have lexical elements added to the utterance, as is the case with sentential negation. Below is an example of nonveridical marking with the aid of a modal adverb. This is how speakers are able to be maximally cooperative, sharing potentially useful insights with the community, while at the same time avoiding blame in case the nonveridically marked proposition proves to be false. ()

a. John went to the store. b. John probably went to the store.

The cooperative principle, particularly regarding the maxim of quality (maxim of truthfulness) as a universal presumption in communication, appears to be linked to the Social Contract and as such has a fundamental role in human communicative interaction, as it expands the reliability and thus duration of any type of cooperation and maximizes the effectivity of negotiation (Chatzopoulou ). Given that markedness is a property of something in relation to something else, it is understandable that nonveridical environments are marked compared to veridical, nonveridical negation is marked compared to default negation and emphatic negation is marked compared to nonemphatic. The notion of economy of some sort is linked to markedness as well, regarding e.g. the markedness of parametric features (Roberts and Roussou , van Gelderen , b, section ...) and although an experimental study that would examine cognitive processing in nonveridical environments is to our knowledge still anticipated, there are studies that indicate excessive cognitive activity in the case of lying (Ofen et al.  and references therein) that signal deviation from the truthfulness maxim as more costly and less economical in terms of processing, a choice that requires more motivation than the default. We are aware of the gaps in the reasoning between the intuitive tendency toward and anticipation of true or sincere utterances and nonveridical marking in language. However, it is not inconceivable that this descriptive tendency in human interaction may have after some point become grammaticalized over a period of a few hundred thousand years from when Homo sapiens and earlier relatives are proven to have



Introduction

been present on the planet (Chatzopoulou ). The role of Pragmatics in language evolution may indeed have been understudied (cf. Sperber and Origgi , Scott-Phillips ), with the field of Evolutionary Pragmatics named relatively recently by Wildgen (). The present study, through the pragmatic grounding of nonveridicality marking crosslinguistically, as a reflection of the maxim of truthfulness, it is hoped will provide a slight improvement in this area, as well as promote better communication across linguistic fields.

. Roadmap In chapter  the (non)veridicality theory of polarity is presented with data from Standard Modern Greek and all the relevant phenomena are introduced. This theory was developed (since Giannakidou  et seq.) to account for limited distribution elements, such as the English anyone and the Modern Greek κανένας /kanénas/ ‘anyone/no one’ that have restrictions on their licensing environments. Crucially, the (non)veridicality theory of polarity of Giannakidou () places no categorial restrictions on the items that manifest negative polarity behavior. Several such elements are found in Modern Greek, both in the lexical domain (indefinites, minimizers, cf. Garzonio and Polleto ) and in the functional domain (subjunctive, perfective nonpast verb form). It is shown that the (non)veridicality theory of polarity is relevant for negator choice as well, not only in the dual-negator system of Greek, but for multiple other languages inside and outside the Indo-European family that have a distinct negator for nonveridical environments (e.g. prohibition, conditional protasis, epistemic and deontic modality, future tense, etc.). The second negator in these languages, N, is a negative-polarity item, an item licensed exclusively in nonveridical semantic contexts. Some evidence is provided from other old Indo-European languages, given that Greek inherited the polarity N from the Proto-Indo-European reconstructed form *meH. The journey begins with examples from Homeric Greek (eighth century ) in which the contrast and complementarity between N and N is evident. Chapter  is dedicated to Classical Greek, namely the Attic dialect, and provides a thorough examination of the functions of N and N at this stage (fifth-fourth centuries ) through qualitative and quantitative evidence. The full distribution of N and N can be directly captured by the (non)veridicality theory of polarity: all attestations of N and N, whether with finite or nonfinite forms (infinitives and participles) or even DP-internal, are accounted for with reference to the semantic property of (non)veridicality, as a property of the licensing environment in the sense of Giannakidou. It is argued that N is the marked form, occurring only in nonveridical environments, while N is the unmarked form of standard negation in the sense of Payne (). N is further identified as having a number of nonnegative functions, most significantly as a particle introducing yes/no questions

Roadmap



and a complementizer introducing verba timendi complements. Thus, the defining property of N that can unify all its manifestations in Attic is not negativity, but nonveridicality. This agrees with Joseph and Janda’s () treatment of N in Modern Greek, as a morphological constellation: an element with distinct but interrelated functions that center around a common core. This common core proves to be the property of nonveridicality in the distribution of N in Attic (as well as in Modern Greek). In chapter  the Hellenistic Koine stage of the Greek language is examined based on non-atticizing texts mainly from the first century  to the second century . General developments of the language are presented that support a treatment for nonveridicality as encoded in a syntactic projection, independent of morphological mood and independent of the C position. A treatment of the licensing of polarity items is proposed (among which is the Greek N) in terms of syntactic agreement in the sense of Chomsky (, ). Nonveridical operators are taken to introduce the Nonveridicality Phrase (NVP) in syntax, encoding the observation that nonveridical environments tend to be morphologically marked in ways that can be distinct from mood marking. According to this view, N in Koine Greek (as well as in the previous and the following stages) can be represented as bearing an uninterpretable [uNV] feature that is checked against the overt or covert head of the NVP. The differentiation of this projection from Laka’s () SigmaP and Culicover’s () Pol(ality)P is discussed and a correlation is also made with Cinque (). The next major stage of vernacular Greek that is dealt with regarding the distribution and basic properties of N and N is Late Medieval Greek, in chapter , with texts from the twelfth to the fourteenth centuries . Spoken Greek of this period depicts a stage of variation in the forms of N and N. Both elements have developed counterparts with which they appear in free variation, as the former N-thing οὐδέν /udhén/ and N-thing μηδέν /midhén/ indefinites have bleached into plain sentential negators. This variation would be permanently resolved by the eighteenth century. The new form of N (οὐ)δέν /(u)dhén/ would replace the N oὐ(κ) of Classical and Koine Greek. N μη /mi/, however, would remain stable in function and form. Two instances of parameter resetting in Late Medieval Greek are also discussed in this chapter: (i) the specifier-to-head shift in the syntactic status of the negators, which is predicted from the van Gelderen () Head Preference Principle, and (ii) the loss of N from the conditional protasis, explained through the upward micromovement of N on the Cinque () hierarchy, a movement to a position already occupied by the conditional particle of Late Medieval and Modern Greek. Thus the N was lost from the conditional antecedent for reasons that relate to its syntactic status shift and exact location on the expanded CP. Chapter  discusses the Greek negator transformations, from Attic Greek to Standard Modern, in relation to Jespersen’s Cycle (Jespersen , term coined by



Introduction

Dahl ), as described in the Jespersen’s Cycle literature. The developments of N and N in Greek do not properly qualify as instances of Jespersen’s Cycle according to the traditional understanding of the phenomenon, as it did not manifest a doubling stage (cf. Willmott ). A new approach for Jespersen’s Cycle is proposed, which accommodates not only traditional Jespersen’s Cycle languages, but also Greek and a number of other atypical languages that deviate in one way or another from the traditional morphosyntactic description of the phenomenon. It is proposed that Jespersen’s Cycle is a diachronic phenomenon whose regularities are to be found in the semantics: the bleaching of emphatic negation to plain propositional negation. A two-stage account for Jespersen’s Cycle is presented, an account prefigured from a cognitive-functional perspective in Furtado da Cunha () and de Cuypere (). A refinement for the intuitive notion of bleaching is also presented as scalar endpoint lexicalization, while the representations of diachronic change are described as instances of leftward lexical micromovement, a view that is in agreement with the perspective and micro-terminology introduced in Traugott and Trousdale () and Roberts (, ). The final part of chapter  offers an overview of the diachronically stable functions of N: (i) in negative directives, (ii) as a particle introducing yes/no questions, (iii) introducing verba timendi complements. The generalization is that all these are the C-related functions of N μη, while the use of N in conditionals (where it is neither in a C position nor related to one) was not maintained. The function of N as lexical negation, which is still preserved in Standard Modern Greek, appears to have been externally induced (through lexical borrowing from French and English during the twentieth century, cf. Efthimiou ). This function is very scarce in Late Medieval texts and is found only with participles, which are CP-projecting elements in these functions. It is argued that N μη did not eventually renew, because of the inertial pressures of its several nonnegative functions, which, being nonnegative, were not affected by Jespersen Cycle phenomena.

2 The (non)veridicality theory of polarity and negator selection In this chapter the central theoretical background of this study is presented, the (non)veridicality theory of polarity (Giannakidou  et seq., Zwarts ) with emphasis on the aspects which are relevant for the purpose of accounting for negator selection in the history of Greek. As the theory was developed primarily with reference to Modern Greek, I present here the relevant data of this synchronic linguistic stage. (Non)veridicality immediately explains limited distribution elements in Standard Modern Greek, which can be either lexical or functional. Among the latter is the Modern Greek N: μη(ν) /mi(n)/.

. Negative polarity: precursors and (non)veridicality (Giannakidou ) “Polarity” refers to the behavior of certain families of lexical elements that have restrictions on their distribution and cannot appear just anywhere with the same freedom that, e.g., simple noun phrases have. (b) gives an example of one of the most famous and intensively studied polarity items, a member of the any-paradigm of English. ()

a. I saw a boy yesterday. b. *I saw anybody yesterday.

Polarity phenomena have been identified within the field of linguistics since Buyssens’ () and Klima’s () articles on negation and negative environments in English. The phenomenon was originally thought to depend on the contrast between affirmation and negation, resulting in the characterization of such lexical elements as negative polarity items (Jackendoff , Baker ). Affirmation appears to disallow such limited distribution elements, while negation licenses them. ()

a. *I saw anyone yesterday. b. I didn’t see anyone yesterday.

Negation and Nonveridicality in the History of Greek. First edition. Katerina Chatzopoulou. © Katerina Chatzopoulou . First published  by Oxford University Press.



The (non)veridicality theory of polarity

The term affective was also used in order to refer to negative polarity items as requiring either overt negation for their licensing or some sort of negative entailment (Baker ) or implicature (Linebarger , ). Another dominant view that preceded the theory of (non)veridicality as an attempt to unify the environments in which negative polarity items are found arose with Ladusaw (), who proposed downward entailingness as the semantic licenser of negative polarity items. Example () below gives an environment that qualifies both as affective, according to the first tradition, and downward entailing, in relation to Ladusaw’s observations. ()

Few students saw any movies. Negative implicature of (): Most students didn’t see any movies. Downward entailment: Few students saw any French movies. ⟦French movies⟧ ⊆ ⟦movies⟧

Downward entailing operators such as few allow inference from sets to subsets; French movies denotes a proper subset of movies, thus, according to Ladusaw (), the negative polarity item any in () is rendered grammatical. However, both the idea of polarization (affirmation–negation) and monotonicity (Ladusaw ), although fruitful (Hoeksema , Zwarts , , van der Wouden , Kas , Dowty ), proved inadequate to account for the full distribution of negative polarity items. Most prominently, their presence in questions, imperatives, conditionals, generic statements, and under intensional operators, among others, is left unaccounted for, as it is not clear how these qualify either as negative or as monotonic (for full discussion see Giannakidou : –, : –, and ). The property under which all these environments—affective, downward entailing, and the rest—are unified was first identified in Giannakidou (), and was further developed in following publications, namely the property of (non) veridicality Giannakidou , et seq., Zwarts , see recently Giannakidou and Zeijlstra ). The primary motivation was to account for the distribution of the Modern Greek negative polarity items kanénas/KANENAS (non-emphatic and emphatic respectively), supported also through evidence from other languages.1 The (non)veridicality theory was able to account for limited distribution lexical items in ways previous approaches had failed to and since it was introduced it has found various applications in a number of languages, e.g., in Dutch (Hoeksema ), in Italian (Bernardi ), Romanian (Farkas , Falaus ), Salish (Matthewson ), Navajo (Fernald and Perkins ), Russian (Partee , Borschev et al. ), German (von Stechow ), in some southern dialects of Dutch (van Craenenbroeck ), in Chinese (Lin ), in Japanese (Yoshimura ), Korean (J. H. Lee , Giannakidou and Yoon ) and other languages (see Haspelmath a for a list of 1 The term (non)veridicality was introduced earlier in Montague (, ) with a somewhat different understanding, one that links the term to existence and truth (see Brandtler : ).

Negative polarity: precursors



languages that may be relevant for investigation under the rubric of (non)veridicality). This study is one more application of the (non)veridicality theory, in that it strengthens the empirical domain of the notion by showing that (non)veridicality is the factor that regulates negator choice in the history of Greek. (Non)veridicality is formally defined as a property of propositional operators (definition from Giannakidou : ). () (Non)veridicality for propositional operators (i) A propositional operator F is veridical iff Fp entails or presupposes that p is true in some individual’s epistemic model M(x); otherwise F is nonveridical. (ii) A nonveridical operator F is antiveridical iff Fp entails that not p in some individual’s epistemic model: Fp →¬ p in some M(x). () Definition for polarity items A linguistic expression α is a polarity item iff: (i) The distribution of α is limited by sensitivity to some semantic property β of the context of appearance; and (ii) β is (non)veridicality. A function F is veridical if Fp entails or presupposes that p is true in some individual’s epistemic model: the set of possible worlds compatible with a person’s beliefs or anchored to an individual (cf. the individual anchor; Farkas a). A veridical operator entails the truth of p in all worlds in the model, while a nonveridical operator expresses uncertainty: there are some worlds w where p is true, and some worlds w’ where it is not. Within the class of nonveridical operators is the class of antiveridical ones, among which is negation; antiveridical operators entail the falsity of p. In other words, veridical operators reflect the speaker’s certainty and commitment to the truth of the proposition which is uttered, whereas nonveridical operators reflect uncertainty and lack of commitment.2 Some prototypical nonveridical environments are given in table . next to examples from the English any-paradigm. The Greek negative polarity items κανένας /kanénas/ ‘nobody/anybody’ lexicalize a subset of the uses of English any, the rest being instances of free choice items. Free choice items, lexicalized separately in Greek by οποιοσδήποτε /opjosdhípote/ ‘anybody’, are a particular sort of polarity-sensitive lexical elements, elements dependent on nonveridicality, which, however, will not be discussed here to any extent, as they posit additional restrictions on their contextual felicity, namely the property of nonepisodicity (see Giannakidou  for Modern Greek free-choice items and Giannakidou ,  for the semantics of free choice in general, also Haspelmath  for free-choice items in general). It is important to note that the (non)veridicality theory of polarity poses no categorical restrictions on the elements that depend on it (Giannakidou : –, –). 2

For epistemic models see also Heim () and Tancredi () among others.



The (non)veridicality theory of polarity TABLE . Prototypical nonveridical environments

Nonveridical

Antiveridical

Semantic environments

Examples with any in English

Negation without-clause

Ariel didn’t talk to anyone. Phillip entered without anyone noticing him.

before-clause

Ella left before eating anything.

conditional protasis

If you see any wolves, go inside and lock the door.

imperatives

Kiss any frog.

interrogatives

Did you eat any berries?

modal verbs

She could see anyone from the balcony.

generics

Any wolf eats pigs.

downward entailing

Few dwarfs brought any diamonds.

This is made clear in the next section, where a brief overview of negative polarity items found in Standard Modern Greek is presented. These include lexical (indefinites, minimizers, focus particles), as well as functional (mood choice, negator choice) manifestations of sensitivity to (non)veridicality.

. Polarity phenomena in Standard Modern Greek .. Negative polarity items In Standard Modern Greek two varieties of lexical negative polarity items can be found: (i) those that are sensitive only to negation and negative-like operators and (ii) those that are sensitive to nonveridical operators in general (Giannakidou ). The distinction between the two, for the most frequent polarity items in Modern Greek, lies in the use of emphatic accent, an observation already made in Veloudis (). Although the two paradigms are segmentally the same, those that are sensitive only to negation bear emphatic accent, whereas those that are sensitive to nonveridicality in general do not. Below are the two paradigms. The nominal ones are marked for gender and case. In () masculine nominative forms are cited. ()

kanέnas/ KANENAS ‘anyone, anybody/ no one, nobody’ kanέnas N/ KANENAS N ‘any N-singular/ no N-singular’ típota Nplural ‘any N-plural/ no N-plural típota/ TIPOTA ‘anything/ nothing’ potέ/ POTE ‘ever/ never’ puthεná/ PUTHENA ‘anywhere/ nowhere’ kathólu/ KATHOLU ‘at all/ not at all’ kathólu Nmass/ KATHOLU Nmass‘any Nmass/ no Nmass’

Polarity phenomena in Standard Modern Greek



Emphatics exhibit a strong dependency upon nonveridicality and are grammatical only within the scope of antiveridical operators (negation, without-clauses, and before-clauses), while nonemphatics have a weaker dependency, in that they are grammatical with nonveridical operators in general (antiveridical operators included), as in the antecedent of conditionals, in questions, disjunctions, perhapsclauses, modals, downward entailing DPs, habituals, and in the complement of negative verbs and intensional verbs. In their cooccurence with sentential negation the Modern Greek negative polarity items give rise to structures known in the literature as Negative Concord (NC) of the strict kind, in that the presence of negation is required (see the Negative Concord typology in Giannakidou , ). Some examples of the Modern Greek negative polarity items in representative environments appear below. ()

Ο Γιάννης δεν είδε κανέναν/ΚΑΝΕΝΑΝ. o Jánis dhεn ídhε kanέnan/ KANENAN the. Jiannis. N see.. [n-body./N-BODY.] ‘John didn’t see anybody/ANYBODY.’ (emphatics and nonemphatics with negation)

()

Αν θέλεις κανένα /*ΚΑΝΕΝΑ γλυκό να μου πεις. An thέlis kanέna/*KANENA ghlikó na mu pis if want. n-/ n- candy  I. say.. ‘If you want any/ *no candy, tell me.’ (only nonemphatics in conditional protasis)

()

Είπες τίποτα/ *ΤΙΠΟΤΑ; ípεs típota/ *TIPOTA say... n-thing Did you say something? (only nonemphatics in interrogatives)

() Έφυγε χωρίς να πει τίποτα/ ΤΙΠΟΤΑ. Έfighε xorís na pi típota/ TIPOTA leave... without  say.. n-thing/N-THING ‘S/he left without saying anything/ANYTHING.’ (emphatics and nonemphatics with negation) Another class of lexical items that exhibit strong negative polarity behavior which has in fact crosslinguistic representation is the class known as minimizers (see Zanuttini , Garzonio and Polleto , and for Greek in particular, Giannakidou , , Veloudis , , Kiparsky and Condoravdi , and Roussou ), predicates that are to various extents idiomatized and which denote minimal amounts. In Modern Greek these items are grammatical only in the scope of antiveridical operators (exx , ) and they are generally focused, while in



The (non)veridicality theory of polarity

other nonveridical environments they are more marginal, although intuitions can vary (ex. ). They are, however, banned in positive episodic environments (ex. ). () Δεν είπε den ípε N say... ‘S/he didn’t say a word.’

λέξη. lέksi word

() Δεν έφαγε dεn έfaghε N eat... ‘S/he didn’t eat a bite.’

μπουκιά. bukjá bite

() *Χθες ο Γιάννης hthεs o Jánis yesterday the John ‘John ate a bite yesterday.’

έφαγε έfaghε eat...

() ??Είπες λέξη για να ípεs lέksi ja na say... word in-order to ‘Did you say a word to defend them?’

μπουκιά. bukjá. bite

τους tus them

υπερασπιστείς; ipεraspistís? defend..

.. The focus particle ούτε /útε/: an instantiation of a negative polarity EVEN Another manifestation of polarity sensitivity that will be of particular relevance to our discussion is found in the Modern Greek counterparts of the English focus particle even. The semantics and pragmatics of even have received considerable attention since Horn () and Fauconnier (), followed by Karttunen and Peters (), and a number of studies, among which are Rooth (), Horn (), Kay (), König (), Wilkinson (), Giannakidou () and Israel (). Standard Modern Greek employs three different expressions in lexicalizing the functions of the English even, all three sensitive to (non)veridicality: (a) a positive polarity element, ακόμη και / akómi kε/ (morphologically: ‘still and’), (b) a negative polarity element, ούτε /útε/ (etymologically: ‘not-and’)and (c) a flexible scale element, έστω /έsto/ (etymologically: ‘be..’). Our focus here is on the negative polarity item ούτε /útε/, optionally followed by the expression καν /kan/ (etymologically: ‘and-if ’), but examples of all three of them are provided below (see for full discussion Giannakidou ). () Ακόμη και ο Γιάννης akómi kε o Jánis EVEN the Jánis ‘Even Jánis came.’

ήρθε. írthε come.. (positive polarity EVEN)

Polarity phenomena in Standard Modern Greek



() Ούτε (καν) ο Γιάννης (δεν) ήρθε. útε (kan) o Jánis (dhεn) írthε EVEN the Jánis N come.. ‘Even Jánis didn’t come.’ (negative polarity EVEN ) () Ήρθε έστω ο Γιάννης; írthε έsto o Jánis? come.. EVEN the Jánis ‘Did even (read as: ‘at least’) Jánis come?’

(flexible scale EVEN )

Similarly to their English counterparts, all three expressions involve reference to a scale in terms of a presupposition or conventional implicature, but the negative polarity EVEN ούτε /útε/ must be in the scope of an antiveridical operator (negation, without-clause) in order to be licensed, either overtly or covertly (Giannakidou : –). The scale that both the negative polarity EVEN ούτε /útε/ and the positive polarity EVEN ακόμη και /akómi kε/ presuppose is that of likelihood, understood as a possibility scale (Horn , ). The corresponding English examples from Giannakidou  make these observations more clear. ()

The Dean invited even Bill. i. ∃x [x ≠ Bill ∧ C(x) ∧ invited (Dean, x)], and ii. ∀x [x ≠ Bill → [likelihood (Dean inviting x) > likelihood (Dean inviting Bill)]]

()

The Dean didn’t invite even Bill. i. ∃x [x ≠ Bill ∧ C(x) ∧ ¬invited (Dean, x)], and ii. ∀x [x ≠ Bill → [likelihood (Dean inviting Bill) > likelihood (Dean inviting x)]]

Thus, the presuppositional contribution of the negative polarity item ούτε /útε/ is identical to the one of the positive polarity item ακόμη και /akómi kε/, as represented in their denotations in (), with the exception that in the case of the negative polarity item ούτε /útε/ we have an instance of scale reversal under negation and Jánis from being the least likely person to come in example () becomes the most likely person to come in example (), similarly to Bill of examples () and () in receiving an invitation from the Dean. ()

a. ⟦akomi ke⟧: λx λP: ∃y [y ≠ x ∧ C(y) ∧ P(y) ∧ ∀y [y ≠ x → (likelihood (P(x)) > likelihood (P(y))]. P(x) b. ⟦oute (kan)⟧ = λx λP: ∃y [y ≠ x ∧ C(y) ∧ ¬P(y)] ∧ ∀y [y ≠x → (likelihood (P(x)) > likelihood (P(y))]. P(x)

As a result, ακόμη και /akómi kε/ lexicalizes a low likelihood EVEN, while ούτε (καν) /útε (kan)/ lexicalizes a high likelihood EVEN, which is in support of Rooth’s



The (non)veridicality theory of polarity

() ambiguity analysis regarding the item even in English. Still they are both sensitive to (non)veridicality, in that the positive polarity ακόμη και /akómi kε/ must escape the scope of antiveridical operators, while the negative polarity ούτε (καν) /útε (kan)/ must be in the scope of an antiveridical operator. The syntactic analysis of ούτε /útε/ in Giannakidou () attributes an uninterpretable negative feature [uN] to ούτε /útε/ that must be checked against a potentially null negation, Ø N, a negative polarity negation that carries both interpretable/categorial and an uninterpretable/inflectional negative feature [Ø N: CAT [iN], INFL [uN]], in order to account for the fact that in preverbal position the sentential negation marker is optional (), while in postverbal it is required (). This analysis will be relevant in chapters  and  where the negative indefinites of Attic and Hellenistic Greek are discussed that manifest a similar pattern. () Ούτε (καν) ο Γιάννης útε (kan) o Jánis EVEN the Jánis ‘Even Jánis didn’t come.’

(δεν) (dhεn) N

() Δεν ήρθε ούτε (καν) dhεn írthε útε (kan) N come.. EVEN ‘Even Jánis didn’t come.’

ήρθε. írthε come..

ο o the

Γιάννης. Jánis Jánis

Furthermore, the flexible scale item έστω /έsto/ also exhibits polarity behavior, in that it is licensed by nonveridical operators, such as interrogatives and imperatives, but not under negation (cf. Giannakidou : –). As opposed to the negative polarity ούτε (καν) /útε (kan)/ and the positive polarity ακόμη και /akómi kε/, the item έστω /έsto/ makes reference to a scale that is not specifically that of likelihood, but a scale made salient by the context. This scale can relate to quantities and minimal amounts. () Έφαγες έστω ένα μήλο; Έfaghεs έsto έna mílo? eat.. EVEN one apple ‘Did you even (read as: ‘at least’) eat an apple?’ The behavior of the flexible scale item έστω /έsto/ will be relevant in the account proposed for Jespersen’s Cycle and the formalization for the particular kind of grammaticalization and bleaching that Jespersen’s Cycle instantiates. In Greek and many other languages Jespersen’s Cycle proceeds through the transformation of a scalar indefinite (possibly negative) into propositional negation. In the case of Greek, this scalar item was the Classical Greek negative indefinite οὐδέν /u:den/ ‘nothing’

Polarity phenomena in Standard Modern Greek



(morphologically: ‘not-even-one’). As has been argued for in Giannakidou and Yoon (), it is not the case that scalarity is identified with negative polarity, as there are plenty of nonscalar NPIs, among which is also the Greek N. However, the ούτε /útε/ variant of the Greek EVEN is both an NPI and a scale-evoking element. So far, multiple instances of lexical items exhibiting polarity sensitivity in Modern Greek have been presented. In the next section we turn toward the ways in which (non)veridicality can get grammaticalized in the functional domain. .. (Non)veridicality and mood choice Standard Modern Greek distinguishes three morphological moods: indicative, subjunctive, and imperative. Apart from the regulating factor of (non)veridicality in mood choice, which is anticipated and based on the split between (i) nonindicative (in nonveridical environments, cf. Giannakidou , ) and (ii) indicative (unmarked), Modern Greek grammaticalizes a particular kind of verb form, which is deficient in that it carries no formal mood marking: the perfective non-past (PNP). The PNP verb form is dependent (Joseph and Philippaki-Warburton , Holton, Mackridge and Philippaki-Warburton , Haberland ), in that it cannot stand on its own in the way the other temporal-aspectual combinations can, as can be seen in table . (examples from Giannakidou ).

TABLE . Tense and aspect combinations in Standard Modern Greek Aspect Imperfective

Perfective

Γράφω. ghráf-o (IN) ‘I am writing.’

*γράψω ghráps-o (PNP) ?

Past

Έγραφα. éghraf-a (IP) ‘I was writing.’

Έγραψα. éghrapsa (PP) ‘I wrote.’

Tense

Non-past

The perfective non-past of Modern Greek has been analyzed in Giannakidou (, ), within a pronominal treatment of tense (Partee , , Kratzer ), as exhibiting polarity behavior of a temporal kind in that it carries a deficient time variable that cannot receive a time value from context alone, as opposed to what is generally the case with nondeficient verb forms. It rather requires the presence of one of the nonveridical particles of Modern Greek that apparently can perform this function: the subjunctive particle να /na/, the future particle θα /tha/, the conditional



The (non)veridicality theory of polarity

αν /an/, the before particle πριν /prin/, the after particle αφού /afú/, the optative particle ας /as/, or the Modern Greek N μη(ν) /mi(n)/, either in its use as a prohibitive or in its complementizer function (a point we return to in the following section). () *Γράψει ghrápsi write.. ??John write

ο o the

Γιάννης. Jánis Janis (PNP on its own is ungrammatical)

() Να γράψει ο Γιάννης. na ghrápsi o Jánis  write.. the Janis ‘John should write/ Let John write.’ () Θα γράψει ο Γιάννης. tha ghrápsi o Jánis  write.. the Janis ‘John will write.’ () Ας γράψει ο Γιάννης. as ghrápsi o Jánis  write.. the Janis ‘Let John write.’ () Αν/Πριν/Αφού γράψει an/prin/afú ghrápsi if/before/after write.. ‘If/before/after John writes . . . ’

ο o the

Γιάννης . . . Jánis Janis

() Μη γράψεις. Mi ghrápsis N write.. ‘Don’t write.’ () Φοβήθηκα μη γράψει ο Γιάννης. Fovíthikan mi ghrápsi o Jánis fear.. N write.. the Janis ‘I was afraid that John would write.’ (N in complementizer position) The unacceptability of the perfective non-past verbal form in positive unembedded assertions is thus one more instance of a semantic sensitivity to (non)veridicality found in Standard Modern Greek, a fact already observed in Giannakidou and Zwarts ().

Polarity phenomena in Standard Modern Greek



Furthermore, as mentioned at the beginning of this section, the subjunctive, which in Modern Greek is analytical and not synthetic—contrary to the morphological subjunctive of Classical and Hellenistic Greek—is itself a polarity item of the functional kind (Giannakidou , , ). The Modern Greek subjunctive consists of the subjunctive particle να /na/ combined with either the imperfective non-past (IN), the imperfective past (IP) or the previously discussed PNP verb form (see table .). In its unembedded functions it carries directive force (nonveridical) and can express suggestion, invitation, as well as counterfactuality (see examples below). () Να γράφει. na ghráfi  write.. ‘S/he should be writing.’ () Να έγραφε. na ghrápsi  write.. ‘S/he should have written.’ () Να γράψει. na ghrápsi  write.. ‘S/he should write.’

(directive)

(directive counterfactual)

(directive)

The subjunctive particle να /na/ can also combine with the perfective past (PP in table .),3 but only in dubitative yes/no questions (nonveridical), where we can claim that it is linked to a null complementizer that holds the position of a question particle, see (a) (for the dubitative use of να /na/ see also Rouchota ). Nα /na/ can also introduce ordinary (non-dubitative) yes/no questions in combination with the perfective non-past (PNP in table .) verb form, see (b). () a. Να έγραψε; na έghrapsε?  write.. ‘Has s/he written (I wonder)?’ b. Να γράψει; na grápsi?  write.. ‘Should s/he write?’

(dubitative question)

(yes/no question)

Να /na/ is also found with imperfective non-past in interrogative function. In this use it can be described as borderline between a dubitative and an epistemic modal. In either case, this is again a nonveridical environment. 3



The (non)veridicality theory of polarity

This analysis is in agreement with the one proposed in Giannakidou (, ) regarding the previous uses of the unembedded subjunctive that have directive force, according to which the particle να /na/ heads the MoodP (Philippaki-Warburton , ; Philippaki-Warburton and Veloudis , Tsimpli , Giannakidou ), but is also linked to a covert complementizer that carries directive illocutionary force. All these uses can also be characterized as modal. The Modern Greek subjunctive, however, derives its characterization as a ‘subjunctive’ in that it is the mood par excellence of subordination and embedded propositions. Given that (i) Modern Greek does not exhibit nonfinite complementation, as opposed to its ancestors at least until Early Medieval and (ii) that the imperative mood does not appear in embedded contexts, in agreement with the crosslinguistic tendency regarding morphological imperatives (cf. Giorgi and Pianesi , Portner ), complementation in Modern Greek can take place either through the indicative or through the subjunctive. () Γνωρίζω ότι έρχεται. ghnorízo óti έrhεtε know.. that come.. ‘I know s/he is coming.’

(factive verb with indicative complement)

() Θέλω να έρχεται. thέlo na έrhεtε want..  come.. ‘I want him/her to be coming (now and then).’ (volitional verb with subjunctive) The regulating factor that determines mood choice in embedded contexts in Modern Greek has been shown to relate to the the embedding propositional attitude verb, the semantics of the verb form that licenses either the indicative or the subjunctive (Giannakidou , , , , , more recently a/b, ).4 This again boils down to (non)veridicality, rendering the subjunctive a polarity item of the functional kind, as it is selected by nonveridical verbs, such as volitionals or directives, that express lack of commitment on the truth of the embedded proposition. A list with representative verbs is given below (cf. Giannakidou , also Veloudis : – for discussion).

For different approaches that highlight different factors in mood choice see Farkas (, a,b) for an explanation based on modality and Rivero and Terzi () for an account in terms of illocutionary force. 4

Polarity phenomena in Standard Modern Greek ()



Nonveridical verbs select the subjunctive volitionals: θέλω /thέlo/ ‘I want’, ελπίζω /εlpízo/ ‘I hope’ directives: διατάζω /diatázo/ ‘I command’, συμβουλεύω /simvulεvo/ ‘I advise’ modals: πρέπει () /prέpi/ ‘must’, μπορεί () /borí/ ‘may’, μπορώ /boró/ ‘I can’ permissives: επιτρέπω /εpitrέpo/ ‘I permit’ commissives: υπόσχομαι /ipóshomε/ ‘I promise’, ορκίζομαι /orkízomε/ ‘I swear’ negative: αποφεύγω /apofέvgo/ ‘I avoid’, αρνούμαι /arnúmε/ ‘I deny’ verbs of fearing (verba timendi): φοβάμαι /fovámε/ ‘I fear’, διστάζω /dhistázo/ ‘I hesitate’

On the other hand, verbs that reflect commitment of at least one epistemic agent (the speaker or the subject of the main verb) to the truth of the complement sentence select the factive complementizers ότι /óti/ or που /pu/ with the indicative (Giannakidou : , : –). ()

Verbs with indicative assertive: λέω /lέo/ ‘I say’, διαβάζω /djavázo/ ‘I read’, ισχυρίζομαι /ishirízomε/ ‘I claim’ fiction verbs: ονειρεύομαι /onirεvomε/ ‘I dream’, φαντάζομαι /fandázomε/ ‘I imagine’ epistemic: πιστεύω /pistέvo/ ‘I believe’, νομίζω /nomízo/ ‘I think’ factive: γνωρίζω /ghnorízo/ ‘I know’, μετανιώνω /mεtanjóno/ ‘I regret’ semifactive: ανακαλύπτω /anakalípto/ ‘I discover’, θυμάμαι /thimámε/ ‘I remember’

While some verbs like γνωρίζω /ghnorízo/ ‘I know’ have the option of selecting either the indicative or the subjunctive, the choice of mood still reflects a dependency to (non)veridicality and variation in mood choice is usually accompanied by a shift of meaning of the verb (Giannakidou , Quer ), in that the choice of mood highlights different aspects of the selecting verb form. But even when this is not the case, as with the verb θυμάμαι /thimámε/ ‘I remember’, this still agrees with the view that the subjunctive is the marked mood in terms of nonveridicality, while the indicative is unmarked. Therefore Modern Greek grammaticalizes (non)veridicality within the language system in many ways, regarding both lexical and functional categories. As noted, in Giannakidou’s theory of polarity there are no category restrictions on the elements that can manifest this sensitivity on their licensing environment, in terms of the property of (non)veridicality. In what follows I present one more instance of sensitivity to (non)veridicality manifested in Greek as well as other languages within the realm of functional categories, which is the topic of this study: choice of negator.



The (non)veridicality theory of polarity

.. (Non)veridicality and negator choice Negative polarity items such as the ones we have been examining so far have been described as “les satellites de la negation” (Gaatone ), the satellites that revolve around negation. In this section we see that negation itself can be a polarity item, in that it can exhibit an allomorph conditioned by the semantic environment, namely the property of (non)veridicality. This allomorph in Standard Modern Greek is the N μη(ν) /mi(n)/, which, although itself an antiveridical operator in its prototypical uses, is a negative-polarity item at the same time. In this chapter evidence is presented from Modern Greek, accompanied by crosslinguistic analogies; and our historical journey begins with an overview of the Proto-Indo-European negator system, as reconstructed through the oldest Indo-European languages, among which is also Homeric Greek (eighth century ). ... Standard Modern Greek negators Negation in logic refers to the one-place propositional operator whose function is to reverse the truth value of the proposition it applies to. ()

[[¬φ]] =  iff [[φ]] =  [[¬φ]] =  iff [[φ]] = 

In Standard Modern Greek, as in all Greek, this function can be performed by two expressions that are in complementary distribution. These are N δε(ν) /dhe(n)/ and N μη(ν) /mi(n)/ and they both can be the vehicles of logical negation (Tzartzanos , Mirambel , , Browning , Veloudis , Giannakidou , ). () Ο Γιάννης o Jánis the. Jánis. ‘John did not come.’

δεν/*μην dhen/ *min N/*N

() Μη/*δεν φέρεις mi / *dhen féris N/*N bring.. ‘Don’t bring John (along).’

ήρθε. írthe came.

τον ton the.

Γιάννη! Jáni Jánis.

In terms of the typological division on the expression of negation in Miestamo (a, b) and Miestamo and van der Auwera (), Standard Modern Greek manifests an instance of symmetric standard negation (negation in declaratives, following the terminology of Payne ). The presence of N δε(ν) /dhe(n)/ in () is the only difference the proposition has to the corresponding positive (). No other morphological or syntactic alterations take place between the positive and

Polarity phenomena in Standard Modern Greek



the negated form of a declarative, as opposed to what the situation is e.g. in English, (see translated examples below). () Ο Γιάννης δεν o Jánis dhεn the Jánis N ‘John did not come.’ () Ο Γιάννης o Jánis the Jiannis ‘John came.’

ήρθε. írthε came. (negated declarative)

ήρθε. írthe came. (corresponding affirmative)

Things are more complicated when it comes to Modern Greek N, a point I return to in the next section. Regarding their syntactic status, it has been argued that the Modern Greek N and N are heads, as their behavior is similar to that of the negative markers in Romance (cf. Zanuttini , , see Giannakidou  for discussion). The head status of Modern Greek negators is represented in figure ., along with the relative ordering between the MoodP and the NP in Standard Modern Greek. N δε(ν) /dhe(n)/ has been described as the negator of indicatives, and N μη(ν) /mi(n)/ as the negator of nonindicatives (Joseph and Philippaki-Warburton , Giannakidou ). The explanation, however, of the generalization has remained mysterious. Attempts have been made to identify the single semantic parameter that explains it, as in the paper of Philippaki-Warburton and Spyropoulos () who propose that negator selection in all Greek—Modern

MoodP Mood’ Mood

NegP Neg’

Neg

a. b.

Ø na a. ‘S/he didn’t come.’

TP

dhεn (Neg1) min (Neg2)

írthε (come.pp.3sg) érthi (come.pnp.3sg)

b. ‘Let him/her not come.’

FIGURE . The syntactic status of the Modern Greek negators



The (non)veridicality theory of polarity

Greek included—can be reduced to the contrast between epistemic and deontic modality: N is claimed to relate to epistemic modality and N to deontic. The insufficiency of this view is a point we soon return to. Joseph and Janda () provide a detailed overview of the uses of the Modern Greek N μη(ν) /mi(n)/, proposing a treatment of the N as a morphological constellation, a family of synchronically—and possibly diachronically—related manifestations of a lexical element that center around a common semantic core. Morphological constellation: A group of elements which share at least one characteristic property of form, but are distinguished by individual idiosyncrasies—of both form and function—that prevent their being collapsed with one another. (Joseph and Janda : )

It is, however, unclear what exactly this semantic core—or Gesamtbedeutung, in terms of Jakobson ()—is. In this section, it is shown that the distribution of the Modern Greek N and N immediately falls from the (non)veridicality theory of polarity: the Modern Greek N is itself a polarity item that is licensed exclusively in nonveridical contexts. ... The distribution of N and N in Standard Modern Greek As noted in the previous section, the standard negation of declarative sentences in Modern Greek is N δε(ν) /dhe(n)/. N is not only the negator of matrix clauses, but also appears in embedded environments that are in the indicative. () Ο Γιάννης δεν o Jánis dhen the Jiannis N ‘John did not come.’

ήρθε. írthe come..

() Γνωρίζω ότι δεν ghnorízo óti dhen know.. that N ‘I know s/he did not come.’

(negated main clause) ήρθε. írthe come.. (negated complement of a factive verb)

N μη(ν) /mi(n)/, on the other hand, is the negator found in non-declaratives, most prominently the negator of the subjunctive (Veloudis , Joseph and Janda ), either (i) embedded, as when selected by a nonveridical verb form, or (ii) unembedded as a surrogate for true negative imperatives (TNIs). Modern Greek is among the languages that lack TNIs (see also Joseph and Philippaki-Warburton , Rivero and Terzi , Han ) for reasons that relate to the relative ordering of MoodP and NP in Modern Greek (see fig. .) and the head status of the negators (which intereferes with the movement of the verb to CP in the construction of the imperative, cf. Giannakidou : –). In this sense the Modern Greek N exemplifies an instance of asymmetric

Polarity phenomena in Standard Modern Greek



negation, in the terminology of Miestamo and van der Auwera (): the morphological imperative has no negative counterpart. Examples of all these uses are given below. () a. Ο Γιάννης παρακάλεσε τη Μαρία να μην έρθει. o Jánis parakálεsε ti María na min έrthi the Jannis request.. the Maria  N come.. ‘John asked Maria not to come.’ (embedded subjunctive in the scope of a directive) b. Ο Γιάννης ντύθηκε για να μην κρυώσει. o Jánis díthike já na min kriósi the Jannis dress... in.order  N get.cold.. ‘John got dressed in order not to catch a cold.’ () *Μην έλα. min έla N come.. (‘Don’t come!’)

(TNIs unavailable in Standard Modern Greek)

() Να μην Na min  N ‘Don’t come.’

(unembedded subjunctive as surrogate for TNIs)

έρθεις. έrthis come..

In its unembedded directive function N can appear simply followed by the perfective non-past or the imperfective non-past, while the subjunctive particle να /na/ can be dropped, as in () and (), or it can be replaced by the hortative particle ας /as/ (etymologically related to the Medieval  imperative άφες /áfεs/ ‘let’), see (a). The meaning remains that of a prohibition, and unaffected. It feels slightly more polite in the case of the hortative ας /as/, which can also be dropped in some cases, especially when the verb form is in plural, see (b). All these uses, however, can be described as directive. () Mην έρθεις. min έrthis N come..  ‘Don’t come.’ () Mην έρχεσαι. min έrhεsε N come.. ‘Don’t come.’

(na particle dropped)



The (non)veridicality theory of polarity

() a. Ας μην έρθεις. as min έrthis  N come.. ‘Don’t come (You have permission not to come).’ (N with exhortative ας /as/) b. (Ας) μην ξεχνάμε ότι ... (as) min ksεhnámε óti () N forget.. ‘Let’s not forget that . . . ’ (exhortative ας /as/ optional) So far it has been shown that subjunctives and the perfective non-past are negated only through N. However, a split between the two negators determined by mood choice, considering N to be the negator of the indicative and N of the nonindicative, would not be accurate. Modern Greek exhibits two additional environments in which N can appear with the indicative: (a) as a question particle (Joseph and Janda , Chatzopoulou , a, a/b), and (b) as a complementizer selected by verbs of fearing or other nonveridical verbs. In these uses N μη(ν) /mi(n)/ is compositionally nonnegative. In introducing yes/no questions, N has been described as dubitative and it does have some sort of pragmatic effect, usually that of the undesirability of the embedded proposition, but this is not general. In this function it is interchangeable with μήπως /mípos/ ‘lest’, which has in fact been analyzed as a sort of evaluative negation (see van der Wouden : , Yoon : –). () Είπε ότι θα τηλεφωνήσει, αλλά μην / μήπως ξέχασε; ípε óti tha tilεfonísi alá min / mípos ksέxasε? said.. that  call.. but N / Q forget.. ‘S/he said s/he will call, but maybe s/he forgot?’ We see here that N μη /mi/ and μήπως /mípos/ (morphologically: N.how) translate with a modal particle in English and the question is not a negative question. In the use below, we have the Modern Greek N in the N N serialization, resulting to only one instance of logical negation. Again, N appears compositionally vacuous, in that it does not contribute negative meaning, though it does seem to associate with modality (), or desirability (). () a. Φοβάμαι μην / μήπως fovámε min / mípos fear.. N / lest ‘I fear s/he may come.’

έρθει. έrthi come..

b. Φοβάμαι μη / μήπως δεν fovámε mi / mípos dhεn fear.. N / lest N ‘I fear s/he may not come.’

έρθει. έrthi come..

Polarity phenomena in Standard Modern Greek



CP

C’

C

TP

min/mípos Neg2/QP ‘Lest s/he forgot?’

ksέxasε? forget.pp.3sg

FIGURE . The Modern Greek N in complementizer position

In these functions N is structurally higher than NP, occupying the head of the CP (figure .). This development from negator to question particle and complementizer, as a sort of expletive negation, is not uncommon crosslinguistically (Heine and Kuteva , van Gelderen a, ). What may be uncommon is that the Greek N has been inert in these positions for twenty-five centuries (see chapter ). This function of N relates to the general question particle of Modern Greek, άραγε /áraje/ (Giannakidou : ). N μη(ν) /mi(n)/, however, is particular to polar questions, while άραγε /áraje/ has no such restriction. There is a morphological link between the two, which goes back to Attic Greek: the ἆρα μή /a:ra mε:/ serialization, discussed in section ... Furthermore, N has a function as anaphoric negation that can stand in a clause on its own to express prohibition of an action that is happening or is just about to happen (see also Joseph ). ()

Context: John reaches for the last slice of baklava, which Maria is saving for herself. MARIA: Μη! mi N ‘Don’t!’ or ‘No!’

Although this use can be viewed as an instance of ellipsis, an elliptical fragment (cf. Merchant ), through the omission of a perfective non-past or imperfective non-past form of the verb, this instantiation of the Modern Greek N is best analyzed as a mere indexical, with no morphological mood marking at all.



The (non)veridicality theory of polarity

In the light of these data, the split between the indicative and the nonindicative as a determinant of the division of labor between N and N in Standard Modern Greek is partial and only epiphenomenal. What can be argued instead is that N is a negative polarity item licensed exclusively in nonveridical environments. The cooccurence of N with the nonindicatives, the subjunctive να /na/ or the exhortative ας /as/ particles, rests in that both the Modern Greek N μη(ν) /mi(n)/ and the nonindicatives are dependent on nonveridicality too. Table . summarizes the uses of N and N from which the polarity behavior of N is evident: all environments in which N appears qualify as nonveridical according to the definition for (non)veridicality that was presented in (). Table . includes also the use of N as the negation of non-finite elements (cf. Joseph , , ), negation found in adjectives, participials and less frequently on nouns. Although NP and DP internal negation lie in the periphery of this investigation, our main focus being the expression of sentential negation, some examples of this use of N follow, as it is among the functions of the Greek N that go back at least as early as the Classical era (fifth-fourth centuries ) where N did covary with N (see chapter ). In Modern Greek N δεν /dhεn/ is never permitted in these environments. () Ο μη / *δεν o mi / *dhεn the N / *N ‘the non-smoker’

καπνιστής kapnistís smoker

() Ο μη / *δεν o mi / *dhεn the N / *N ‘the unintelligible’

κατανοητός katanoitós intelligible

TABLE . The distribution of N and N in Standard Modern Greek ENVIRONMENT Unembedded assertions Indicative complement of veridical verbs Prohibition (anaphoric) Question particle (non-negative) Complementizer selected by verbs of fearing (non-negative) In the complement of nonveridical verbs With nonveridical particles: subjunctive να /na/, exhortative ας /as/ Negation of the non-finite (lexical negation: with nouns, adjectives, participles)

N dhεn √ √

√

N mi(n) √ √ √ √ √ √

Polarity phenomena in Standard Modern Greek



() Μη / *δεν γνωρίζοντας τί άλλο να πει, έφυγε. mi / *dhεn ghnorízontas ti álo na pi, έfighε N /N know.. what other  say. leave.. ‘Not knowing what more to say, s/he left.’ () and () are characterized as lexical negation (Veloudis , Giannakidou ). Our definition of (non)veridicality does not exactly accommodate this function of the Modern Greek N, because (non)veridicality is defined as a property of propositional operators and no such operator can be claimed to be present in a DP or an NP. But this environment can still be taken to qualify as nonveridical, or averidical (Giannakidou ), in a more distant sense: similarly to interrogatives and prohibitives, the environment of a DP or an NP cannot receive a truth value. The characterization of NPs as averidical relates also to Brandtler’s () understanding of the notion of evaluability and its application in accounting for polarity phenomena. Evaluable utterances are the utterances that can receive a truth value, either as true or false, while nonevaluable utterances, such as NPs in our case, cannot. Brandtler () uses this distinction between evaluable and nonevaluable contexts to explain the structural division between different types of clauses in Swedish. This division is not sufficient by itself to account for polarity phenomena, as Brandtler (: ) discusses. Clearly there are evaluable environments, such as negated assertions, that can license polarity items. Yet it does provide support in treating the NP as an environment that can license negative polarity items, such as the Greek N, given that NPs are nonevaluable, an observation that actually goes back to Aristotle’s De interpretatione. . . . περὶ γὰρ σύνθεσιν καὶ διαίρεσίν ἐστι τὸ ψεῦδός τε καὶ τὸ ἀληθές. τὰ μὲν οὖν ὀνόματα αὐτὰ καὶ τὰ ῥήματα ἔοικε τῷ ἄνευ συνθέσεως κὶ διαιρέσεως νοήματι, οἷον τὸ ἄνθρωπος ἢ λευκόν, ὅταν μὴ προστεθῇ τι· οὔτε γὰρ ψεῦδος οὔτε ἀληθές πω. σημεῖον δ’ ἐστὶ τοῦδε·καὶ γὰρ ὁ τραγέλαφος σημαίνει μέν τι, οὔπω δὲ ἀληθὲς ἢ ψεῦδος, ἐὰν μὴ τὸ εἶναι ἢ μὴ εἶναι προστεθῇ ἢ ἁπλῶς ἢ κατὰ χρόνον. (Aristotle, De interpretatione I.a.–). ‘ . . . because truth and falsity depend on combination and separation. Nouns and verbs themselves are like thoughts without combination or separation; like ‘man’ and ‘white’, when nothing is added; they are not yet either true or false. In proof of this, consider the following; even the word ‘goat-stag’ has significance, but there is no truth or falsity about it, unless ‘is’ or ‘is not’ is added, either simply or in some other tense.’ (Translation based on Edghill and Pickard-Cambridge )

In this section the distribution of N and N in Modern Greek has been presented: N μη(ν) /mi(n)/ is marked in terms of nonveridicality, in that it appears only in environments that qualify as nonveridical. Nonevaluable environments, like questions or the NP, are a subcase of the nonveridical in a broader sense. N δεν /dhen/ is unmarked in terms of nonveridicality, being the default negator of



The (non)veridicality theory of polarity

Standard Modern Greek. Yet N and N are in complementary distribution, which means that the presence of N in the environments it appears is obligatory and it cannot be replaced by N. This relates to the syntactic projection of nonveridicality, argued for in chapter , which imposes a sort of nonveridical concord between the Nonveridicality Phrase (NP) and the NP.

. Crosslinguistic analogies Greek is far from being alone in lexicalizing what is described here as a polaritysensitive negator. Most prominently in the expression of prohibition—which is one of the diachronically stable functions of Greek N—there are languages that employ a negator different from the one used in declaratives, according to studies on the typology of prohibitives by Kahrel and van den Berg (), Kahrel (), Hovdhaugen and Mosel () and Xrakovskij (). In the typological study on prohibitives, by van der Auwera and Lejeune (), based on a corpus of  languages, three different types of languages are distinguished regarding the expression of prohibition in the second person singular: Type : The prohibitive uses the verbal construction of the second singular imperative and a sentential negative strategy found in (indicative) declaratives –  out of  languages. Type : The prohibitive uses the verbal construction of the second singular imperative and a sentential negative strategy not found in (indicative) declaratives –  out of  languages. Type : The prohibitive uses a verbal construction other than the second singular positive imperative and a sentential negative strategy found in (indicative) declaratives –  out of  languages. Type : The prohibitive uses a verbal construction other than the second singular positive imperative and a sentential negative strategy not found in (indicative) declaratives –  out of  languages.

Modern Greek is a type  language, along with  other languages, because it lacks true negative imperatives, as was discussed earlier (section ...), and it employs a sentential negative strategy not found in declaratives, namely a distinct negative marker. The relevant map from van der Auwera and Lejeune’s () study (figure .) locates these languages geographically and labels them as having both a special imperative and a special negative marker for prohibition. As can readily be seen, there is no particular geographic restriction on these languages. Zulu is an example of a language typologically unrelated to Greek that manifests a negator particular to prohibition. The negator of declaratives is the prefix a- (d), while the negator of prohibitives is the prefix mus- (b). The examples are from Poulos and Bosch (: ).

Sweden Iceland

Canada

South Korea

United States

North Pacific Ocean

Japan

Russia

Norway

United Kingdom

North Atlantic Ocean

Finland

GermanyPoland Ukraine

Spain

Italy

Turkey

Libya Egypt

Mauritania Mali Niger

Sudan

Venezuela

Ethiopia

Columbia

New Guinea

DR Congo Brazil Peru

Australia

New Zealand

South Pacific Ocean

Bolivia

Chile

Kenya Tanzania

Angola

South Atlantic Ocean

Argentina

FIGURE . Geographical distribution of languages with a distinct negator for prohibition (types  and ) Adapted from van der Auwera and Lejeune ().

Iran Afghanistan Pakistan Saudi India Arabia Iraq

Algeria Mexico

Indonesia

Kazakhstan

France

Namibia Botswana South Africa

Madagascar

Indian Ocean



The (non)veridicality theory of polarity

() a. Shay-a inja! Hit-. dog ‘Hit the dog!’ b. Mus-a uku-shay-a N..- -hit- ‘Do not hit the dog!’ c. U-ya-shay-a -..hit- ‘You hit the dog.’

inja! dog

inja. dog

d. A-wu-shay-i N..--hit-N.. ‘You do not hit the dog.’

inja. dog

Apart from its use in prohibition the second negator in these languages can have other functions as well, the discussion of which is beyond the scope of this study. All their functions, however, center on nonveridicality, in that they appear in environments that are characterized by the pretheoretical notion of the irrealis (van der Auwera, personal communication). Regarding the presence of the Modern Greek N in nonnegative environments, as a sort of expletive negation (see table .), the study of Honda () is relevant, discussing the expression of sentential negation in a sample of fifty-six languages around the world. Although focusing on negation in declaratives, Honda () provides examples from eleven languages (e.g. Salishan, Blackfoot, Cherokee, Mayan languages, New Guinean languages; see Honda : –) in which the expression of sentential negation includes an element also found in other types of sentences—as in questions, exhortatives, imperatives, protasis of conditionals, and sentences expressing future events (Honda : –)—which he describes as sentences referring to unrealized events. Below is an example from Blackfoot, an Algonquian language, in which sentential negation is expressed in a syntactically discontinuous way, through two morphemes: máát- and waiksaawa. The second one is also found in other environments like yes/no questions and is described as the ‘non-affirmative’ (Frantz : , –; , , Honda : –). The examples below are from Frantz (). ()

Máátomáyo'kaawaiksaawa. máát-omá-yo'kaa-waiksaawa N-yet-sleep-.NA ‘They aren’t asleep yet.’

()

Áyo'kááwaiksaawa? á-yo'kaa-waiksaawa -sleep-.NA ‘Are they sleeping?’

Crosslinguistic analogies



Interrogatives are among the environments that qualify as nonveridical according to the Giannakidou definition in () (see also tables . and .), which means that waiksaawa in () and () exhibits a polarity behavior, similarly to that of the Modern Greek N μη(ν) /mi(n)/ (see also Louie  for a negative polarity treatment of such elements in Blackfoot). They both relate to the expression of sentential negation and appear exclusively in nonveridical environments in functions that aren’t always negative. Another one among numerous languages in which the expression of sentential negation can be accounted for by reference to (non)veridicality is Bukiyip, a language of New Guinea.Apart from distinguishing between a negator for the past (wo/wotak) and a negator for the future (kobwi) (see Conrad and Wogiga : , cited by Honda : ), Bukiyip additionally indicates sentential negation through an infix on the verb (-u-), which is necessary for the construction of negative declaratives (). This infix, however, also appears in future assertions () and thus manifests, in our terms, polarity sensitivity, a dependency on nonveridical contexts (examples from Conrad and Wogiga : –). () Nabotik gamu m-a-klupw-ech yesterday well .--wash-C.. ‘Yesterday we washed the things well.’ (positive declarative) () Nabotik wo gamu m-u-klup-ech e yesterday N well .--wash-C.. N ‘Yesterday we didn’t wash things very well.’ (negative declarative) () Kaman ch-u-naki tomorrow .--come ‘Tomorrow they will come.’

(future declarative)

The expression of the future is one more environment that is characterized by nonveridicality (Giannakidou , , Giannakidou and Zwarts ), as the future is a domain of uncertainty and contrasts with the past, which is veridical and reflects speaker commitment on the realization of an event (Giannakidou , Giannakidou and Mari ). All this relates to the discussion in chapter  where I argue for a Nonveridicality projection, in order to account not only for the Greek data synchronically and diachronically, but generally for all crosslinguistic lexicalizations of polarity distinctions regarding either indefinites or mood marking or the expression of negation itself. In Modern Greek this nonveridicality indicator is fused with the expression of sentential negation μη(ν) /mi(n)/. Other languages, like Bukiyip in examples () through (), follow an analytical strategy that teases apart negation from the nonveridicality indicator, making the presence of the latter evident.



The (non)veridicality theory of polarity

Standard negation (in declaratives)

Prohibitives

Possibility conditional protasis

Embedded in the scope of nonveridical predicate

Counterfactual conditional protasis/wishes

Directives and other nonveridical (future and conditionals)

Exhortatives

TABLE . The seven negative markers of Romeyka

Romeyka

Neg1 u(tš)(i)

Neg2 mi

Neg3 miðen

Neg4 xe

Neg5 mutš

Neg6 tšen

Neg7 utšas

Standard Modern Greek

Neg1 ðen

Neg2 mi(n)

Neg1 ðen

Classical Greek

Neg1 u:(k)

Neg2 mε: (rarely Neg1 u:(k))

Neg2 mi(n) Neg2 [for directives]/ mi(n) Neg1 ðen [for conditionals]

(Chatzopoulou and Sitaridou )

One last example of the typological possibilities of realization of sentential negation and the relevance of (non)veridicality is the case of Romeyka Greek, a variety still spoken in north-east Turkey (Sitaridou , a/b), which manifests not two but seven noninterchangeable instances of sentential negation, six of which are particular to specific nonveridical environments. Table . gives an overview of the negators compared to their counterparts in Standard Modern Greek and Attic Greek and examples of the Romeyka uses follow (see Chatzopoulou and Sitaridou ) () Utš eporesa mairepsini. N could. cook.INF ‘I could not cook.’

(declarative)

() Mi tros! N eat. ‘Don’t eat!’

(prohibitive)

() Amiðen paθanis u manθanis. IF.N suffer. N learn. ‘If you don’t suffer, you don’t learn.’ () Esi θelis eɣo you. want. I. ‘You want me not to eat.’

xe N

na 

(possibility conditional) troɣo. eat. (embedded subjunctive)

Beginning the journey



() na mutš ixa xasini ton paran, xar n’  N had. lose. the money now  ‘If I hadn’t wasted the money, I would’ve built a house.’

epina ospit. made.  house (counterfactual)

() as to naksemi tšen na troi. till the morning N  eat. ‘He won’t be eating till the morning.’

(future declarative)

() Alis utšas erte. Alis N. come. ‘Alis should not come.’

(exhortative)

The above examples of negation in Romeyka Greek are an indication not only of the connection of (non)veridicality and negator choice, but also of the link between negator choice and functional projections, such as the projection of Nonveridicality introduced in this study. Furthermore the Romeyka data support the link between negator choice and not only support major functional projections, but also point toward the applicability of the fine-grained CP analysis of Cinque (), further discussed in chapter . We have seen so far that the (non)veridicality theory is necessary in order to explain negator choice in a significant number of languages crosslinguistically with different levels and different types of complexity, most of which are typologically and genetically unrelated to the Greek language. In the final section of this chapter I turn to the Proto-Indo-European origin of the two-negator contrast, as manifested in some of the oldest language relatives.

. Beginning the journey: the oldest language relatives and the first attestations .. Indo-European origins and affinities Three sources can be claimed as explaining the presence of a linguistic phenomenon in a language: (i) inheritance, (ii) contact, and (iii) chance, all three necessarily reflecting language universals. For Greek, the contrast between two complementary negators appears to have been part of its inheritance from Proto-Indo-European. The Proto-Indo-European negators have been reconstructed as *ne and *meH through the comparative method, based on the reflexes of either one or both negators in Sanskrit, Latin, Greek, Hittite, Lithuanian, Old Church Slavonic, and Germanic (Joseph , Fortson ). The particulars of the reconstruction of the Proto-IndoEuropean negators *ne and *meH are beyond the scope of our investigation, but their instantiations in some of the oldest daughter languages (Hittite, Sanskrit, Greek, Old Avestan, Old Persian) and some more recent relevant relatives (Armenian and Albanian) are given in table ..



The (non)veridicality theory of polarity TABLE . Negators in some of the oldest Indo-European languages Hittite Vedic Sanskrit Greek Old Avestan Luwian Old Persian Latin Armenian Old English Albanian Old Church Slavonic Lithuanian

First attestation th c.  nd m.  th c.  nd m.  th c.  th c.  th c.  th c.  th c.  th c.  th c.  th c. 

N natta na u:(k[h]) nôit nawa naiy ne/non č-’ ne nuk/s’ ne nè

N le: ma: me: ma: niš ma: mi mos -

The grammars and sources consulted are the following: for Hittite, Hoffner and Melchert (): –; for Vedic Sanskrit, Joseph : –, Whitney : ; for Old Persian, Meillet : –, Kent : ; for Avestan, Jackson : , Swarup Misra : ; for Luwian, Hawkins : ; for Classical Armenian, Roszko : , Schmitt : . See also Fowler () for all Indo-European languages, as well as Davies () for Anatolian in particular.

We observe that it is the oldest Indo-European languages that exhibit the two-negator contrast—Hittite, Vedic Sanskrit, Old Avestan, and Old Persian—while for the languages that do not preserve this distinction it is plausible to assume that there has been a merger at the expense of the marked negator N, which had a more limited distribution. Whether the reflexes of *ne and *me: in some of these languages are the result of inheritance or areal contact, the distribution of the second negator N in all Indo-European languages that have one can be accounted for though the leading idea of this study: that sentential negation itself can be a polarity item triggered exclusively in nonveridical contexts. The licensing environments of N are not the same across these languages, neither synchronically nor diachronically, but they are always nonveridical in the sense of Giannakidou (). The common core in all Indo-European manifestations of N is its correlation with directive force: in the intersection of all semantic environments in which N in these languages is found lies its use in prohibition. Representative examples of both negators in the Indo-European languages that exhibit such a contrast follow (tables .–.).5 Hittite () natta-an u:k tarnahhun. N it I release.. ‘It was not I who released it.’ (Hoffner and Melchert : ) 5

For more information on each of these languages see Hock and Joseph ().

Beginning the journey



() le: ume:ni. N see.. ‘Let us not see.’ (Hoffner and Melchert : ) () -e-wa nekku Land-/..- N(?) harkan? destroy./.. ‘The land is not destroyed somewhere, is it?’

kuwapikki somewhere

TABLE . Negative particle distribution in Hittite Environments in Hittite N natta

N le:

N(?) nekku

assertions [can negate any prohibitions [with present indicative or imperative], predicate other than the wishes [with optative] imperative], constituent negation, in rhetorical questions

question particle, (contrasts in this use with N in rhetorical questions, in that N is unbiased or has a weaker bias than N)

For the N(?) column, see Hoffner and Melchert (: –); see also Hahn (: –), Friedrich (: ), Josephson (: ), and Eichner (: –).

Vedic Sanskrit () Mama . . . áha švášuro ná á: my . . . yet father.in.law. N hither ‘My father-in-law has not come yet.’ (Dahl : .)

jaga:ma. come..

() Kálayo má: bibhi:tana! Kalis. N fear... ‘O Kalis, do not fear!’ (Dahl : .)

TABLE . Negator distribution in Vedic Sanskrit Environments in Vedic Sanskrit N /na/

N /ma:/

everywhere except for prohibitions with injunctive

prohibitions [with the injunctive usually, but sporadically attested with optative, subjunctive and imperative]*

* See Hoffmann (: –) and Joseph ().



The (non)veridicality theory of polarity

Old Persian () Ima hasiyam naiy duruxtam. This true N false. ‘This is true, not false.’ (DB, –) () Patiyazbayam daiva ma: yadiyaisa. proclaim... daiva N worship. ‘I proclaim, the daivas should not be worshipped.’ (XPh–) TABLE . Environments of negative particles in Old Persian Environments in Old Persian N /naiy/ assertions, [with indicative; subjunctive in future relative clauses, future conditionals and purpose clauses, optative in the apodosis of a future less vivid]

N /ma:/ prohibitions, negative wishes, [with subjunctive, optative, injunctive, and imperative]

Meillet (: –), Kent (: ).

Classical Armenian () Oč kamêr N want... ‘He didn’t want . . . ’ (Schmitt : ) () Mi layk‘ i veray im. N cry.. for me. ‘Don’t cry for me . . . ’ (Schmitt : ) TABLE . Negative particles in Classical Armenian Environments in Classical Armenian N /oč/

N /mi/

assertions, questions [with indicative prohibitions [with the imperative mood], inhibitions and all other functions not performed [with hortative subjunctive] and negative wishes [with the optative] by N] Mann : , Roszko : , Schmitt : , .

Albanian () Nuk e hapni derën. N it.. open. door.the ‘You are not opening the door.’

Beginning the journey



() Mos e hapni derën. N it.. open. door.the ‘Don’t open the door!’ (Tomić : ) () në mos gaboj . . . if N err.SG ‘if I am not mistaken . . . .’ (Joseph ) TABLE . Environments of negative particles in Albanian Environments in Albanian N / nuk/s’/

N /mos/

assertion [with indicative] prohibition, conditional protasis, lexical negation

.. Homeric Greek negators and traces of the Proto-Indo-European negatives Homeric Greek (eighth century ) is the closest surviving ancestor of Classical Greek, the language of the Iliad and the Odyssey, and has been used along with Hittite and Vedic Sanskrit for the reconstruction of the Proto-Indo-European negators *ne and *meH. Although written in a language that is considered artificial and particular to the Greek heroic epos with admixtures from dialects such as Aeolic on an Ionic language base, Homeric Greek was comprehensible to the speakers of the Classical era and being the outcome of a long oral tradition, it can provide reliable data to a significant extent for earlier stages of the Greek language. Most importantly, it is a stage in which the two negators of Classical Greek are already attested, N οὐ(κ) /u:(k)/ and N μή /mε:/. Below are examples of the Homeric Greek negators in their prototypical environments, an assertion for N () and a directive for N (). () Θέτις δ’ οὐ λήθετ’ ἐφετμέων παιδὸς ἑοῦ. Thetis d’ u: lε:thet’ ephetmeo:n paidos heu: Thetis P N forget.. requests child. her. ‘Thetis did not forget the requests of her son.’ (Iliad .– ) (N in an assertion) () ἐξαύδα, μὴ κεῦθε νόῳ. eksauda, mε: keuthe noo:i speak-out... N hide. . mind.  ‘Speak out; hide it not in your mind.’ (Iliad . ) (N in a prohibitive) The origin of N οὐ(κ) /u:(k)/ is unknown, although an etymology has been proposed by Cowgill ()—considered also in Chantraine –, and supported more



The (non)veridicality theory of polarity

recently in Beekes , , as well as Joseph : —that N οὐ(κ) /u:(k)/ comes from a pre-Greek phrase *ne oiu kwid, with the original meaning ‘not ever in my life’ from *ne (Proto-Indo-European N), *oiu (‘life, age’), and *kwid (‘something’), see also van Gelderen (: ). The similarity of the Homeric N μή /mε:/, however, to the Proto-Indo-European N meH is evident; not only in form, but most importantly, in its polarity behavior. All the environments in which Homeric Greek N is found, as summarized in Willmott (, ) (see table .), are nonveridical, while its function as a nonnegative question particle is already attested (). () μή πού τινα δυσμενέων φάσθ’ ἔμμεναι ἀνδρών; mε: pu: tina dysmeneo:n phasth emmenai andro:n? N somehow someone enemy. say.. be. men. ‘Do you say he is an enemy?’ (Odyssey .) (N in a yes/no question) TABLE . Environments of negative particles in Homeric Greek Environments in Homeric Greek N /u:(k[h])/

N /mε:/

statements of obligation, statements of ability, directives, wishes, most conditional assertions, conditional consequents antecedents, purpose clauses Willmott 

In fact, all the environments of Homeric Greek N in table . correspond to assertions, sentences expressing speaker commitment to the truth of the proposition expressed by the sentence uttered, while N μή /mε:/ is clearly marked in terms of nonveridicality; it is the nonveridical allomorph of N and has additional nonnegative, yet nonveridical uses, like its use in introducing yes/no questions; see (). Thus (non)veridicality explains negator selection in Proto-Indo-European, Homeric Greek, and its oldest relative languages, Hittite, Vedic Sanskrit, Old Avestan, and Old Persian. .. The homotics νή /nε:/ and μά /ma/: remnants of (Proto-)Indo-European negators? Possible remnants of the Proto-Indo-European negators in the Greek language can be found in Homeric as well as later stages of Greek in the homotic

Beginning the journey



particles νή /nε:/ and μά /ma/,6 which are particular to oaths and are in free variation.7 () νὴ Δία/ (Pherekydis, Fragmenta a.) θεούς (Perseus Phil., Fragmenta .) νὴ τὸν/τὴν Δία (Aristophanes, Aves ) / Ἡρακλέα (Demosthenes, De corona .) / Ἥραν (Xenophon, Memorabilia ...) nε: Dia/ theu:s  nε: ton/tε:n  Zeus/ gods  the./ ‘by Zeus/ the gods/ Hercules/ Hera’

Dia/ Hε:raklea/ Hε:ra Zeus/ Hercules/ Hera

() μὰ Δία (Theognis, Eleg .) / θεούς (Plato, Symposium c ) μὰ τὸν/τὴν Δία (Plato, Crito .b.) / Ἡρακλέα (Aeschines, In Ctesiphontem .) / Δήμητρα (Aristophanes, Ranae ) ma Dia/ theu:s  ma ton/tε:n  Zeus/ gods  the./ ‘by Zeus/ the gods/ Hercules/Demeter’

Dia/Hε:raklea/Dε:mε:tra Zeus/Hercules/Demeter

The homotic particles tend to cooccur with negation () or affirmation (), usually as parentheticals (Fletcher ). The Proto-Indo-European negative particle νή /nε:/ relates to the *n̥- prefix which turned in Greek into the alpha privative ἀ(ν)- /a(n)-/ (Moorhouse : –), e.g., ἀνέλπιστος8/anelpistos/ ‘beyond expectation/hope’ (/a(n)-/ + /elpizo:/ ‘I hope’ + /-tos/ deverbal suffix creating adjectives), but previous forms of the alpha privative as νή /nε:/ are also attested in lexical composition, e.g., νηλεής9 /nε:leε:s/ ‘pitiless’ from /nε:/ + /eleos/ ‘mercy’ (Liddell, Scott, and Jones ), νήπιος10 /nε:pios/ ‘young, infantile’ from /nε:/ + /epos/ ‘word’ (Chantraine ), νῆστις11 /nε:stis/ ‘not eating’, from /nε:/ + ἐσθίω /esthio:/ ‘I eat’. Although the etymological relation of the homotics νή /nε:/ and μά /ma/ to the Proto-Indo-European negators is speculative, what is relevant for my purposes is that of the two homotic particles νή /nε:/ and μά /ma/, only the second one will make its way to the subsequent stages of the language (), persisting right into Standard Modern Greek (). Similarly to the Greek N μή /mε:/, the homotic μά /ma/ survived until after the Hellenistic times, being used interchangeably with νή /ni/ (see , ), but is extinct in low-register vernacular Greek by the Early Medieval stage. νή /ni/ does not survive in Modern Greek, but μά /ma/ is quite frequent. Only the deities have changed.

See Apollonius Dyscolus, De constructione .... Although Dionysius Thrax in Ars grammatical (...) assigns to them a specialization as νή /nε:/ occuring with affirmation (κατωμοτικόν) and μά /ma/ with negation (ἀπωμοτικόν) (see also Householder : ); this, however, is not confirmed by my findings. 8 9 Sophocles, Trachiniae . Iliad .. 10 11 Odyssea .. Iliad .. 6 7



The (non)veridicality theory of polarity μὰ τὸν Δία, ὦ Σώκρατες ma ton Dia o: So:krates u: Neg hom the Zeus, voc Socrates ‘No, in the name of Zeus, dear Socrates’ (Plato, Crito 43.b.3)

(88) οὐ

(89) ναὶ

μὰ Δία Dia nai ma yes hom Zeus ‘Yes, in the name of Zeus’ (Aristophanes, Pax 416)

(90) νὴ Δία nε: Dia hom Zeus ‘by Zeus’ (Plutarch, Solon 6.2.4) (91) μὰ Δία ma Dia hom Zeus ‘by Zeus’ (Plutarch, Solon 28.1.6)

(fifth century bc)

(first century ad)

τον άγιον Δημήτρη (fourteenth century  Greek) () Μά ma ton ághion Dimitri  the saint Demetrius ‘In the name of Saint Demetrius’ (Ptocholeon ) () Μα το ma to  the ‘By God’

Θεό. Thεό God

(Standard Modern Greek)

. Summary In this chapter I have presented the central theory of this study, along with the relevant evidence from Standard Modern Greek and argued that negator choice in Modern Greek must be understood as a polarity phenomenon. It was shown that N μη(ν) /mi(n)/ is a polarity item licensed in nonveridical contexts. It is for this reason that N cooccurs with subjunctives, questions, and other nonveridical elements. Furthermore, crosslinguistic extensions were identified of the idea that negator choice depends on nonveridicality to a number of Indoeuropean (Vedic, Hittite, Armenian, Albanian), as well as typologically unrelated languages, such as Zulu and Algonquian. In the following four chapters the particulars of negator selection in previous stages of the Greek language are presented, beginning from Classical Greek: the language spoken in Athens during the fifth and fourth centuries .

3 Negation, mood, and (non)veridicality in Classical Greek (fifth to fourth centuries ) In this chapter the relevant evidence from Classical Greek is presented and the regulating factor of (non)veridicality in negator choice is revealed. Classical Greek is the Attic dialect, the language spoken in fifth–fourth-century  Athens, which is included in the ancestral lineage of Standard Modern Greek, as opposed to other Ancient Greek dialects like Doric, Aeolic, and Macedonian. Attic Greek is in a sense the joy of the historical linguist. The abundance of texts preserved and digitized, combined with the absence of diglossia, make it ideal for the purposes of this study. At the same time Attic Greek is the root of all that is prescriptive. The prestige of fifth–fourth-century  Athens and the valued content of many texts in this stage of the language, in rhetorics, poetics, philosophy, ethics, and metaphysics, served to turn the Attic dialect into “Great Attic,” a literary standard and norm for the learned in all subsequent centuries.1 The two kinds of negation that Attic Greek exhibits are N οὐ(κ) /u:(k)/ and N μή /mε:/. In the present chapter, their complementary distribution is reduced to a polarity phenomenon (Chatzopoulou and Giannakidou , Chatzopoulou a/ b/c, a/b, ), a sensitivity to (non)veridicality (Giannakidou  et seq.) in the functional domain. N οὐ(κ) /u:(k)/ is the unmarked negator in terms of (non) veridicality; it appears in both veridical and nonveridical semantic contexts, while N μή /mε:/ is marked: it appears in nonveridical environments exclusively. The relation of the Attic negative particles to mood is discussed, moving toward the conclusion that although both negation and mood exhibit polarity sensitivity, negator choice and mood choice in Attic Greek are basically independent. Evidence is provided that syntactically both negators are phrasal. However, N οὐ(κ) /u:(k)/2

1

Denniston (), Dover () among others; more recently Vessela (), Horrocks (). The allomorph οὐ /u:/ appears if the following word begins with a consonant, οὐκ /u:k/ appears if the following word begins with a vowel, while οὐ(χ) /u:(kh)/ is used when the following word begins with [h], 2

Negation and Nonveridicality in the History of Greek. First edition. Katerina Chatzopoulou. © Katerina Chatzopoulou . First published  by Oxford University Press.



Negation, mood, and (non)veridicality

behaves as the prototypical negator: it is the standard negation in terms of Payne (), the negator of declaratives, while μή /mέ:/ has additional nonnegative functions and its representation as a morphological constellation of independent yet diachronically and synchronically related functions is already justified (cf. Joseph and Janda  for Modern Greek, section ..). This chapter is structured as follows. In section . basic facts about the Attic dialect are introduced, regarding verbal morphology (grammatical information on verbs, infinitives, and participles) and word order. Section . includes argumentation on the phrasal status of the Attic negators and their inconsistent relation to mood is exposed. The regulating role of (non)veridicality in negator selection is presented in section .. Section . provides an overview of the nonnegative functions of N μή /mε:/, as a particle introducing yes/no questions, a complementizer, and an attitudinal marker. Section . extends the polarity behavior of the Attic Greek negators to the morphologically negative paradigms of Attic indefinites: N-words (e.g. οὐδείς /u:de:s/ ‘no one’, οὐδέν /u:den/ ‘nothing’, οὐδέποτε /u:depote/ ‘never’) and N-words (e.g. μηδείς /mε:de:s/ ‘no one’, μηδέν /mε:den/ ‘nothing’, μηδέποτε /mε:depote/ ‘never’). Nwords are exclusively licensed in nonveridical environments. Section . summarizes the chapter.

. The Attic dialect Attic Greek was the most influential of the Ancient Greek dialects. It was spoken mainly in the region of Athens during the fifth and fourth centuries  (and for some texts written in the third century ). This section provides basic information on the structure of the particular dialect that relates to our discussion. .. Morphology of verbs and verbals Attic Greek morphologically distinguishes seven tense/ aspect combinations, denoted as ‘tense’ in traditional grammars: present, imperfect, aorist, future, perfect, pluperfect, and future perfect; four moods: indicative, subjunctive, optative, and imperative (mood is usually fused with tense, in that a single morpheme can indicate both tense and mood; see table .); three voices: active, passive, medio-passive. With some variation in terminology, these descriptions are given in all traditional grammars of Classical Greek (Goodwin , Jannaris , Kühner and Gerth , Tzartzanos , Rijksbaron ).

transcribed in the Greek script as a rough breathing on the initial segment—usually vowel—of the following word, e. g. oὐχ ὁρῶ /u:kh horo:/ ‘I don’t see’.



The Attic dialect

TABLE . Paradigm of active third person singular of λύω (/ly:o:/, ‘set free’) in Attic Greek Indicative Subjunctive

Optative Imperative

Participle

Infinitive

Present

λύει /ly:e:/

λύῃ /lyε:i/

λύοι /ly:oi/

λυέτω /ly:éto:/

λύων.masc /ly:o:n/ λύουσα.fem /ly:u:sa/ λύον.neut /ly:on/

λύειν /ly:e:n/

Imperfect

ἔλυε /ély:e/

-

-

-

-

-

Future

λύσει /ly:se:/

-

λύσοι /ly:soi/

-

λύσων.masc /ly:so:n/ λύσουσα.fem /ly:su:sa/ λύσον.neut /ly:son/

λύσειν /ly:se:n/

Aorist

ἔλυσε /ély:se/

λύσῃ /ly:sε:i/

λύσαι/-ειε /ly:sai/ /ly:se:e/

λυσάτω /ly:sáto:/

λύσας.masc /ly:sas/ λύσασα.fem /ly:sasa/ λύσαν.neut /ly:san/

λῦσαι /ly:sε/

Perfect

λέλυκε /lély:ke/

λελυκώς ᾖ /lely:ko:s ε:i/

λελυκὼς εἴη /lely:kó:s é:ε:/

λελυκὼς ἔστω /lely:kós ésto:/

Pluperfect

ἐλελύκει /elely:ke:/

-

-

-

-

-

Future Perfect (rare)

λελυκὼς ἔσται /lely:kó:s éstai/

-

λελυκὼς ἔσοιτο /lely:kó:s ésoito/

-

λελυκὼς ἐσόμενος.masc /lely:kó:s esómenos/ λελυκυῖα ἐσομένη.fem /lely:kuí:a esoménε:/ λελυκὸς ἐσόμενον.neut /lely:kós esómenon/

λελυκὼς ἔσεσθαι /lely:kó:s ésesthai/

λελυκώς.masc λελυκέναι /lely:kénai/ /lely:kó:s/ λελυκυῖα.fem /lely:kuí:a/ λελυκός.neut /lely:kós/



Negation, mood, and (non)veridicality

Furthermore, Attic Greek is a language with participles (marked for gender, case, number, voice, tense, and aspect) and infinitives (marked for voice, tense, and aspect) (cf. also Sevdali ). I focus on the expression of negation in all three verbal forms, for which a unified explanation regarding the distribution of the two negators is proposed (section .). The paradigm of a regular verb in active voice is given in table ..3 .. Attic Greek word order Attic Greek word order cannot easily be described in terms of grammatical categories such as subject, object, and verb. There seems to be a tendency for verb types to appear in final position, but the grammatical information present on the nominal categories (i.e., case marking on nouns and adjectives, which additionally have to agree in gender) results in a very liberal word-order pattern. Aristotle’s remark on this issue in his treatise On interpretation sums up the general situation: Μετατιθέμενα δὲ τὰ ὀνόματα καὶ τὰ ῥήματα ταὐτὸν σημαίνει, οἷον ἔστι λευκὸς ἄνθρωπος – ἔστιν ἄνθρωπος λευκός·.4 ‘You can transpose the subject and predicate, without any change in meaning. Thus we say ‘man is white’ or ‘white is man’.

This mobility carries over to most grammatical categories, for which there does not appear to be a fixed order, neither for the relative position of dependent and independent clauses, nor for the order of grammatical categories within each (cf. Denniston , Dover ). A pragmatic approach on Ancient Greek word order has been more fruitful in the spirit of Weil (), recently applied by Dik (, ). Dik () gives a pragmatic account of word order in the language of Herodotus, which was followed by a study on the word order of Attic Greek tragic dialogue (Dik ). Her work is based on information structure, in the view also of Chafe (, ), and the notions of Topic and Focus, both observed to precede the verbal category, in a basic syntactic template Topic–Focus–Verb. To the extent that crosslinguistically subjects tend to be topics, this is immediately translatable to SXV word order (conventionally described as SOV). SOV is claimed by Devine and Stephens () to be the underived Attic word order. On the other hand MeierBrügger (a, b) supports the Mittelstellung of the Ancient Greek verb following the line introduced by Kiecker (). Devine and Stephens () revise their earlier position tending toward a characterization of Ancient Greek as a non-configurational language, a claim already made by Cervin () in his dissertation “Word order in

3

For Attic Greek phonology cf. Allen (), Devine and Stephens (), and more recently Probert (). 4 Aristotle, De interpretatione b –.

The Attic dialect



ancient Greek: VSO, SVO, SOV, or all of the above?”5 A non-configurational treatment of Attic Greek may seem plausible, along the lines of Kiss (). Attic negators, however, have a fixed relative position: they always have the category they negate—overt or implied, in case of ellipsis (section .)—in their syntactic scope. () oὐ πείθομαι, ἐπεὶ δικαστὰς οὐχ ὁρῶ u: pe:thomai epe: dikastas u:kh horo: N obey... because judges. N see... καθημένους. kathε:menu:s sit... ‘I don’t believe you; for I do not see the Dicasts sitting.’ (Aristophanes, Nubulae –) () μὴ προδοῦναι λίσσεται. mε: produ:nai lissetai N betray.. beg... ‘He begs (her) not to abandon (him).’ (Euripides, Alcestes ) This fact, combined with the order in which grammatical information appears on the nonirregular verb of Attic, could enable us to propose, in the spirit of Pollock (), the following order of projections as reflecting the vertical structure of Attic Greek (cf. Rivero  for this approach to Modern Greek): ()

οὐκ u:k

οὐκ Neg1

NegP F*Move > F*Merge > F

The Late Merge Principle suggests a shift from F*Move to F*Merge, which is considered to be more economical and less marked. For diachronic change this implies a view of grammaticalization as a tendency for elements to go ‘up-the-tree’, reanalyzing in an ‘upward’ fashion. The Hellenistic ἵνα /ína/ exemplifies such an instance of Late Merge or upward lexical micromovement in our terminology, if we assume that the Nonveridicality projection is originally higher than the ἵνα /ína/ particle, because nonveridicality is introduced by the category that selects for the ἵνα /ína/ clause (e.g., ability modals, directive force, question operator). The structures in () reflect the transition in the function of INA from complementizer to nonveridical marker in the [clause][clause] structure in which it appeared in Attic Greek () and the [clause [SUBJ clause]] developing during Hellenistic and Roman times (). In () the INA clause is unselected, see (a). In () it is selected as an internal argument of the verb κέλευσον /kéleuson/ ‘order/request’ (aorist imperative), see (b), a verb that could appear either as transitive or as intransitive with an elided object both in Classical and in Hellenistic Greek. Example () gives an instance of the verb κέλευσον /kéleuson/ in Attic Greek as transitive with a final infinitive, which was its canonical complementation in Attic (for the infinitives of Classical Greek within a generative perspective and the accusativus cum infinitivo in particular see Sevdali ). () αὖθις κέλευσον, ἵνα σαφῶς μάθωσί authis keleuson hina sapho:s matho:si again order... so.that clearly understand... σου. (Attic Greek) su: you.. ‘Repeat the order, so that they may clearly understand you.’ (Euripides, Helena ; translation by Coleridge ()) [clause][clause] () κέλευσον kéleuson order... δρυμοῦ dhrymú forest. ‘Order that beasts .–)

ἵνα ína INA

ἐξέλθωσιν eksélthosin come.out

θηρία thiría beast..

ἐκ ek from

τοῦ tu the. (Koine Greek)

come out of the forest . . . ’ (Testamentum Abrahae [clause [INA clause]]



Developments in Hellenistic-Roman times

() κέλευσον ἐλθεῖν ὡς ἐμὲ τὸν keleuson elthi:n o:s eme ton order... come.. to me the. Ἑρμογένην (Attic Greek) Hermogenε:n Hermogenes. ‘Tell Hermogenes to come to me.’ (Xenophon, Memorabilia ...) (clause [accusativus cum infinitivo]) ()

a.

(Attic Greek: INA clause as VP adjunct)

CP VP VP VP

CP

authis keleuson again order.imp.2sg

b.

su: hina sapho:s matho:si so.that clearly understand.subj.3pl you.gen (Koine Greek: INA clause as V complement)

CP VP VP V

kéleuson order.imp.2sg

CP

ína ekséltosin thiría ek tu dhrymú to come.out beasts.nom from the forest

The effect of structure simplification—if we take clause-union to qualify as such—is apparently a violation of syntactic inertia, but given that this takes place owing to external pressures—namely the homophony in the mood system, as was argued in section ..—this is in accordance with the predictions of Keenan (, ), elaborated and expanded in Longobardi’s () Inertial Theory with the main axiom that “syntactic change should not arise, unless it can be shown to be caused” (Longobardi : , but see Waltereit and Detges  for critical discussion and counterexamples, e.g., syntactic change driven by syntactic factors).

Changes in the mood and complementation system



In the final part of chapter  I propose that what is in fact inert is the semantics itself and that inertia in syntax can be a mere reflection of that—to the extent that it is—given that the communicational and expressional needs of the speakers appear to have been the same (at least in historical times). And there are conspiratorial forces in language that make sure these needs are fulfilled under all conditions crosslinguistically and diachronically. It is these forces that induced the semantic broadening of INA in Hellenistic Greek into a marker of nonveridicality, because qualifying reality and speaker commitment is important in human language and pragmatically driven (cf. the maxim of quality/truthfulness in Grice ). .. The Nonveridicality projection (NVP) Our assumption is that nonveridical operators, such as the directive verb κέλευσον /kéleuson/, introduce the Nonveridicality projection into syntax and it is to this projection that both INA and the Greek N are linked. The vanishing structure in figure . offers a visualization of the transition of INA from purpose complementizer (unselected) in Attic Greek to nonveridical marker (selected) in Koine Greek. This representation aims to capture the notion of clause-union and the relation of selection that develops between the matrix verb, e.g. the directive κέλευσον /kéleuson/, and the INA phrase in Hellenistic times. An embedded C can still be assumed to project even in Koine Greek and after, depending on the needs of the phenomenon we want to represent. For our present purposes, this level of abstraction is exactly where the upward reanalysis of INA is evident. This direction is reminiscent of the AFFECTcriterion in Haegeman (), but in a way that is inclusive to all nonveridical environments and remaining agnostic for now on whether or not locality restrictions apply. There does appear to be a tendency for NPI licensing in c-commanding

VP V’ V

NonVerP NonVer’ NonVer

CP C’ C

KOINE GREEK: kéleuson ATTIC GREEK:

kéleuson

ína ína

FIGURE . Cross-clausal grammaticalization: INA from Attic to Koine

TP



Developments in Hellenistic-Roman times

positions of the nonveridical head based on the Greek data. However, crosslinguistic and language structure considerations limit us at this point from making any stronger claims on locality (see Giannakidou and Zeijlstra  for discussion and references therein). Thus in the transition from Attic to Koine Greek the INA particle has resettled into a higher position outside of its original clause and in this sense it manifests an instance of Late Merge. We take it that the Nonveridicality projection was always there, even in Attic Greek, and nonveridical operators such as the verb κέλευσον /kéleuson/ ‘order’ introduced the Nonveridicality phrase even in the case of final infinitives, see (), where the notion of mood is irrelevant. This explains the attestation of N μή /mε:/ with final infinitives that was discussed in chapter . In Hellenistic times the Nonveridicality projection gets lexicalized through the INA particle. The structure in figure . can account for the rest of the nonveridical operators, the question operator, the imperative, the conditional, modals, etc.: they introduce the Nonveridicality projection that can host Koine Greek INA or the modal ἄν /an/ of Classical and Koine Greek (see chapter ) or is introduced by the conditional particles εἰ /i/, ἄν /an/, ἐάν /eán/, and at the same time it correlates with the Greek N. Evidence for this projection in the case of final infinitives (infinitives selected by nonveridical predicates) are examples found in papyri of Late Koine Greek in which we find INA preceding an infinitival form. These intermediate cases of INA plus infinitive are also discussed in Ljungvik () and Burguière (), cited in Joseph ( []: ). () ᾔτησέν μοι πρᾶξαι αὐτῇ ἵνα πέμψαι αὐτοὺς . . . ítisén mi prákse autí ína pémpse autús ask... me. make.. her. INA send.. them. ‘He asked me to make her (so as to) send them’ (P.Oxy. . (– c) f; translation by Joseph ( [])) The Nonveridicality projection is introduced in this case by the nonveridical predicate πρᾶξαι /prákse/ ‘to make/force’. The Nonveridicality projection relates to, but cannot be identified with, morphological mood and nonindicative mood in particular. Examples were provided from Modern Greek and Classical Greek (see chapters  and ) and will follow for the subsequent stages, in which the N particle, as well as the Modern Greek να /na/ particle, can cooccur with morphological indicative (viewed as the mood of unembedded assertions) in its interrogative/dubitative or strong volitional uses. In (a) the finite form, glossed as perfective past, can cooccur with either N or the να /na/ particle with an interrogative/dubitative function and it is indistinguishable from the finite form of the unembedded assertion in (b). Similarly the finite forms of (a), (c), and (d), this time in imperfective past, are indistinguishable from the finite form in the unembedded assertion (b).

Changes in the mood and complementation system () a. Μην/Να έγραψε; min/na έghrapsε? N/ write.. ‘Has s/he written (I wonder)?’

b. Η Άννα έγραψε. i Ana éghrapse the Anna write.. ‘Anna wrote.’

() a. Να έγραφε. na éghrafe  write.. ‘S/he should have written.’ (counterfactual) ‘If only s/he would write.’ (strong volitional)

b. Η Άννα έγραφε. i Ana éghrafe the Anna write.. ‘Anna was writing.’

c. Να μην έγραφε. na min éghrafe  N write.. ‘S/he should not have written.’ (counterfactual)



d. Να έγραφε; na éghrafe?  write.. ‘Should s/he have written?’ ‘Do you think s/he was writing?’

It is these observations that have led to the positing of a zero indicative morpheme (Giannakidou , following Philippaki-Warburton , ) that would account for (b) and (b) and the claim that the finite forms of Modern Greek basically bear only tense and aspectual marking (Holton, Mackridge, and PhilippakiWarburton ), an assumption I am implicitly following by using for the Modern Greek finite forms glosses that refer only to their temporal and aspectual marking, e.g., perfective past. Yet motivation from the history of the language and from other phenomena such as negation and evidentiality within Greek and other languages strongly implies the presence of a projection distinct from morphological mood. The complementizer ὅτι /óti/, discussed in section ..., seems to point in the same direction, if viewed as an evidential that can have reportative or inferential functions that qualify speaker commitment to whatever follows. I propose that the difference in use and meaning between ὅτι /óti/ and INA rests in the lexical semantics of the two elements and not their structural position, as both can be claimed to relate to the Nonveridicality projection during the Koine Greek linguistic stage. The ὅτι /óti/ particle would later on develop into a purely assertive complementizer (see chapter ), but in this use of Hellenistic Koine it merely functions as a particle that indicates direct speech, indirect speech, or causality (see Sim : ). All three uses have been linked to evidentiality in one way or another (Izvorski , Faller ). Thus, whether directivity or modality can also be traced and although aspectual and temporal considerations are relevant as well, I argue that it would be meaningful to posit a syntactic projection that can account for all the functions of INA and all the manifestations of N in the history of Greek. The only notion that can unite all the environments of both elements, as well as those of the να /na/ particle of



Developments in Hellenistic-Roman times

Late Medieval and Modern Greek, is the notion of nonveridicality. Nonveridicality regulates negator selection in languages that have to choose among multiple negators (see chapter ), mood selection, and the distribution of other polarity items, like negative polarity indefinites; and relates to modals and evidentials crosslinguistically. Negation itself is one kind of nonveridical marking, officially introduced as a syntactic projection in Pollock (), followed by Ouhalla () and Zanuttini (). Besides, given that our ultimate goal is the unambiguous representation of sentential meaning, in a way that captures as many fine distinctions as possible, some room can be found in the syntactic representations for a notion that gets so heavily grammaticalized crosslinguistically and in the history of Greek in particular, which is the focus of this study. Overall, the viewpoint I propose offers a syntactic treatment of the notion of ‘licensing’ that has been discussed in connection with polarity phenomena since Ladusaw (). Syntactic analyses regarding the appearance or ‘triggering’ of negative-polarity items—as are the Greek N and the Modern Greek να /na/ particle—have already been proposed in Linebarger (, ) and Progovac (, ). In this study the notion of the ‘c-commanding trigger’ of negative polarity items is unified into one sole projection, the Nonveridicality projection, that may overlap e.g. with negation or mood, but can also be independent. Further motivation for a syntactic handling, through the notion of agreement (as in Chomsky ), of a traditionally semantic phenomenon comes from the diachronic development of the Modern Greek perfective non-past (PNP), as will be shown in section ... .. Precursors and differentiation: Cinque (), Culicover (), Laka () The expanded CP analysis known as the ‘cartographic approach’ introduced in Rizzi () and further developed in Cinque () relates strongly to what I am proposing and it has been discussed in connection with the Homeric and Classical Greek N in Willmott (). However, empirical evidence indicates that the Nonveridicality projection I am arguing for is not only independent from mood, but it is also independent from C. N and INA are attested in a significant variety of contexts, among which are also the infinitives and in nominals/nominalizations. Even if we accept Sevdali’s () analysis of the Ancient Greek infinitive as projecting its own CP, the presence of both N and INA with nominals and even DP internal (in the case of Late Medieval and Modern Greek να /na/) strongly implies that Nonveridicality can project even in the absence of a CP. This fact also differentiates the NVP I am arguing for from the Polarity phrase (PolP) introduced in Culicover (, more recently Poletto and Zanuttini , De Clercq , ), which appears nevertheless to have been motivated to account for similar phenomena (see Zanuttini b, Progovac , Breitbarth and Haegeman , Haegeman and Lohndal , van Gelderen b, , Biberauer ,  among others). The Polarity phrase is taken to project below C and indicate whether a sentence is positive or negative. A functional projection of a similar nature, the Sigma

Changes in the mood and complementation system



phrase (ΣP) is in a sense more relevant. The Sigma phrase (ΣP) was introduced earlier in Laka () within her discussion of negation in English, Spanish, and Basque, in order to syntactically host, not only affirmation and negation (especially when correlating with emphasis), but also n-words—a term coined by Laka—for which she proposed a unified account as Negative Polarity Items (Laka : ). According to the Giannakidou ( et seq.) theory, such items are dependent on nonveridicality, as a semantic property of their licensing context. The Nonveridicality projection I propose is an actual polarity phrase (not in the literal, but in the current conventional sense): it does not simply indicate affirmation or negation, but rather marks an environment as nonveridical, an environment that does not add to the common ground. As a result, I argue that in the same way that negation is marked compared to affirmation (Horn , Speranza and Horn ) and a NP projects only in the presence of negation, the NVP projects only in nonveridical environments and is not a projection that can receive two values (+ or ), as opposed to the PolP. In others words, we take that there is no VP counterpart in the syntax—or as a matter of fact, in the semantics—by assuming that veridicality is the default. I ground this primarily on our data (nonveridical environments appear to be morphologically marked) and secondarily on the Gricean quality maxim, the maxim of truthfulness (since Grice ), that truth commitment from the part of the speaker is among the expectations of the hearer according to the Principle of Cooperativity: if an apparent falsity is spoken unqualified and unflouted, the listener has the right to complain. The way, however, nonveridicality gets grammaticalized can be language-particular. In terms of the overlapping projections that were introduced in our pretheoretical representations of INA reanalysis (figures . and .), nonveridicality can be parasitic on mood or parasitic on C (cf. Matthewson, Davis, and Rullmann , for a similar treatment of evidentiality), but it can also be independent. As a result of Late Merge, then, ἵνα /ína/ in Late Hellenistic Greek begins to lexicalize the Nonveridicality projection in an analytical and isomorphic way by expanding its functions throughout the Nonveridical Space (fig. .) at the heart of which lies negation. As is further discussed in chapter , our main claim is that the Greek N particle, similarly to the N-words of Attic Greek, has an uninterpretable nonveridicality feature [uNV] and it can appear only in environments where nonveridicality projects, that is in the environments that are considered to be nonveridical from the viewpoint of the logical subject. .. Feature economy: what happens to the finite verb after the reanalysis of INA One final point that remains to be discussed regarding the INA+subjunctive mood reanalysis is what happens to the finite verb form after the broadening of the INA particle. This can be explained in terms of feature economy (van Gelderen , b), but from the perspective of the verb form. Example (a) is repeated in () below.



Developments in Hellenistic-Roman times

() πρῶτόν ἐστιν, ἵνα ἐγὼ κοιμηθῶ. proton estin, ína eghó kimithó first is INA I sleep... ‘The first thing is that I should sleep.’ (Epictetus, Dissertationes ab Arriano digestae ...; translation by Sim (: )) The formal description I provided was that the uninterpretable [uS] feature carried by the Attic Greek ἵνα /hina/ complementizer, in Hellenistic Koine gradually becomes interpretable [iS], both because of the phonological weakening of the contrast between the subjunctive and the indicative and because of the diachronic tendency of upward featural micromovement. From the viewpoint of the finite verb form, as is the form κοιμηθῶ /kimithó/ in example (), the interpretable [iS] feature it used to bear now becomes uninterpretable [uS] and it needs to be in the scope of a subjunctive assigning element, or in other words, a nonveridical operator—given that we have now established a link between the subjunctive and the Nonveridicality projection—in order to surface. This is immediately explained through Feature economy as formulated in van Gelderen (b), mainly to account for transformations like those from prepositions to complementizers (as in the English for) and from verbs to auxiliaries (as in the English will) among other phenomena of grammaticalization. The definition below is from van Gelderen (b: ) (see also van Gelderen b: ). () Feature Economy Minimize the semantic and interpretable features in the derivation: semantic > [iF] > [uF] The transformation that is relevant for the finite remnant of the INA reanalysis is the one from [iF] to [uF] which sheds light on the polarity behavior of perfective nonpast (PNP) finite verb forms in Early Medieval and surviving in Modern Greek (as discussed in ..). Below is example (b) of perfective non-past in Modern Greek repeated in (). I take that the DP is an environment in which nonveridicality projects by default. () b. Το πιο σημαντικό είναι (το) να κοιμηθούμε. to pjo simandikó íne (to) na kimithúme the most important is (the) NA sleep.. ‘The most important thing is for us to sleep.’ The perfective non-past forms have their origin in the aorist subjunctive of classical and Hellenistic times. Mood indication was weakened in Hellenistic times, but all finite forms—and up to a certain stage nonfinite forms as well—would continue to bear aspectual as well as temporal information. As a result, the transformation of the aoristic subjunctive into the perfective non-past, from an overt indicator of nonveridicality, one instance of which is the subjunctive mood, into a polarity item, an

Change in word order: from OV to VO



element that needs to be in an agreement relationship with nonveridicality, follows from the diachronic tendencies of language change within a generative perspective. Thus the exact change that happens to the aorist subjunctive remnant, according to the viewpoint I introduce, is one from [iNV] to [uNV]. This explains the full distribution of the perfective non-past presented in chapter , given that the perfective non-past cooccurs not only with the Modern Greek subjunctive particle να /na/, but also with the exhortative ας /as/, the conditional particle αν /an/, and with N as well (see section ..). We have seen so far how the transformations in the language system during Hellenistic times had the effect of the isomorphic emergence of nonveridicality in syntax. It is with reference to the Nonveridicality projection that negator selection in Greek can be syntactically accounted for. In the remaining part of this chapter one more change that relates to the diachronic development of sentential negation in Greek is discussed, the change in underived word order and the statistical evidence on the distribution of N and N in Koine Greek follow.

. Change in word order: from OV to VO One final development that takes place in the transition from Classical Greek to Hellenistic Koine that is presented here, as it relates to our discussion, is the shift in underlying word order: from SOV to VSO. Although the configurationality of Attic Greek is rather problematic, as was discussed in chapter , the standard, underived tendency of word order is for the verb to be in sentence final position, according to a number of studies (Ebeling , Joseph  [], Devine and Stephens , Taylor , Dik , ). In Hellenistic times there is a major shift towards verb initial patterns (Blass and Debrunner : , Lightfoot : –, Horrocks , Taylor , Aitchison , () : –, –, Haug , Horrocks : –, Deligianni a), although word order in general remains very flexible and for the most part driven by pragmatic factors (see also Karali’s  study on Greek word order in epigraphy from the fifth century  to the second century ). Given that Greek is diachronically a pro-drop language, this shift can be represented as one from OV to VO and from head final to head initial structures. SVO structures are still available and frequent, as is generally the case in languages with VSO underived word order (which is the sixth universal of Greenberg, see Greenberg : , also Friberg : –). () οἱ . . . σώφρονες οὐδὲν δήπου ἄδικον οὐδὲ ἀνόσιον hoi so:phrones u:den dε:pu: adikon u:de anosion the prudent N-thing thus injust. neither unrighteous.



Developments in Hellenistic-Roman times ποιοῦσιν poiu:sin do... ‘ . . . right-minded men commit no unjust or unrighteous deed.’ (Xenophon, Hellenica ...–; translation from Brownson ( and )) (Attic Greek: OV)

() οἱ μαθηταί σου ποιοῦσιν ὃ i mathité su piúsin o the students your do... that ποιεῖν ἐν σαββάτῳ. piín en sabháto do. on Saturday ‘ . . . your disciples are doing what is unlawful to Testamentum, Secundum Mattheum ..–; ())

οὐκ uk N

ἔξεστιν éksestin allow...

do on the sabbath.’ (Novum translation by Senior et al. (Koine Greek: VO)

This change in the unmarked word order may correlate with the Greek Jespersen’s Cycle processes discussed in chapter , where the novel approach for Jespersen’s Cycle is presented, as a phenomenon that involves the transformation of predicate negation into propositional. For Greek the predicate negation that underwent Jespersen’s Cycle was the negative indefinite οὐδέν /udhén/ N-thing, which originated as an object, a verbal complement, in Classical Greek. I suggest that the change in word order from SOV to VSO in Hellenistic Greek facilitated the reanalysis of οὐδέν /udhén/ as a nonargument of the verb at some point during the cycles of acquisition. The arguments of the verb, both external and internal, are now anticipated to its right. If οὐδέν /udhén/ occurred to the left of the verb in the input of the language learner, then it had to be something other than an argument and was thus reanalyzed as an element of a projection higher than the VP, either a VP adjunct and therefore an adverb (with a meaning of notat-all, cf. Arndt and Gingrich :  for μηδέν /midhén/,  for οὐδέν / udhén/ with negative adverb meaning), or eventually propositional negation itself (figure .). A correlation between word order change and Jespersen’s Cycle in general has already been made since Vennemann (). Vennemann claims that it is the shift in word order that triggers the negative cycle. His argumentation and predictions, however, do not apply to the Greek data, for reasons that are discussed in detail in chapter . Jespersen’s Cycle and its exact manifestation in a language seem to arise in the context of multiple parameters. One of them may be word order change, not as the sole or main reason, but rather in the sense of Schwegler (), as a pathway for change (cf. also the change from OV to VO in Middle English, Lightfoot : –, Bech ).

Change in word order: from OV to VO



(Classical Greek: OV)

NegP

Neg’

VP

Neg Ø

NP

V

[uNeg] οὐδέν /u:den/ Neg1-thing ‘I saw nothing.’

εἶδον /e:don/ see.aor.ind.1sg

NegP

(Koine Greek: VO)

Neg’

Neg

VP

V

οὐδέν /udhén/ Neg1

(NP)

εἶδον /ídhon/ see.aor.ind.1sg

‘I didn’t see.’

FIGURE . N-thing reanalysis from Classical Greek to Koine Greek

This shift during Koine Greek toward head initial structures is supported also by evidence from prepositional phrases, whose frequency rises at the expense of oblique case marking and in particular at the expense of the former dative (cf. Young : ). This is one more instance of preference for analytical structures. Moreover, apart from their increase in frequency, prepositional phrases now establish a fixed internal structure: they are always head initial and postposing is no longer available, at least in the low-register texts (New Testament, Epictetus). In Classical Greek both orders were possible with a number of prepositions, although the preference towards



Developments in Hellenistic-Roman times

preposing is already evident. There is variation in the selection of case for the prepositional complement and a phonological effect of stress shift, from /perí/ to /péri/, in postposing in the examples () and () below, but the basic meaning of the preposition is the same. () τυφλοῦται γὰρ περὶ τὸ φιλούμενον typhlu:tai gar peri to philu:menon blind... P about the love... ὁ φιλῶν. (Attic Greek) ho philo:n the. love... ‘For the lover is blind in his view of the object loved.’ (Plato, Leges ..e.–; translation based on Bury  and ) () λόγων γὰρ οὐ νῦν ἐστιν ἀγών, ἀλλὰ σῆς ψυχῆς πέρι. logo:n gar u ny:n estin ago:n alla sε:s psykhε:s peri words. P N now is struggle but your life. about. ‘The effort is not about words, but about your life.’ (Sophocles, Electra –; Translation based on Sir Richard Jebb ) (Attic Greek) () οὗτος ἐστιν περὶ οὗ γέγραπται (Koine Greek) útos estin perí u ghéghrapte he. is... about whom. write.-.. ‘This is the one about whom it is written . . . ’ (Novum Testamentum, Secundum Mattheum ..; translation by Senior et al. () Thus the shift toward head initial structures may relate to negator renewal in the history of Greek. At the Koine Greek stage, however, the former N οὐ(κ) /u(k)/ remains the prototypical and most frequent negator of assertions and veridical environments in general. Yet the N-thing οὐδέν /udhén/ is on its way to turning into propositional negation, a change, which like so many others, had already begun in the Classical, or early postclassical period (Horrocks : ).

. Negator distribution in Koine Greek The environments in which N and N are found in Koine Greek are generally the same as the ones of the Classical era (see table .). Their syntactic status remains phrasal and their full distribution can be accounted for with reference to nonveridicality: N remains a negative polarity item. The basic differentiations detected in this stage are two: (i) N μή /mi/ gets more specialized as lexical negation and as the negation of the nonfinite at the expense of N οὐ(κ) /u(k)/; that is, N cooccurs with infinitives and participles in several veridical environments, where

Negator distribution in Koine Greek



formerly N would have been anticipated. This development is already observed in Gildersleeve’s  paper ‘Encroachments of μή on οὐ in Later Greek’; (ii) N increases in frequency in its metalinguistic/constituent negation function, either as οὐ(κ) /u(k)/ or as the emphatic variant οὐχί /ukhí/, which in some cases may be a stylistic preference within the figure of speech litotes (cf. Horn : ), but it appears to have had a permanent effect in the expression of constituent negation in the subsequent language stages. Representative examples of the full distribution of N and N from texts that are at the heart of the Koine Greek stage follow. .. The distribution of N οὐ(κ) /u(k)/ in Koine Greek In Koine Greek N οὐ(κ) /u(k)/ remains the prototypical negator of assertions, unembedded (), as well as embedded (), and is found in other veridical environments, typically: (i) with infinitives () and participles () in the scope of veridical predicates, (ii) with adverbial participles that induce a veridical context, such as causal/modal () and concessive participles, and (iii) as metalinguistic/local constituent negation (). It is also found, as previously, in positive bias questions () and in its adverbial/elliptical uses (), which furthermore prove the phrasal status of N () according to the why no(t)? test introduced in Merchant (). () ἠγέρθη, οὐκ ἔστιν ὧδε. ighérthi, uk éstin ódhe rise... N be... here ‘He has been raised; he is not here.’ (Novum Testamentum, Secundum Marcum ..; translation by Senior et al. ()) (unembedded assertion) () ἵνα γνῷς, φησίν, ὅτι οὐ ψευδῆ ταῦτά ἐστιν. ína gnos phisin oti u pseudhíi tauta estin INA know... say... that N false these are ‘So that you know, he says, that these are not false.’ (Epictetus, Dissertationes ab Arriano digestae ...–) (embedded assertion) () οὐ συμμεῖναι . . . φασιν αὐτούς, ἀλλ’ u simíne phasing autús, al N stay.. say... them but ἀπελθεῖν εἰς τὴν Βοττιαίαν. apelthín is tin Bhotiéan depart.. to the Bottiaea ‘ . . . they did not stay together there, it is said, but went off to Bottiaea.’ (Strabo, Geografica ...–; translation by Jones ()) (infinitive in veridical scope)



Developments in Hellenistic-Roman times

() οὓς ᾔδεσαν οὐ μετασχόντας τῶν us ídhesan u metashóntas ton whom. know... N participate... the. ὅρκων órkon oaths. ‘ . . . those that they knew had not taken part in the oaths’ διὰ τὸ παῖδας ἔτι ὄντας. dhia to pédhas éti óntas for the boys. still be..... ‘because they were still children.’ (Strabo, Geog ...–; translation based on Jones ) (participle in veridical scope: epistemic verb) () Παρθενίας οὐχ ὁμοίως τοῖς ἄλλοις ἐτίμων parthenías ukh omíos tis ális etímon Partheniae. N similarly the. rest. honor... ‘ . . . they would not honor the Partheniae (with civic rights) like the rest, ὡς οὐκ ἐκ γάμου γεγονότας. os uk ek ghámu gheghonótas as N from marriage become..... ‘on the ground that they had been born out of wedlock.’ (Strabo, Geog ...–; translation by Jones ()) (causal participle) () οὐκ ἐμὲ δέχεται ἀλλὰ τὸν uk emé dhéhete alá ton N me. accept... but the. ἀποστείλαντά με. apostílantá me send.... me. ‘(he) receives not me but the One who sent me.’ (Novum Testamentum, Secundum Marcum ..; translation by Senior et al. ()) (metalinguistic/constituent negation) () Οὐ δοκεῖ σοι, ἔφη ἡνῶσθαι τὰ πάντα; u dhokí si éphi inósthe ta pánta? N seem... you. say... unite.. the all ‘Do you not think, he said, that all things are united in one?’ Δοκεῖ, ἔφη. Dhokí. seem... ‘I do, he said.’ (Dissertationes ab Arriano digestae ...–; translation based on Long ) (positive bias question)

Negator distribution in Koine Greek



() ἔξεστιν ἡμᾶς Καίσαρι φόρον δοῦναι ἢ οὔ; éksestin imás Késari phóron dhúne i u is.possible .. us. Ceasar. tax. give.. or N ‘Is it lawful for us to pay tribute to Caesar or not?’ (Novum Testamentum, Secundum Lucam ..–; translation by Senior et al. ()) (adverbial/elliptical) () τὸ φιλόστοργον δοκεῖ σοι κατὰ φύσιν τ’ to philostorghon dokí si kata phýsin t’ the affection seem .. you. by nature. and εἶναι καὶ καλόν; íne ke kalón? be.. and good ‘Does affection appear to you to be according to nature and to be good?’ πῶς γὰρ οὔ; pos ghar u? how P N ‘Certainly.’ (lit. How not?) (Epictetus, Dissertationes ab Arriano digestae ...–; translation based on Long ().) (phrasal status of N) N occurs in other environments as well, which are unspecified in terms of (non) veridicality, such as relative clauses (), where it brings about a veridicality effect. I take that the veridicality effect comes about because of the absence of the Nonveridicality projection. Language is truthful by default. Otherwise this will be indicated through nonveridical marking and nonveridicality will project. N remains also the negator of timendi predicates, in which N is already present as a nonnegative complementizer (). Some of the functions mentioned, such as the cooccurence of N with infinitives, are not represented in all the texts (e.g., they are absent in the Synoptic Gospels). () ὃς γὰρ οὐκ ἔστιν καθ’ ἡμῶν, ὑπὲρ os ghar uk éstin kath’ imón ypér who.. P N be.. against us. for ἡμῶν ἐστιν. imón estin us. be.. ‘For whoever is not against us is for us.’ (Novum Testamentum, Secundum Marcum ..; translation by Senior et al. ()) (relative clause)



Developments in Hellenistic-Roman times

() δέδοικας μὴ οὐ σχῇς μάγειρον. dhédhikas mi u shis mághiron fear... N N have... cook. ‘ . . . you are afraid that you will not have a cook.’ (Epictetus, Dissertationes ab Arriano digestae ...; translation by Long ()) (in nonveridical environment – where N is already present) .. The distribution of N μή /mi/ in Koine Greek As mentioned earlier, N in Koine Greek maintains its polarity behavior and appears in all the nonveridical environments it appeared in in Classical times, either as a negator of finite (), (), () and non-finite forms (), (), or as a compositionally nonnegative complementizer: with verbs of fearing () and in its question particle function (). The distribution of N and N in the New Testament in particular has been discussed in Klein’s () comparative study ‘Negation and Polarity in the Greek, Gothic, Classical Armenian and Old Church Slavic Gospels’. Although Klein (: ) uses the term modal negation to refer to the N of Greek and the N of Armenian—which was also /mi/ (see chapter )—he makes all the relevant observations and categorizations regarding the environments in which N in Koine Greek is attested from which its polarity behavior is evident. Below are examples of the full distribution of N in nonatticizing texts. The main departure from the Attic Greek N distribution is its stronger representation with nonfinite forms, even in some former ‘veridical’ environments, such as causal participles, where previously only N was found or in the case of oratio obliqua (indirect speech) infinitives (). We can view this as a result of its further specialization as lexical/internal negation in the nonveridical environment of DPs () and all sorts of nominals in general. () Μὴ φοβοῦ, μόνον πίστευε. mi phovú, mónon písteue N fear... only believe... ‘Do not be afraid; just have faith.’ (Novum Testamentum, Secundum Marcum ..–; translation by Senior et al. ()) (unembedded directive) () ἵνα μὴ μωρὸς ᾖ, ἀλλ’ ἵνα ína mi morós i, al’ ína INA N naïve be... but INA ‘So that he wouldn’t be a fool, but in order to learn’

μάθῃ, máthi, learn... (purpose clause)

ἃ ἔλεγεν ὁ Σωκράτης. a éleghen o Sokrátis those say... the. Socrates. ‘what Socrates taught.’ (Epictetus, Dissertationes ab Arriano digestae ...–; translation based on Long )

Negator distribution in Koine Greek



() ἐὰν μὴ πυγμῇ νίψωνται τὰς χεῖρας οὐκ eán mi pyghmí nípsonte tas híras uk if` N fist. wash... the. hands. N ἐσθίουσιν. esthíusin eat... ‘ . . . if they do not wash their hands, they don’t eat.’ (Νovum Testamentum, Secundum Marcum ..–) (conditional protasis) () ἔδει σῴζειν τὰ σύμβολα édhi sózin ta sýmvola must... preserve.. the symbols ‘We ought to preserve the signs which God has given,’

τοῦ tu the.

θεοῦ, theú god.

ἔδει αὐτὰ μὴ καταπροίεσθαι. édhi autá mi katapríesthe must... these N dispose.. ‘we ought not to throw them away.’ (Epictetus, Dissertationes ab Arriano digestae ...–; translation by Long ()) (in the scope of a deontic modal) () μὴ ἐχόντων αυτῶν ἀποδοῦναι ἀμφοτέροις mi ehónton autón apodhúne amphotéris N have.... them. return.. both. ἐχαρίσατο. eharísato spare... ‘Because they didn’t have (money) in order to pay back, he spared them both (the debt).’ (Νovum Testamentum, Secundum Lucam ..–) (causal participle (absolute)) () δέδοικα μὴ ἐκκρουσθῶ. dhédhika mi ekrusthó fear... N knock.out... ‘I fear that I may get refuted.’ (Epictetus, Dissertationes ab Arriano digestae ...–.; translation based on Long ) (complementizer with verbs of fearing) () μή δύνανται οἱ υἱοὶ τοῦ νυμφῶνος, mi dhýnante i yí tu nymphónos N can... the servants the. groom. ‘Can the wedding guests’



Developments in Hellenistic-Roman times ἐνᾧ ὁ νυμφίος μετ’ αὐτῶν ἐστιν, νηστεύειν; en o o nymphíos met’ autón estin nistéuin? while the groom with them. is... fast.. ‘fast while the bridegroom is with them?’ (Novum Testamentum, Secundum Marcum ..–; translation by Senior et al. ()) (introducing yes/no questions)

() φασὶ . . . τοὺς Πισάτας μὴ μετασχεῖν τοῦ phasí tus Pisátas mi metashín tu say... the. Pisates N participate.. the. Τρωικοῦ πολέμου. Troikú polému Trojan war. ‘They say that the Pisates did not participate in the Trojan war . . . .’ (Strabo, Geog ...–) (infinitive in veridical scope (oratio obliqua)) () Ὁ ἔχων δύο χιτῶνας μεταδότω τῷ o éhon dhýo hitónas metadhóto to the have... two cloaks give... the. μὴ ἔχοντι. mi éhonti N have... ‘Whoever has two cloaks should share with the person who has none.’ (Novum Testamentum, Secundum Lucam ..; translation by Senior et al. () (lexical negation) .. Statistics of N and N attestations and comparison with Attic Greek Figure . provides the distribution of our -negator sample in non-atticizing texts dated from the first century  to the first century  of mid and low register (for the complete list of texts see the appendix) and figure . compares it to that of Attic (chapter ). As discussed in chapter , the statistics regarding each use of N and N are given primarily in order to verify the claim that ‘probabilistic considerations,’ reflected in frequency of use, ‘have nothing to do with grammar’ (Chomsky : ). They are merely the reflection of preservation or change, not the driving force behind either (see also van Gelderen : –). N maintains its polarity behavior, appearing exclusively in nonveridical environments, among which is its function as lexical negation that rises (either as DP internal or infinitival negation at the expense of N). N remains the prototypical negator of declaratives and veridical environments, but being diachronically unmarked in terms of nonveridicality it can appear in nonveridical environments as well, as in

The negative indefinites of Koine Greek



Conditional protasis Conditional participle Relative clause Purpose/result clause With infinitive in veridical environment Causal/modal participle Question particle or with verba timendi With infinitive in nonveridical environment (Positive bias) question Unembedded directive (prohibitive) Embedded factive (veridical verbs, because clause, relative clause) DP internal/lexical/participial Constituent negation Assertions 0

50

100

150 Neg1

200

250

300

350

Neg2

FIGURE . The distribution of N and N in Koine Greek

the conditional protasis that is now noticeable (figure .), and as the negator of verba timendi complements, where N is already present in its nonnegative complementizer function (not noticeable in the quantitative data, but identified through examples, see (), and traditional grammars, cf. Moulton , Moule , Robertson , Porter ). These are the basic developments that would be less plausibly reduced to pragmatic or stylistic factors, as opposed to the increase of N in unembedded directives in Koine, which may relate to the religious and socially prescriptive character of many Koine Greek texts or the higher representation of N in positive bias questions in the Attic Greek texts, that are of a more dialectical nature.

. The negative indefinites of Koine Greek: N-thing and N-thing In the closing section of this chapter qualitative and quantitative evidence regarding the distribution of the indefinites N-thing οὐδέν /udhén/ and N-thing



Developments in Hellenistic-Roman times

Conditional protasis Conditional participle Relative clause Purpose/result clause With infinitive in veridical environment Causal/modal participle Question particle or with verba timendi With infinitive in nonveridical environment (positive bias) question Unembedded directive (prohibitive) Embedded factive (veridical verbs, because clause, relative clause) DP internal/lexical/participial Constituent negation Assertions 0

100

200

Neg2 in Attic Neg2 in Koine

300

400

Neg1 in Attic Neg1 in Koine

FIGURE . Comparison of N and N distribution between Koine Greek and Attic Greek

μηδέν /midhén/ in Koine Greek is presented, and the relevant data are compared to those of the Attic Greek stage (see figures . and .). The main observation is that their attestation in Negative Concord structures (in the scope of overt N or N or a corresponding negative compound in the case of negative spreading) is significantly reduced, based on a sample of over  nthings (N-things plus N-things) from each stage, most likely an effect of the

The negative indefinites of Koine Greek



Koine Greek Neg1-and Neg2-thing 22% 2% 4% 72%

udhén… u(k)…udhén (NC) mi…midhén (NC) midhén…

FIGURE . Rates of Negative Concord in Koine Greek (Negative Spreading included) Attic Greek Neg1- and Neg2-thing 23% 4% 12%

61%

u:dεn… u:(k)…u:dεn (NC) mε:…mε:den (NC) mε:den…

FIGURE . Rates of Negative Concord in Attic Greek (Negative Spreading included)

N-First Principle, since Horn (). Other than that, N-thing, like Nwords in general, exhibits the same dependency on nonveridicality, in that it appears in the environments in which N appears in its negative functions. Representative examples of their polarity behavior are given in () through (). () αὐτὸς δὲ οὐδὲν ἀπεκρίνατο αὐτῷ autos dhe udhén apekrínato autó he P N-thing respond... him. ‘He gave him no answer.’ (Novum Testamentum, Secundum Lucam ..–; translation by Senior et al. ()) (N-thing in an assertion) () ὁ δὲ ἐσιώπα καὶ οὐκ ἀπεκρίνατο o dhe esiópa ke uk apekrínato the P be.silent... and N respond... οὐδέν. udhén N-thing ‘He was silent and answered nothing.’ (Novum Testamentum, Secundum Marcum ..–; translation by Senior et al. ()) (N-thing in an assertion (NC))



Developments in Hellenistic-Roman times

() εἰ μηδὲν ἕτερον, τῷ χρόνῳ γνώσῃ, ὅτι I midhén éteron to hróno gnósi óti if N-thing else the. time. learn... that οὐδέν ἐστιν. udhén esti N-thing is ‘If you learn nothing else, you will learn from time that (the body) is nothing.’ (Epictetus, Dissertationes ab Arriano digestae ...; translation based on Long ) (N-thing in a conditional antecedent) (N-thing in a conditional consequent (assertion)) () μὴ φοβεῖσθαι μηδέν. N fear..11 N-thing ‘Do not be afraid of anything.’ (Epictetus, Dissertationes ab Arriano digestae ...) (N-thing in a directive (NC)) () οὐδενὶ οὐδὲν εἶπαν udhení udhén ípan N-body. N-thing. say... ἐφοβοῦντο γάρ. ephobhúnto ghar fear... because ‘They didn’t say anything to anyone, because they were afraid.’ (Novum Testamentum, Secundum Marcum ..; translation based on Senior et al. ) (N-thing in an assertion (negative spread)) () ὅρα μηδενὶ μηδὲν εἴπῃς. óra midhení midhén ípis look... N-body. N-thing. say... ‘See that you tell no one anything.’ (Novum Testamentum, Secundum Marcum ..; translation by Senior et al. ()) (N-thing in an embedded directive (negative spread))

. Summary In this chapter the basic changes regarding polarity and negation that pertain to the Koine Greek linguistic stage have been presented. I argued that the phonological developments of this stage, namely the homophony in the mood system, threatened the (non)veridicality-based mood distinction, a fact which subsequently triggered 11

This is the imperatival use of the infinitive, rare, but attested, especially in Epictetus.

Summary



the syntactic emergence of Nonveridicality—lexically represented by the INA particle—as a projection related to, but distinct (i) from morphological mood, (ii) from C. The Nonveridicality head can also be considered to host elements that have been claimed to be ‘polarity’ or ‘Sigma’ heads. The Nonveridicality head qualifies commitment to truth, so it can naturally be thought of as ‘modal’, while it can also be linked to speech-act operations in C (as suggested in Giannakidou ). Among the polarity items that benefit from this treatment is the Greek N. The distribution of N—as any of the various N compounds, e.g. the N-thing μηδέν /midhén/—remains dependent on nonveridicality and according to the proposed analysis carries an uninterpretable Nonveridicality feature [uNV], when checked against an overt or covert nonveridical head. The modal readings it gets are consistent with this. Furthermore, in this stage, N gets more specialized in its lexical negation function at the expense of N, while Negative Concord structures get significantly reduced, a change linked to Greek word-order particulars. This development would have an important impact on the exact manifestation of Jespersen’s Cycle in Greek, further discussed in chapter .

5 Late Medieval and beyond: the renewal of N The third major stage of spoken Greek we are examining regarding the distribution of N and N is Medieval Greek, also known as Byzantine Greek. This period covers roughly ten centuries, from the fifth century  to the fifteenth century  and it is partitioned into Early (fifth to tenth centuries) and Late Medieval (eleventh to fifteenth centuries). Owing to distinct historical factors there are hardly any texts from the Early Medieval period that can be taken to reflect a spoken variety or be securely chronologized at that stage (cf. Markopoulos : –, Horrocks : , Willmott ), although there is an unprecedented abundance of atticizing texts, as well as texts with a problematic chronologization and textual tradition (layers of interpolation, contamination, etc.).1 As a result, our focus here is on the Late Medieval period, in which most of the changes that began during the previous periods are now stabilized (Gignac –). The texts included in our corpus for Late Medieval Greek are all in verse (Ptohoprodromos, Digenis Akritis, Song of Armouris (Vasileios Digenis Akritis kai to Asma tou Armouri’), Livistros and Rodamne, Imberios and Margarona, Chronicle of Morea, Ptoholeon),2 because the vast majority of prose written in this period—as well as in the previous one, Early Medieval—contains various degrees of formal atticizing discourse and manifests distinct traits of artificiality. These atticizing texts are surely worth examining from a linguistic point of view as well, but in a separate study, given that they are ruled out by our ‘vernacular Greek’ criterion, posited regarding text selection in chapter . Although Late Medieval Greek, similarly to the previous stages, was a period of dialectal variation (Horrocks : ), this variation is for the most part reflected in

1 The problem of textual tradition is relevant for all stages of Greek, Classical and Koine Greek included. However, it is more prominent during the Early Medieval stage, because of the religious nature of the majority of texts, as well as Atticism, whenever the scribes felt that either the content or the language of the texts should be altered. 2 See the bibliography for the exact versions and editions selected. Some of these texts, e.g. Digenis Akritis, survive in more than one versions. Our criterion of selection was proximity to the spoken language (cf. also Horrocks : –, – for general information on these texts).

Negation and Nonveridicality in the History of Greek. First edition. Katerina Chatzopoulou. © Katerina Chatzopoulou . First published  by Oxford University Press.



Late Medieval and beyond: the renewal of N

the phonology, which is less relevant to our discussion for this stage. For the territories that remained within the Byzantine empire—for as long as they remained—there was an established form of written Greek, even for low-register vernacular Greek, of the kind that is examined in the present chapter. This linguistic stage, as is the case with most Greek after the Classical era, has not yet received systematic attention, in the form of a comprehensive grammar. With the exception of Jannaris’ () historical Greek grammar, which laid important foundations in this direction, most works are of a philological or historical nature, devoted to particular authors or texts. More recently, the Late Medieval Greek period has been discussed within diachronic overviews of the Greek language, either in more general works (Joseph , Adrados , Horrocks ) or in studies focused on specific phenomena, e.g., the infinitive (Joseph []), the future tense (Joseph [], Pappas , Pappas and Joseph , Joseph and Pappas , Markopoulos ), aspects of inflectional morphology (Pappas , , Mertyris ), and word order (Deligianni a). This study is placed among the latter, as it tracks the expression of negation throughout the history of Greek. Thus in this chapter I am focusing on N and N developments during the Late Medieval period, in anticipation of a more general work in the future.3 The most basic change that is characteristic of this stage, and that has already began during the Early Medieval period (Horrocks ), is the emergence of the former indefinite (N-thing) οὐδέν /udhén/ as unambiguous propositional negation of declaratives and other veridical environments at the expense of the former N οὐ(κ) /u(k)/, with which, however, it coexists in free variation until the end of Late Medieval (see Rijksbaron ). The former N-thing μηδέν /midhén/ shows the same tendency to replace N μη /mi/ in its negative uses, but to a lesser extent and it does not eventually get generalized (section .). Furthermore, the N μη /mi/ itself undergoes a change in syntactic status, from phrase to head (like all the negators of this period) and from being a general negator of nonveridical environments to being a C-related element (section .). These facts have two immediate repercussions: (i) the loss of True Negative Imperatives (section ..), something which also relates to the stabilization of να /na/—former ἵνα /ina/— as a subjunctive mood marker projecting above the NP and (ii) the ban of N μη /mi/ from the conditional protasis (section .). Section . presents evidence from Early Modern Greek (eighteenth and nineteenth centuries ) in verse, in which the negator system of Standard Modern Greek is recognizable: N has received its phonetically reduced form δεν /dhen/, as the sole negator of declaratives (standard negation) and of other environments (mostly veridical, but nonveridical as well, given that N is diachronically unmarked in terms of nonveridicality). N μη /mi/ is the negator of nonveridical environments. Section . summarizes the chapter. 3 Such work of broader scope has in fact begun within the Medieval Greek Grammar Project at the University of Cambridge, under the supervision of David Holton and Geoffrey Horrocks.

Negation in Late Medieval Greek



. Negation in Late Medieval Greek: basic developments In this section the core differentiation of Late Medieval Greek from the previous stages is presented regarding negation, which is variation in the expression of sentential negation. This is an intermediate stage, where previous forms of sentential negation coexist with new ones. The language of these texts already has most of the distinctive properties of Standard Modern Greek and especially the ones that relate to our discussion: the former INA particle, now appearing as να /na/, has increased in frequency, indicating the subjunctive in all its Modern Greek uses, while the finite forms lack formal mood marking, bearing only the temporal and aspectual information, like the verbs of Standard Modern Greek. This information is indicated in the glosses of the examples from this stage. .. How many negators? Medieval Greek is a period of diversity regarding the expression of sentential negation, not only among registers, owing to the diglossia (see chapter ); different forms are found even in one and the same text (see also Willmott ). This is an anticipated stage in the diachrony of a negator system, in that it has been attested in the historical course of negation in other languages as well (cf. van der Auwera a). The previous N οὐ(κ) /u(k)/ and N μη /mi/ remain the statistically most frequent instances of sentential negation and their distribution is once more explainable with reference to (non)veridicality (Giannakidou  et seq.). However, both N οὐ(κ) /u(k)/ and N μη /mi/ develop a variant with which they appear to be in free variation: the former indefinite N-thing οὐδέν /udhén/ (‘not-evenone’) is interchangeable with N οὐ(κ) /u(k)/ and similarly the former indefinite N-thing μηδέν /midhén/ (‘not-even-one’) is interchangeable with N μη /mi/ in its negative uses. Furthermore, the Medieval N οὐδέν /udhén/ is also detectable in its phonetically reduced form, δεν /dhen/, at least as early as the twelfth century , after the loss of initial unstressed vowels in a significant number of words (Gignac : –, Horrocks : –), e.g., ὀψάριν /opsárin/ > ψάριν /psárin/ ‘fish’, ὀσπίτιν /ospítin/ > σπίτιν /spítin/ ‘house’, ατόν /atón/ > τον /ton/ ‘him’ (weak clitic pronoun), εἰς τον /is-ton/ > στον /ston/ ‘to the’ (.). Representative examples of all morphological variants of N and N follow. The new forms, N (οὐ)δέν /(u)dhén/ and N μηδέν /midhén/, are discussed, along with extensive data, in section ... () ὁ εἰς τὸν ἄλλον οὐκ ἀγαπᾶ μόνον με πονηρία. o is ton álon uk aghapá mónon me poniría the one the. other. N love... only with cunningness ‘They do not love each other, but only with cunningness.’ (Chronicle of Moreos ) (N in an assertion)



Late Medieval and beyond: the renewal of N

() ὀϊμέν, ἀδέλφι μας καλόν, οὐδὲν σὲ θεωρούμεν oimén adhélfin mas kalón, udhén se theorúmen alas sibling. our. good. N you. see... ‘Alas our dear sibling, we shall not see you.’ (Digenis Akritis ; the futurate meaning comes from the general context.) (N in an assertion) () τέως γουν δεν εβάστασα να μην τον ερωτήσω. téos ghun dhen evástasa na min ton erotíso any.more thus N bear.. NA N him. ask.. ‘I could not reserve any longer from asking him (I could not bear not to ask him).’ (Livistros and Rodamne ) (N in an assertion, N in the scope of NA) () και σώπα, μη χολομανής ke sópa, mi holomanís and silence... N be.angry.. ‘and silence, do not be angry…’ (Livistros and Rodamne ) (N in an unembedded directive) () Αὐτόθε στέκου, Μαξιμού, ὧδε μηδὲν περάσεις aftóthe stéku Maksimú, ódhe midhén perásis there stand.. Maksimou. here N pass.. ‘Stay there, Maximu, do not come here.’ (Digenis Akritis ) (N in an unembedded directive) In these examples we see that the former indefinites οὐδέν /udhén/ and μηδέν /midhén/ no longer refer to entities. They are rather equivalent to ‘not’. Figure . gives a clear picture of this diversity in the twelfth–fourteenth century vernacular texts that were examined, as well as the preservation of the status of N (either μη / mi/ or μηδέν /midhén/) as a negative polarity item according to the Giannakidou definition presented in chapter . N appears only in nonveridical environments. Such a stage of variation is anticipated and crosslinguistically attested in the diachrony of the expression of negation, a point we return to in chapter . .. Oὐδέν /udhén/ and μηδέν /midhén/: diagnosing plain propositional negation The diagnosis of οὐδέν /udhén/ and μηδέν /midhén/ as plain sentential negators is based on four factors: (i) their cooccurrence with saturated predicates, where they cannot be in any way interpreted as among the arguments of the verb; (ii) the possibility of further reinforcing οὐδέν /udhén/ and μηδέν /midhén/ with other intensifiers, such as ποσῶς /posós/ ‘at all’, resulting in strict Negative Concord structures (Giannakidou ), identical to those found in Standard Modern Greek (see chapter ); (iii) their cooccurrence with accomplishments, predicates denoting bounded, nongradable eventualities, in which οὐδέν /udhén/ and μηδέν /midhén/

Negation in Late Medieval Greek



Conditional protasis

Selected by nonveridical predicate (embedded)

Relative clause

Purpose/result clause

Causal/modal participle With verba timendi

Question particle

(Positive bias) question

Unembedded directive (prohibitive)

Wh-clause

Embedded factive (veridical verbs, becauseclause)

DP internal/lexical/participial

Neg1 dhen

Constituent negation

Neg2 midhén Neg1

Assertions

Neg1 udhén Neg2 mi Neg1 u(k)

0

100

200

300

Neg2

FIGURE . Negator distribution and diversity in Late Medieval Greek

cannot be interpreted as contributing any sort of intensification; (iv) metalinguistic evidence from the Suda dictionary, as early as the tenth century , where the meaning of udhén is explicitly identified with the meaning of N οὐκ /uk/. ... Oὐδέν /udhén/ and μηδέν /midhén/ with saturated predicates The attestations of oὐδέν /udhén/ and μηδέν /midhén/ in environments where they cannot receive a theta role are abundant in this stage. They are found, both with intransitives, as well as transitives with overt objects (we recall that oὐδέν /udhén/ and μηδέν /midhén/



Late Medieval and beyond: the renewal of N

in Classical Greek were indefinites of neuter grammatical gender; as a result they would be incompatible as indefinites in subject position, if the subject is indicated as animate, male or female). () αὐτὸς μικρὸς οὐδὲν εἶδεν τὸ τοῦ λοετροῦ κατώφλιν. aftós mikrós udhén ídhen to tu loetrú katóflin he little N see.. the the. washroom. entrance ‘When he was young he didn’t see the entrance to the washroom.’ (meaning: he didn’t bathe) (Ptohoprodromos III ) (with saturated predicate) () η κόρη ουκ i kóri uk the maiden N ηξεύρει. iksévri know.. ‘The maiden does not Rodamne )

εγνωρίζει eghnorízi know..

με, me me.

τις tis who

είμαι íme am..

ουδέν udhén N

know me, she doesn’t know who I am.’ (Livistros and (with saturated predicate)

() όσοι ουδέν θέλουν να ελθούν εδώ ας απομείνουν. ósi udhén thélun na elthún edhó as apomínun those N want.. NA come.. here  stay.. ‘Those who do not want to come, let them stay here.’ (Digenis Akritis ) ((embedded) with saturated predicate) () ουδέν θυμάσαι τέκνον μου τι εποίκεν ο παπούς udhén thimáse téknon mu ti epíke o papús N remember.. child my what did.. the grandfather σου; su? your ‘Do you not remember, my child, what your grandfather did?’ (Digenis Akritis ) (rhetorical question) () πρόσεχε, μηδέν ξηροχασμάσαι. prósehe midhén ksirohasmáse be.careful... N yawn.. ‘Be careful, do not yawn.’ (Ptohoprodromos III ) (with an intransitive) () αυτά εις εμέν μηδέν τα ειπής. aftá is emén midhén ta ipís these to me N them say.. ‘Do not say these things to me.’ (Digenis Akritis ) (with saturated predicate)

Negation in Late Medieval Greek



() Μηδέν σκοπήσετε άρχοντες ότι δια κούρσον ήλθεν. midhén skopísete árhontes óti dia kúrson ílthen N think.. rulers that for spoils come.. ‘Do not think, o rulers, that he has come for spoils.’ (Chronicle of Moreas ) (with saturated predicate) ... Oὐδέν /udhén/ and μηδέν /midhén/ further reinforced: Strict Negative Concord Apart from being banned from argument position, oὐδέν /udhén/ and μηδέν /midhén/ can no longer function as intensifiers (negative adverbs) in this stage, as they appear with other elements that already do this, giving rise to the Negative Concord Structures of Late Medieval Greek. This kind of Negative Concord pattern is strict, according to the typology of Giannakidou (), as is the case in Standard Modern Greek (see chapter ): the negator (whether N or N) is always required. Especially the attestation of oὐδέν /udhén/ and μηδέν /midhén/ with the minimizer ποσῶς /posós/ ‘at all’ indicates that they can no longer perform this function on their own. () ἐπαίρνει ἡ epérni i take.. the δίδει. dhídhi give.. ‘The maiden takes Rodamne )

κόρη kóri maiden

τὸ βεργίν, καμμίαν οὐδὲν τὸ to verghín kamían udhén to the stick, no one.. N it

the stick, she doesn’t give it to anyone.’ (Livistros and (strict negative concord)

() οὐδὲν ἐδιάβησαν ποσῶς κανένας μήνας δύο udhén edhjávisan posós kanénas minas dhío N pass.. at.all even.one month two ‘Not even a month or two passed by.’ (Chronicle of Moreas ) (strict negative concord) () ποτέ μηδέν οκνήσετε, μη νύκταν μηδέ μέραν. poté midhén oknísete mi níktan midhé méran never N be.idle.. N night N-either day ‘Never be idle, neither night nor day.’ (Digenis Akritis ) (strict negative concord) () Λέγει «πληροφορήθησε, ποσῶς μηδὲν λυπᾶσαι». léghi pliroforíthise posós midhén lipase say.. find.out.. at.all N be.sad.. ‘He says: find out, do not be sorrow at all.’ (Livistros and Rodamne ) (strict negative concord)



Late Medieval and beyond: the renewal of N

() στανεό μου ουδέν staneó mu udhén forcefully me. N καυχάται kafkháte boast.. ‘They did not speak to me Akritis )

μου εμίλησαν, τινάς μη το mu emílisan tinás mi to me. speak.. anyone N it

against my will, may no one brag this.’ (Digenis (strict negative concord)

We also identify the τις /tis/ indefinite (τινάς /tinás/ in the accusative ex. ) as an unambiguous NPI and also in emphatic position. The strict Negative Concord patterns of Late Medieval will not be further discussed here, but we direct the reader for further analysis to Horrocks (),4 who tracks Negative Concord developments from Attic Greek to Late Medieval Greek and provides an account of the relevant indefinites at each stage, along with plausible explanations and further crosslinguistic implications.5 In any case these structures point to one more parameter reset in the transition from Koine (nonstrict Negative Concord) to Late Medieval and Modern Greek. This parameter reset is attested in the diachrony of other languages, e.g., in Italian (Zeijlstra ), but in the opposite direction, from strict (Old Italian) to nonstrict (Modern Italian). Zeijlstra () draws a correlation between strict Negative Concord and the availability of True Negative Imperatives in Italian. As is discussed in section .., the situation was exactly the reverse for Greek. I have no explanation for this fact at this time, other than the consideration that underived word order in each stage may bear relevance to Negative Concord patterns, as it appears to have an effect on Jespersen’s Cycle as well (see section ..). ... Evidence from Aktionsart and gradability render emphatic readings unavailable Further evidence in support of the negator status of oὐδέν /udhén/ and μηδέν /midhén/ in Late Medieval Greek can be drawn from the lexical aspect (or aktionsart) of some predicates they appear to negate, namely accomplishments. Based on Vendler’s (, ) traditional categorization of predicates into: (i) states, e.g. ‘love’, (ii) activities, e.g. ‘run’, (iii) accomplishments, e.g. ‘drown’, and (iv) achievements, e.g. ‘sneeze’, accomplishments are the sort of predicates that denote

Horrocks analyzes the τις /tis/ indefinite paradigm in Attic Greek as ambiguous between the NPI and a positive polarity item (PPI); we find it in Late Medieval Greek in NPI function. See also Veloudis () on the morphological/etymological link between NPIs and PPIs in Standard Modern Greek, regarding in particular the ka(n)- in the NPI κανένας /kanenas/ and the PPI κάποιος /kapojos/ paradigms (cf. also Margariti ,  on the distributive determiner κάθε /kathe/ in Standard Modern Greek). 5 Cf. Ladusaw () and Larrivée () on such transformations. See also Willis et al. () on crosslinguistic parallels to what has been described as the ‘Quantifier cycle’, often attested transformations and tendencies in n-words. 4

Negation in Late Medieval Greek



eventualities, which are durative, internally heterogeneous, and bounded, in that they have a natural endpoint (see also Dowty , Hinrichs , Verkuyl , Krifka , among many others). The aspectual properties of verbal predicates, regarding lexical or grammatical aspect, in the history of Greek is a vast topic that deserves separate examination. For the Koine Greek stage this work has been done to a significant extent (Porter , Fanning , , C. Campbell , a, b), though mostly with the focus on grammatical aspectual distinctions. For the purposes of the present study we need only to point out that, given these properties of accomplishments (durativity, heterogeneity, and boundedness), such predicates are expected to be incompatible with intensified negation (*John didn’t drown at all), as well as intensification in general, because they do not permit a gradable reading. During Late Medieval oὐδέν /udhén/ and μηδέν /midhén/ are attested with such predicates, as well as predicates in general that do not allow a quantifiable interpretation. Some examples of this sort follow. () το κάστρο εζητήσασιν, κ’ εκείνοι ουδέν το δίδουν. to kástro ezítísasin k’ ekíni udhén to dhídhun the castle ask.. and they N it give.. ‘They requested the castle and they are not giving it (the castle).’ (Chronicle of Moreas ) (accomplishment) () ἄλλον νὰ μὴ ἴδης τὴν Συρίαν, ἄν οὐδεν ὑποστρέψεις. álon na mi idhís tin Sirían, an udhén ipostrépsis again  N see.. the Syria if N return.. ‘May you not see Syria again, if you do not return (from Syria to Greece).’ (Dighenis Akritis, based on the Escorial manuscript, ed. S. Alexiou (), ) (accomplishment) () Αὐτόθε στέκου, Μαξιμού, ὧδε μηδὲν περάσεις. aftóthe stéku Maksimú, ódhe midhén perásis there stand.. Maksimou. here N pass.. ‘Stay there, Maximu, do not come here.’ (Dighenis Akritis ) (accomplishment) () Ὢ μά τὸν Θεόν, Φιλοπαππού, οὐδὲν εἶμαι ἐγὼ προδότης. o ma ton Theón Filopapú udhén íme eghó prodhótis   the God Filopapu. N am.. I traitor ‘By God Filopappu, I am not a traitor.’ (Dighenis Akritis ) (nongradable) These attestations of oὐδέν /udhén/ and μηδέν /midhén/ with accomplishments and nongradable predicates in general can only mean that both oὐδέν /udhén/ and



Late Medieval and beyond: the renewal of N

μηδέν /midhén/ have bleached to plain propositional negation (the notion of semantic bleaching is further discussed in chapter ). ... Metalinguistic evidence from the Suda dictionary (tenth century ) Finally, a Byzantine encyclopedia of the tenth century  (earlier than the stage we are discussing) known as the Suda dictionary, an extensive work of over , entries, explicitly identifies the function of oὐδέν /udhén/ with the one of N oὐκ /uk/. The author in a way corrects the cited form with oὐδέν /udhén/, which may still have been considered marginal or vulgar in this use, by explaining to the reader that oὐδέν /udhén/ here should be understood as oὐκ /uk/. () οὐδὲν ἦλθε… ἤθελε γὰρ εἰπεῖν οὐκ udhén ilthe íthele ghar ipín uk N come... want... P say.. N ἦλθε. ílthe come... ‘He udhén came, he wanted to say, he did not come.’ (Suda, omicron..) Based on all these considerations a clear picture emerges regarding the negator status of the former indefinites oὐδέν /udhén/ and μηδέν /midhén/. Horrocks () locates the beginning of this transition regarding oὐδέν /udhén/ at the sixth century , while in their negative adverb functions oὐδέν /u:den/ and μηδέν /mε:den/ occasionally allowed a plain negator reading even since Classical times (see section ..). The N μηδέν /midhén/, however, would not be generalized in this function in the stages to follow, which may be the reason its shift of function during Late Medieval is not much known. N oὐδέν /udhén/, on the other hand, was here to stay, and was also found in its phonetically reduced form δεν /dhen/ every now and then. Figure . summarizes the course of functions of oὐδέν /udhén/ in the history of Greek and manifests the fact that it got generalized as an exponent of plain propositional negation in Late Medieval times. The indefinite μηδέν /midhén/ followed a comparable yet distinct trajectory, further discussed in chapter , where the novel approach of Jespersen’s Cycle is presented along with a formalization of the intuitive notion of semantic bleaching. Regarding its structural position, oὐδέν /udhén/ originates mainly as an element in the VP domain, viewed also as the thematic layer (where theta roles are assigned) (cf. van Gelderen : ) and gradually raises to syntactically higher positions until it reaches the NP. This is in agreement with current generative theories of syntactic change (Roberts and Roussou , van Gelderen , ), as will be discussed in the final chapter.

Change of syntactic status: phrases to heads



The tasks of udhén in the history of Greek argument (subject or object) as in ‘nothing’ argument determiner as in ‘no cat’ adverb as in ‘not-at-all’ adverb determiner as in ‘not-at-all easy’ 275

negator as in ‘not’

148 112 62

45

14

ATTIC GREEK

87 36

54 4

KOINE GREEK

0

0

0

LATE MEDIEVAL

FIGURE . The functions of oὐδέν /udhén/ in three major stages of spoken Greek

. Change of syntactic status: from phrases to heads In chapter  it was shown that the syntactic status of the Attic Greek negators N οὐ(κ) /u:(k)/ and N μή /mε:/ is phrasal, and that they both occupy the specifier of the Attic Greek NP (section .). The phrasal status of N οὐ(κ) /u(k)/ and N μή /mi/ was preserved in Koine Greek. There is, however, evidence that by Late Medieval Greek all exponents of propositional negation ( kinds of N and  kinds of N) are already heads. This explains how we reach a stage where the negators of Standard Modern Greek are heads (see Giannakidou : –), while it also agrees with Zanuttini’s (, ) observations in Romance languages, that preverbal negators tend to be heads. We claim that the syntactic status of Late Medieval Greek negators has shifted from phrasal to capital, based on three observations: (a) they are strictly preverbal and only clitics can intervene between the negator and the verb,6 (b) they do not appear inside the DP and generally they cannot negate anything other than a TP (see figure .), (c) they appear to interfere with head movement, if the unavailability of True Negative Imperatives in Late Medieval is attributed to that (following the analysis of Rivero , Rivero and Terzi  on the availability of True Negative Imperatives, cf. also Zeijlstra ). .. Strict preverbal position All variants of N and N in Late Medieval Greek are found preverbally and only clitics can intervene between the negator and the verb form. This contrasts to 6

See, for the changes in clitic placement, Horrocks ().



Late Medieval and beyond: the renewal of N

the freedom of N and N in Classical Greek and in Koine where the negators need not be adjacent to the verb, while they could also occur in postverbal position through focusing of the verb form (see chapter ). () αὐτός, ὅταν ἐμάνθανε, ὑπόδησιν οὐκ εἶχεν. Aftós ótan emánthane ipódhisin uk íkhen he when study.. shoes N have.. ‘He, when he was a student, didn’t have shoes.’ (Ptohoprodromos IV ) (N οὐ(κ) /u(k)/) () Οὐ μᾶς ἀφῆ ἀπὸ τοῦ νῦν ποιεῖν ἀνδραγαθίας. u mas afí apó tu nín piín andhraghathías N us. let.. from the now do. brave deeds ‘He doesn’t let us, from now on, do brave deeds.’ (Digenis Akritis ) (N οὐ(κ) /u(k)/ with clitic) () Ὢ μά τὸν Θεόν, Φιλοπαππού, οὐδὲν εἶμαι ἐγὼ προδότης. o ma ton Theón Filopapú udhén íme eghó prodhótis   the God Filopapu. N am.. I traitor ‘By God Filopappu, I am not a traitor.’ (Digenis Akritis ) (N οὐδὲν /udhén/) () το στέμμα γαρ το χρύσινον ουδέν το επαρεδέχθη. to stéma ghar to khrísinon udhén to eparedhékhthi the crown P the golden N it accept.. ‘He did not accept the golden crown.’ (Chronicle of Moreas ) (N οὐδὲν /udhén/ with clitic) () τέως γουν δεν εβάστασα να μην τον ερωτήσω. téos ghun dhen evástasa na min ton erotíso any.more thus N bear.. NA N him. ask.. ‘I could not refrain any longer from asking him.’ (Livistros and Rodamne ) (N δεν /dhen/) () Ἐσὺ μικρὸς καὶ esí mikrós ke you little and ‘You are young and Armouris )

ἀνέλικον, anélikon, underage underage,

καβάλλα δὲν σὲ πρέπει kavála dhen se prépi riding.on.horse N you. must you should not ride on a horse’ (Song of (N δεν /dhen/ with clitic)

() και σώπα, μη χολομανής ke sópa, mi holomanís and silence... N be.angry.. ‘and silence, do not be angry…’ (Livistros and Rodamne )

(N μη /mi/)

Change of syntactic status: phrases to heads () στανεό μου ουδέν staneó mu udhén forcefully me. N καυχάται. kafkháte boast.. ‘They did not speak to me Akritis )



μου εμίλησαν, τινάς μη το mu emílisan tinás mi to me. speak.. anyone N it

against my will, may no one brag this.’ (Digenis (N μη /mi/ with clitic)

() «μηδέν φοβάσαι» λέγω την, «πάντης μη εφαντάχθης;» midhén fováse légho tin pándis mi efandákhthis? N fear.. say.. her in.case N dream.. ‘ “Don’t fear,” I tell her. “Did you perhaps have a bad dream?” ’ (Livistros and Rodamne ) (N μηδέν /midhén/) () και μηδέν σου φανή κακόν διατί ονειδίσαμέν σε ke midhén su faní kakón dhiatí onidhísamén se and N you seem.. wrong because scold.. you ‘And don’t be annoyed, because we scolded you.’ (Digenis Akritis ) (N μηδέν /midhén/ with clitic) .. Unattested as DP internal or constituent/metalinguistic negation The distribution of negators in this linguistic stage (figure .) shows that there are hardly any attestations of Ns and Ns either DP internally, as lexical negation, or as constituent/metalinguistic negation. This is in stark contrast with their distribution during Classical and Hellenistic times, where both N and N had a strong representation in these functions. () τοῦ tu: the.

γὰρ gar P

ἀγαθοῦ agathu: good.

τὸ to the.

οὐκ u:k N

ἀγαθὸν agathon good.

ἀπόφασις. apophasis negation. (Attic Greek) ‘the not good is the negation of the good.’ (Aristotle, Analytica priora b –) (N as DP internal negation)

() οὐδεὶς γὰρ τὸ μὴ ἀγαθὸν ἐπαινεῖ. (Attic Greek) u:de:s gar to mε: agathon epaine: n-body.. P the N good. praise... ‘no one praises the not good.’ (Aristotle, Rhetorica a ) (N as DP internal negation)



Late Medieval and beyond: the renewal of N

() οὐκ ἐμὲ δέχεται ἀλλὰ τὸν ἀποστείλαντά με. uk emé dhéhete alá ton apostílantá me N me. accept... but the. sender.. me. (Koine) ‘(he) receives not me but the One who sent me.’ (Novum Testamentum, Secundum Marcum ..; translation by Senior et al. ()) (constituent negation) () Ὁ ἔχων δύο χιτῶνας μεταδότω τῷ μὴ o éhon dhýo hitónas metadhóto to mi the have.. two cloaks give... the. N ἔχοντι. (Koine) éhonti have.. ‘Whoever has two cloaks should share with the person who has none.’ (Novum Testamentum, Secundum Lucam ..; translation by Senior et al. ()) (lexical negation) In Late Medieval such uses of Ns and Ns are unattested. This may relate to the diminishing of the nonfinite elements of the language (infinitives and participles), while Medieval negators of this stage do not occur even with the remaining infinitival forms (see for these forms Joseph []). The word for ‘no’ is by now όχι /ókhi/, a form that goes back to the Classical and Hellenistic Greek emphatic variant of N, οὐχί /ukhi/ (Hatzidakis , Andriotis , Dangitsis , Rijksbaron ; see Joseph  for some complications in this etymology).7 () «και τέως εγνωρίζεις με;» ke téos eghnorízis me? and previously know.. me “And do you know me from before?” I Rodamne )

Λέγω Légho say.. tell him that

τον ton him “no.” ’

ότι «όχι». óti ókhi. that no (Livistros and

Late Medieval Greek negators appear to be capable of negating only tensed phrases and nothing else. An exception to this generalization is the attestation of N μη /mi/ with causal and modal participles (see figure .), a use that resembles its lexical negation function and which in fact survives in Standard Modern Greek, as does its lexical negation function in general (see section ...).

7 N οὐ /u/ is still found in this function only in high, atticizing registers, which are not included in this study.

Change of syntactic status: phrases to heads



() καὶ γὰρ μὴ ἔχων τί φορεῖν μεγάλως ke ghar mi ékhon ti forín meghálos and P N have.-.. what wear. greatly τουρτουρίζω. turturízo shiver.. ‘And because I don’t have what to wear, I shiver greatly.’ (Ptohoprodromos III ) (N with participle) () ἄκων καὶ μὴ βουλόμενος, μέριμνα με ἐσέβην ákon ke mi vulómenos mérimna me esévin unwilling and N want.-.. concern me. enter.. ‘Forced and against my will, I started getting worried.’ (Livistros and Rodamne ) (N with participle) However, such participles, like those of Attic and Koine Greek, should best be treated as CP projecting elements that correspond to full clauses, while for the DP internal attestation of N μη /mi/ in Standard Modern Greek I am assuming that this use was revived by external factors, namely the extensive borrowing from English and French during the twentieth century (Anastassiadis-Simeonidis , ) as a transliteration of the negative prefix non in those languages (Efthimiou ). These functions of N μη /mi/ are of a different nature, in which its former phrasal status is preserved. In its attestation with verb forms, N μη /mi/, similarly to the rest of the negators of this stage, appears to have acquired head status. Further evidence comes from the unavailability of True Negative Imperatives in Late Medieval times. .. Unavailability of True Negative Imperatives: a parameter reset The availability of morphological imperatives, also referred to as True Negative Imperatives or simply prohibitives, is a parameter according to which languages can vary (Joseph and Philippaki-Warburton , Zanuttini , , Rivero , Rivero and Terzi , Tomić , Han , , Zeijlstra , ). In the history of Greek, True Negative Imperatives were available both in Attic Greek () and in Hellenistic Koine (), while surrogate forms through the subjunctive were also attested (see chapters  and ). () μὴ φάθι. mε: phathi N speak.. ‘Don’t speak.’ (Say ‘no’.) (Plato, Gorgias d )

(Attic Greek)

(true negative imperative)



Late Medieval and beyond: the renewal of N

() μὴ πολλῶν mi polón N many. ‘Do not desire many ...)

(Koine Greek) ἐπιθύμει epithými desire... things.’ (Epictetus, Dissertationes ab Arriano digestae (true negative imperative)

By Late Medieval Greek, such structures are nearly unattested in spoken language (only  attestations out of  instances of N μη /mi/ in direct prohibition)8 and only the surrogate forms are possible, using either the perfective non-past or the imperfective non-past forms of the verb. The examples below present instances of negative directives formed by combining with N μη /mi/ with the perfective non-past of φοβάμαι ‘I fear’ in (), while () has an instance of a positive imperative form σώπα /sópa/ ‘silence..’ (which means that the imperative as a morphological category was available and productive) followed by two negative directives in imperfective non-past. () and () provide examples from Standard Modern Greek, where we can actually provide negative evidence on the unavailability of True Negative Imperatives in () and the standard way to form a negative directive in () through the perfective non-past and optionally the να /na/ particle. () μὴ φοβηθῆς τὸν θάνατον παρὰ μητρὸς κατάραν. mi fovithís ton thánaton pará mitrós katáran N fear.. the. death. but mother. curse. ‘Do not fear death, but a mother’s curse.’ (Digenis Akritis ; th century ) (TNIs unavailable) () σώπα, μη χολομανής, τίποτα μη λυπάσαι. sópa, mi holomanís típota mi lipase silence... N be.angry.. nothing N be.sad.. ‘…silence, do not be angry and do not be sad for anything.’ (Livistros and Rodamne ; th century ) (TNIs unavailable) () *Μην έλα. min έla N come.. (‘Don’t come!’)

(Standard Modern Greek)

(TNIs unavailable)

8 These examples are found in Ptohoprodromos I,  and III, , where someone else’s speech is quoted, who is of a socially higher rank than the writer.

Change of syntactic status: phrases to heads () (Να) μην έρθεις/έρχεσαι (na) min érthis/érhese () N come./. ‘Don’t come/be coming.’



(Standard Modern Greek)

(surrogate (im)perfective non-past)

Several analyses have been proposed in the literature on this parameter of crosslinguistic variation, the (un)availability of True Negative Imperatives, which apparently is a parameter that can be reset even in the diachrony of a single language, as was the case in Greek. Zanuttini (, a, ) explains the unavailability of True Negative Imperatives by assuming that in languages that lack them, the imperative mood-marking is in a sense defective (it lacks a certain functional feature), which is problematic in case the negator of that language selects, e.g., a MoodP or a TP, or generally requires that missing feature. I skip further discussion here other than the observation that this analysis cannot easily be carried over to Modern Greek and other Balkan languages, as has been argued in Han (: ). Han (, ), on the other hand, argues that the ungrammaticality of True Negative Imperatives in some languages results because negating an imperative in those languages requires a syntactic structure in which the imperative operator ends up in the scope of negation. This renders these structures uninterpretable, given that illocutionary force cannot be negated, a claim that goes back to Frege () (cf. Dummett , Han : –). Modern Greek is included in Han’s investigation corpus, being among the languages that verify her analysis. However, the applicability of this approach for the case of Late Medieval Greek is difficult to determine, as it involves certain kinds of testing that cannot be performed owing to the nature of this diachronic study. For the purpose of this section, which is determining the syntactic status of sentential negation in Late Medieval Greek, I adopt the account of Rivero () and Rivero and Terzi (), generally supported also in Zeijlstra (). Their analysis links the (un)availability of True Negative Imperatives to the hierarchical structure of functional projections (the locus of the imperative feature) and the syntactic status of the negative marker (see also Giannakidou : –). In languages that form imperatives through V-to-C movement—as does Standard Modern Greek and apparently Late Medieval—True Negative Imperatives cannot emerge, if there is an overt negator of capital status that blocks head movement. A surrogate form is employed instead, whose morphology does not require movement to C, like the Late Medieval perfective non-past or the subjunctive να /na/ particle (while the former ἵνα /ína/ is still attested occasionally in Late Medieval in its unreduced form) that is already base-generated in a position higher than negation after the formal restructuring of the mood system in Hellenistic–Roman times (see section ...). Under this reasoning the Rivero () and Rivero and Terzi () analysis explains the unavailability of True Negative Imperatives in Late Medieval Greek, provided that both N μη /mi/ and N μηδέν /midhén/ are syntactic heads in



Late Medieval and beyond: the renewal of N NegP Neg’ Neg oὐ(κ) μή

XP

Ø

FIGURE . The phrasal status of N and N in Classical and Koine Greek NegP Neg’ Neg

TP

oὐ(κ) oὐδέν δεν μη μηδέν

FIGURE . The head status of negators in Late Medieval Greek

their preverbal prohibitive function. This also agrees with the analysis of Zeijlstra (, ), who generally adopts the Rivero () and Rivero and Terzi () view, but also discusses counterarguments. In fact, Zeijlstra () explicitly links the unavailability of True Negative Imperatives in a language to the head status of the negator: ‘every language that bans TNIs exhibits an overt negative marker X0’ (Zeijlstra : ). Figure . presents the phrasal status of N and N in Classical and Koine Greek, while figure . shows the head status of the Ns and Ns in Late Medieval. Such a development, from phrases to heads, which also holds for the former indefinites—oὐδέν /udhén/ and μηδέν /midhén/—is predicted by current theories of syntactic change, discussed in the following section. In Standard Modern Greek μηδέν /midhén/ went back to being phrasal, but with a different meaning, as we will see in chapter , a development that opposes a strong unidirectional approach for grammaticalization. But at this stage it arguably has bleached to plain nonveridical negation of head status. .. The van Gelderen () Head Preference Principle The transformation of lexical elements from phrases to heads is a phenomenon with crosslinguistic representation in language diachrony, according to van Gelderen’s

The ban of N from the conditional protasis



() study Grammaticalization as Economy, preceded in van Gelderen (), which introduces a viewpoint on grammaticalization and syntactic change that was already applied in the previous chapters, in particular regarding the Late Merge Principle (which explains the ἵνα /ína/ broadening, section ...) and Feature Economy (cf. also Roberts and Roussou , which explains the emergence of the Modern Greek perfective non-past, section ..). One more principle of diachronic change that supports the Phrase-to-Head transformation of the Late Medieval Greek negators is the Head Preference Principle from van Gelderen (: ) presented in () below. ()

Head Preference or Spec to Head Principle: Be a head, rather than a phrase

This is a structure-minimizing principle widely attested, e.g. in the case of English auxiliaries, in the demonstrative that to complementizer and determiner, in the crosslinguistic formation of determiners from pronouns (Heine and Kuteva , Wood ), adverbs to complementizers, to only name a few (see van Gelderen  for a full overview). Most importantly it offers a syntactic explanation for the Negative cycle, further discussed in section ... We have seen so far how the negator system has been enriched in the Late Medieval stage, an anticipated stage of variation in spoken language. The oὐδέν /udhén/ negator was going to be generalized in the stages to follow, in its phonetically reduced form δεν /dhen/, which is the sole negator of declaratives and the general exponent of standard negation in the sense of Payne () that can be found in Standard Modern Greek. N μη /mi/ would not be eventually renewed, but it remains stable and unique in its prohibition function (among other nonveridical functions). This stability and its reasons are discussed in section .. A final development that remains to be seen regarding negator distribution is the exclusion of N from conditionals, as the nonveridical environments that license it have at this stage shrunk by one: the conditional antecedent.

. The ban of N from the conditional protasis One more change that is already settled in Late Medieval Greek and which, similarly to the unavailability of True Negative Imperatives, can be viewed as a parameter reset, is the fact that N no longer appears in the conditional protasis, neither in its former form μη /mi/, nor in its novel, but temporary, variant form μηδέν /midhén/. Only N is generally licensed in the conditional antecedent, either as οὐ(κ) /u(k)/, οὐδέν /udhén/, or δεν /dhen/. Examples () through () contain representative cases.



Late Medieval and beyond: the renewal of N

() Καλῶς ἦλθες, νεώτερε, ἂν οὐκ εἶσαι προδότης. Kalós ílthes neótere, an uk íse prodhótis well come.. younger, if N be.. traitor ‘Welcome, younger one, if you are not a traitor.’ (Digenis Akritis ) Ὢ μά τὸν Θεόν, Φιλοπαππού, οὐδὲν εἶμαι ἐγὼ προδότης. o ma ton Theón Filopapú udhén íme eghó prodhótis   the God Filopapu. N am.. I traitor ‘By God Filopappu, I am not a traitor.’ (Digenis Akritis ) () εἰ δὲ καὶ οὐ θέλεις νὰ ἐλθης, ἰδοὺ ἐγὼ ὑπαγαίνω. i dhe ke u thélis na elthis, idhú eghó ipaghéno if P and N want..  come.. here I go.. ‘And if you do not want to came, here I am going.’ (Digenis Akritis ) () καὶ ἂν οὐδὲν ἔλθης τὸ γοργόν, κατέβην ἔχω εἰς Μάγγε ke an udhén élthis to ghorghón katévin ékho is Máge and if N come.. the soon, go..  to Mage ‘And if you do not come soon, I will go to Mage…’ (Digenis Akritis ) () εἰδέ καὶ δὲν τὸ δέξεται, πάλιν νὰ δευτερώσω. idhe ke dhen to dhéksete pálin na dhefteróso if.P and N it accept.. again NA repeat.. ‘if she does not accept it, I will send again.’ (Livistros and Rodamne ) Figure . depicts the distribution of N and N in the conditional protasis in Late Medieval Greek compared to those of the previous stages, Classical Greek and

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Attic Greek

Koine Greek Neg1 οὐ(κ)/(οὐ)δέν

Late Medieval Greek Neg2 μη

FIGURE . Negator distribution in the conditional protasis in Attic, Koine, and Late Medieval

The ban of N from the conditional protasis



Hellenistic Koine. N μη /mi/ is attested in conditionals in Late Medieval, but to a very limited extent. The change in the distribution of N and N from Attic Greek to Koine is not statistically significant (p-value = .), while the change from Koine Greek to Late Medieval is significant (p-value <  15). This use of N is no longer productive and would not make it to the Standard Modern Greek stage. Some considerations on this change are given in section .., while section .. discusses one such remnant of the use of N in conditionals of Late Medieval: its use in exceptive phrases. .. N fossilized in exceptive conditionals In the Late Medieval stage N μη /mi/ is attested very rarely in the conditional protasis and mainly in fossilized phrases, such as εἴμη /ími/ or εἰμή /imí/ (‘if-not’) as part of their morphology. These structures are not exactly pure conditionals, but have the form of exceptive phrases (translated as ‘other than’). () οὐκ ἔναι καλλιότερον, uk éne kalióteron N is.. better. ‘There is no better thing other Rodamne )

εἴμη ími if.N than the

ἡ φιλία τοῦ πόθου. i filía tu póthu the friendship the. lust. friendship of lust.’ (Livistros and

These kinds of clauses, described as ‘exceptive conditionals’ (see von Fintel  on the function of unless in English) were productive especially during the Koine Greek stage, but would not survive in Standard Modern Greek. () τίς δύναται ἁμαρτίας ἀφεῖναι εἰ μὴ μόνος tis dhýnate amartías afíne i mi mónos who can... sin.. forgive... if N only ὁ Θεός; o theós the. God. ‘Who but God alone can forgive sins?’ (Who can forgive sins, if not only God) (Novum Testamentum, Secundum Lucam ..–; translation by Senior et al. ()) (Koine Greek) Thus, we only have a vestigial presence of N in the environment of the conditional protasis. N no longer recognizes it as a licensing environment. This agrees with the fact that we have posited N as being marked in terms of nonveridicality, whereas N is the unmarked, standard negation of the language. Nonveridicality still projects in the conditional protasis, as it explains the presence of other negative polarity items, items that carry an uninterpretable Nonveridicality feature [uNV]. Such a lexical item is the former indefinite τίς /tis/ ‘someone’, which was unspecified in



Late Medieval and beyond: the renewal of N

terms of nonveridical marking in Classical and Koine Greek. In Late Medieval Greek the τίς /tis/ indefinite appears in nonveridical environments only, such as the conditional protasis in () and in Negative Concord structures, as in (). () εἴ τιναν ἐνετράνισαν τῆς κόρης τὰ ὀμμάτια i tinan enteránisan tis kóris ta omátia if n-body gaze.. the. maiden. the. eyes. ‘If the maiden’s eyes would look at anyone…’ (Imperios and Margarona ) (NPI in the conditional protasis) () στέκει καὶ συνηβάζεται, stéki ke sinivázete stand.. and agree.. ‘He goes and comes to agreement and Margarona )

τινὰς νὰ μὴν το μάθῃ. tinás na min to máthi n-body  N it know.. that no one should know that.’ (Imperios (negative concord)

Therefore, it is not that the conditional antecedent somehow lost the ability to license negative polarity items and it is clear that, according to the syntactic treatment of negative-polarity item licensing that was proposed in chapter , the Nonveridicality phrase still projects in the conditional antecedent. The correlation to morphological mood is not exactly relevant or even at least sufficient (that would be to say that N becomes by Late Medieval the negator of indicatives and N of nonindicatives, which is not the case), because the protasis of conditionals also licenses the perfective non-past form of the verb, which is a polarity item and can in fact occur with both N or N, depending on the semantic environment (see section ..). The reasons of the N unavailability in conditionals by the Late Medieval linguistic stage should be traced elsewhere. .. An explanation: upward reanalysis of N on Cinque’s () hierarchy The unavailability of N in the conditional protasis during Late Medieval (although other polarity items are still licensed in that environment) possibly relates to a repartitioning of labor between N and N that must have taken place during the Early Medieval stage. The picture that emerges for the use of N in Late Medieval— as well as for the stages to follow—is that N became an element that correlates to the C position, in contributing illocutionary force (as in the case of prohibition, interrogation, and introducing verba timendi complements). The conditional protasis does not offer such a position for N. The C position in conditionals is filled by the ἀν /an/ or the εἰ /i/ conditional particles that now compete with N and are in complementary distribution. This is a fact that describes the situation in Standard Modern Greek as well. Only N can appear in the conditional protasis in Standard Modern Greek, as presented in the example below (see also chapter ).

The ban of N from the conditional protasis



() Αν δεν/*μην έρθει, θα στεναχωρεθώ. An dhen/ *min érthi tha stenakhorethó if N/*N come..  be.sad.. (Standard Modern Greek) ‘If s/he doesn’t come, I will be sad.’ (only N in the conditional protasis) In the example above we see that the conditional protasis is unable to license the Modern Greek N μη(ν) /mi(n)/, although it does license the perfective non-past form of the verb έρθει /érthi/ ‘come’, which, as shown in chapters  and , is a negative polarity item (cf. Giannakidou ), the historical development of which provided evidence in support of a syntactic treatment of nonveridicality. One reason for this may be that N, apart from its shift in syntactic status (from phrase to head) may also have reanalyzed as relating to a higher position within C, in its expanded form according to Cinque’s () chartographic approach. N μη /mi/ seems to have risen to a position where it competes with the conditional particle both in Late Medieval (ἀν /an/ or the εἰ /i/) and in Standard Modern Greek (αν /an/). This transition is represented in () below. ()

The upward reanalysis of the Greek N μη /mi/ on Cinque’s hierarchy

MoodSpeechAct MoodEvaluative MoodEvidential ModEpistemic T(Past) T(Future) MoodIrrealis ModNecessity AspFrequentative(I) AspCelerative(I) ModVolitional ModObligation ModAbility/Permission AspCelerative(II) T(Anterior) AspTerminative AspContinuative AspPerfect(?) AspRetrospective AspProximative AspDurative AspGeneric/progressive AspProspective AspSgCompletive(I) AspPlCompletive Voice AspCelerative(II) AspSgCompletive(II) AspRepetitive(II) AspFrequentative(II) AspSgCompletive(II)

ModPossibility AspHabitual AspRepetitive(I)

In Cinque’s account the relevant projections are signified as Mood, but since we have established the independency of N from morphological mood (see chapters  and ), this position can be seen simply as a higher C position, with which the N is now linked, following the diachronic tendency for up-the-tree movement (Roberts and Roussou ). The MoodIrrealis is clearly akin to our nonveridicality notion in its purest form: deprived from all additional connotations, such as speech-act, evaluativity, evidentiality, etc. It has been claimed that a movement from MoodIrrealis to MoodSpeechAct is involved in the synchronic derivation of conditionals in general (Danckaert and Haegeman ). To examine the relevance of Danckaert and Haegeman’s claim with the upward reanalysis of N from a position that merely indicates irrealis (in Classical and Koine Greek) to the locus of illocutionary force would need a more focused study, but it relates to the Roberts and Roussou () perception of



Late Medieval and beyond: the renewal of N

grammaticalization as ‘loss of movement’ (that is, loss of synchronic movement; the elements get permanently reanalyzed as originating in their previous landing site). Therefore, the loss of N from the conditional protasis may be a result of its diachronic movement to a syntactic position in which the conditional particles are already hosted. N’s other functions (as a particle introducing yes/no questions and a complementizer selected by timendi predicates) were already in that position (see also Willmott , who provides a mapping for the uses of N on Cinque’s hierarchy), as was N in its prohibitive use. This account (see also Chatzopoulou a/b, ) agrees with the terminology I propose for regular syntactic change as upward lexical micromovement, while Cinque’s () hierarchy of functional projections has already been applied to the explanation of language change in Roberts (). This viewpoint provides the tools for the description of diachronic change out of which the identification of major diachronic tendencies results. Not all changes are permanent and not all functions of an element elevate at once. But if they do, this is the path they appear to follow. .. Comparison to N’s persistency in purpose clauses: possible support for upward micromovement Yet N μη /mi/ was not lost from all embedded environments in which the C position is filled (as was argued to be the case regarding conditionals). In purpose clauses N has remained as the only possible negator since Classical times. Yet, after the Koine stage, an overt complementizer can be optional. () ἵνα μὴ μωρὸς ᾖ, ína mi morós i, INA N naïve be... ‘So that he wouldn’t be a fool, but in tiones ab Arriano digestae ...–)

ἀλλ’ al’ but order

ἵνα μάθῃ… (Koine Greek) ína máthi, INA learn... to learn…’ (Epictetus, Disserta(purpose clause)

() a. νὰ τὰ παραφυλάσσωμεν, τινὰς νὰ μὴ τὰ μάθη. na ta parafilásomen tinás na mi ta máthi  them protect.. n-body  N them learn.. (late medieval) ‘We should keep these secret, so that no one finds out.’ (Livistros and Rodamne ) (purpose clause: with C) b. Ἐκεῖνος ἀπεκρίθην τον κρυφά, μὴ τὸν ἀκούσω. ekínos apekríthin ton krifá mi ton akúso he respond.. him. secretly N him. hear.. ‘He replied to him secretly so that I wouldn’t hear him.’ (Livistros and Rodamne ) (purpose clause: N-C)

Evidence from Early Modern Greek texts



() Ο Γιάννης ντύθηκε (για) (να) μην κρυώσει. o Jánis díthike (ja) (na) min kriósi the Jannis dress... (in.order) () N get.cold.. (Standard Modern Greek) ‘John got dressed in order not to catch a cold.’ (purpose clause) The Cinque () hierarchy and a view of syntactic change as gradient (Haspelmath , Hopper and Traugott , Lehmann , Lightfoot , Traugott and Trousdale b, Roberts ) allow us to account for the presence of N in purpose clauses with an optional complementizer. In this use N originated in the MoodIrrealis—in the terminology of Cinque—position in Attic Greek. In Late Medieval and Modern Greek the presence of an overt complementizer position filler (e.g., the subjunctive να /na/ or για να /ja na/) is optional. These structures are further evidence for N’s reanalysis on a higher position in Cinque’s hierarchy by the Late Medieval stage, because given that N can introduce purpose clauses by itself, it is clear that it can express illocutionary force on its own in this use. In the presence of overt purpose complementizers we can assume that N lexicalizes along with these elements in the C position of purpose clauses. As a result, the upward reanalysis of N presented in () holds for purpose clauses as well. Alternatively it could be argued the N in purpose clauses is still situated in Cinque’s MoodIrrealis position even in the later stages of Greek, and that a purpose complementizer is always present in such clauses, either overtly or covertly. Either explanation supports the usefulness of the Cinque () fine-grained account of functional projections in explaining diachronic language change, as it provides pathways of grammatical change that have been taken in the history of Greek and elsewhere (see Roberts ). So far we have seen the developments regarding the distribution of negation that are characteristic of Late Medieval Greek. The proximity to the Modern Greek stage is already evident. The final section of this chapter provides some evidence from Early Modern Greek texts. By Early Modern Greek there no longer is variation among different forms of N and N, while the selection between the two remains dependent on (non)veridicality.

. Evidence from Early Modern Greek texts: the end of the journey In the final section of this chapter some evidence is presented on the expression of negation in Early Modern Greek, as represented in a collection of poems and folk songs placed in the eighteenth and nineteenth centuries  (Legrand –, Politis ). The variation stage in the attestations of N and N is over. The negative particles of Greek have been stabilized in their Standard Modern Greek forms, N as δε(ν) /dhen/ and N as μη(ν) /mi(n)/.



Late Medieval and beyond: the renewal of N

() Δὲν ἐλαλοῦσε σὰν πουλί, σὰν ὅλα τὰ πουλάκια. dhen elalúse san pulí san óla ta pulákja N sound.. like bird like all the birds ‘It didn’t sound like a bird, like all the birds sound.’ (A. Politis : ) (N dhen in assertion) () τί ‘ναι μικρὸ κι ἀνήξερο, τ’ ἄρματα δὲν γνωρίζει. ti ‘ne mikró ki aníksero t’ ármata dhen ghnorízi because is.. young and unaware the gun. N know.. ‘Because it’s young and unaware, it doesn’t know (how to use) guns.’ (A. Politis : ) (N dhen in assertion) () Δὲ στό εἰπα, Λιάκο, μιὰ φορά, δὲ στό εἰπα dhe sto ípa Ljáko mjá forá dhe sto ípa N you.it say.. Liako one time N you.it say.. τρεῖς καὶ πέντε; tris ke pénde three and five ‘Didn’t I tell you, Liako, once, didn’t I tell you three and five times?’ (A. Politis : ) (N in positive bias question) () Ὅσο εἶν’ Βελούλας ζωντανός, παιδιά μου μή φοβᾶστε. óso in Velúlas zondanós pedhjá mu mi fováste while is.. Velulas alive children my N fear.. ‘As long as Velulas is alive, my children, do not fear.’ (Halatsas : ) (N mi in directive) () σύμμασ’ τὰ símas’ ta handle.. the Αγραφιῶτες. Aghrafiótes Agrafeans ‘Handle your men so : )

παλληκάρια palikárja men

σου su yours

μὴ mi N

πράζουν prázun bother..

τς ts the

that they don’t bother the people of Agrafa.’ (A. Politis (N mi in purpose clause (or embedded directive))

() μή σέ βαραίνουν τά φλουριά, μή σέ mi se varénun ta flurjá mi se N you. burden.. the coins N you. βαραίνουν τ’ ἄσπρα; varénun t’ áspra? burden.. the money ‘Are you burdened by the coins? Are you burdened by the money?’ (Halatsas : ) (N in yes/no questions)

Evidence from Early Modern Greek texts



Neg2 mi(n) 31%

Neg1 dhe(n) 69%

FIGURE . N and N percentages in eighteenth- to nineteenth-century folk songs ( tokens)

Selected by nonveridical predicates Purpose/result clause Causal/modal participle With verba timendi Particle introducing yes/no questions (Positive bias) question

Unembedded directive (prohibitive) Embedded factive (veridical verbs, becauseclause) DP internal/lexical/participial Assertions 0

20

40 Neg1 dhe(n)

60

80

100

Neg2 mi(n)

FIGURE . Negator distribution in Early Modern Greek (folk songs)

N δεν /dhen/ has replaced all previous forms, both the Classical and Koine Greek N οὐ(κ) /u(k)/ and the former indefinite οὐδέν /udhén/ in its noneroded form. N μηδέν /midhén/, on the other hand, has disappeared completely. It is not attested in these texts. Only the former form of N μη /mi/ appears in all the relevant and nonveridical functions. Figure . gives a clear picture of the non-variation stage that has been reached by the Early Modern Greek stage, while Figure . shows the distribution of N and N in these texts manifesting once more the polarity behavior of N μη /mi/. Not all functions of the negators are attested, especially the embedded attestations of N and N, because of the nature of these texts, in which parataxis is strongly preferred. The purpose of the illustration in to show that the variation stage among



Late Medieval and beyond: the renewal of N

multiple N and N forms has ended: N was renewed, N remained stable, in form as well as in its dependency on nonveridicality.

. Summary In this chapter the expression of negation in Late Medieval Greek was examined, which was a stage of variation among different forms of N and N that yet performed the same functions. Earlier forms coexist with new ones, both for N and for N, as the Classical and Koine Greek indefinites oὐδέν /udhén/ and μηδέν /midhén/ had bleached to plain sentential negators and were in free variation with the former N oὐ(κ) /u(k)/ and N μη /mi/ forms. This variation would have cleared out by Early Modern Greek (eighteenth–nineteenth centuries ) at the expense of the former N oὐ(κ) /u(k)/. N μη /mi/, however, would remain stable in form and in dependency on nonveridical contexts. Apart from variation and its eventual resolution, Late Medieval is also a stage in which two basic parameters relating to N and N particles have been reset (a change that should be placed in the Early Medieval period, for which no reliable vernacular texts survive): (i) the change of syntactic status of Ns and Ns in their sentential negation functions from phrases to heads, a fact that was also linked to the unavailability of True Negative Imperatives, and (ii) the ban of N from the conditional protasis, for which an explanation was proposed that relies on an upward reanalysis of N in the Cinque () hierarchy of functional projections. The applicability of Cinque’s () elaborate expansion of the CP for the explanation of diachronic change has already been explored in Roberts (). Our explanation provides further support in that direction. This path is further pursued in section ., regarding the origin of the diachronically stable uses of N μη.

6 Renewal and stability One full Jespersen’s Cycle and one persistent polarity item We have witnessed change and stability in the history of Greek negation. N was renewed, in that the prototypical function of Classical Greek οὐ(κ) /u:(k)/ as the negator of assertions was in Late Medieval assumed by the former indefinite (N-thing) and negative adverb οὐδέν /udhén/ ‘nothing’ > ‘not-at-all’ > ‘not’, which by general phonological alterations resulted in the Modern Greek N δεν /dhen/. N on the other hand remained intact in its prototypical use in negative directives, enduring only a vowel change μή /mε:/ > μη /mi/ that was among the general sound changes that took place during Hellenistic and Roman times. N was preserved also as DP internal/lexical negation and in its nonnegative complementizer uses, as a particle that introduces yes/no questions and is embedded under verbs of fearing and the like. In this chapter the particular manifestation of negator renewal in Greek is discussed in connection to the phenomenon of Jespersen’s Cycle ( Jespersen , , after Dahl ). It is shown that the diachronic development of the Greek negator system does not properly fall under the traditional morphosyntactic description of the Jespersen’s Cycle processes (Fischer et al. , Zeijlstra , van der Auwera a, Willis, Lucas, and Breitbarth, , among others). As already pointed in Willmott (), Greek did not manifest what is described as a stage II of Jespersen’s Cycle, which is the stage of discontinuous negation (de Swart : ), either for N or for N (see also Vossen , van der Auwera et al. ). In this chapter I propose that the actual generalization regarding Jespersen’s Cycle is located in the semantics, a view already suggested in Furtado da Cunha (), as well as De Cuypere () from a cognitive–functional perspective. The analysis I provide is couched within formal semantics and depends heavily on the notions of scalarity and intensification, which are shown to be diachronically unstable. The proposed analysis of Jespersen’s Cycle is inclusive to typical Jespersen’s Cycle languages, such as French (Bréal /, Clarke , Tesnière , Horn ,

Negation and Nonveridicality in the History of Greek. First edition. Katerina Chatzopoulou. © Katerina Chatzopoulou . First published  by Oxford University Press.



Renewal and stability

Detges and Waltereit , Godard ), English (Horn , Frisch , Mazzon , Wallage , ), Dutch/Flemish (Hoeksema , Zeijlstra , Breitbarth and Haegeman ), Egyptian (Gardiner ), Old Norse (Van Gelderen a), Arabic and Berber (Lucas ), Yiddish (Van der Auwera and Gybels ) among many others, but also to other cases of diachronic negator renewal, viewed as atypical or partial Jespersen’s Cycle manifestations in the traditional sense. Greek is among the latter. The evidence that has been provided from the distinct phases of the Greek language (chapters  through ) shows that both negators, N and N, underwent Jespersen’s Cycle in the proposed sense, but it was only in the case of the former, N, that the Cycle was completed and N οὐ(κ) /u:(k)/ was eventually replaced by N (οὐ)δέν /(u)dhén/. (See table ..) The N-thing of Classical Greek, μηδέν /mε:den/—the nonveridical counterpart of N-thing οὐδέν /u:den/—grammaticalized in a different direction, which follows from the analysis proposed here (see also Chatzopoulou , a/b, ), but it did not replace the former N μη /mi/ in its diachronically stable uses, which is an element that goes back to the Proto-Indo-European *meH (section ..). The discussion proceeds in the following way. In section . some traditional descriptions of Jespersen’s Cycle are presented, along with examples of languages in which it is attested in this precise way. Section . summarizes the developments of N in the history of Greek, from which its divergence is evident and follows prototypical Jespersen’s Cycle manifestations. In section . a novel definition of Jespersen’s Cycle is introduced, which explicitly places the regularity of the phenomenon in the semantics, abstracting away from previous morphosyntactic accounts. In section . the diachronically stable uses of N are presented and the developments of the N-thing μηδέν /midhén/ are discussed in relation to a formal refinement of the notion of bleaching that is implied by the proposed definition for Jespersen’s Cycle, as lexicalization of a scalar endpoint. Section . summarizes the chapter.

TABLE . The two-negator contrast from Proto-Indo-European to Standard Modern Greek N

N

Proto-Indo-European *ne vs *meH ............................................................................................................................................... Homeric Greek u:(k[h]) vs me: Classical Greek u:(k[h]) vs me: Koine u(k) vs mi Late Medieval u(k) (and udhén) vs mi (and midhén) Modern Greek dhe(n) vs mi(n).

Traditional accounts of Jespersen’s Cycle



. Traditional accounts of Jespersen’s Cycle The term ‘Jespersen’s Cycle’ was first used by Östen Dahl in his  paper ‘Typology of sentence negation’, to refer to the process by which the expression of negation in a language tends to increase and decrease in complexity over time in regular ways. The exact, often cited, formulation that Otto Jespersen made by examining data from three languages, English, French, and Danish is the following: The history of negative expressions in various languages makes us witness the following curious fluctuation: the original negative adverb is first weakened, then found insufficient and therefore strengthened, generally through some additional word, and this in its turn may be felt as the negative proper and may then in course of time be subject to the same development as the original word. (Jespersen : ; see also Jespersen : )

French is the prototypical example of a language that exhibits such a development and is obligatorily mentioned in all studies throughout the Jespersen’s Cycle literature, being among the three languages of Jespersen’s original corpus. Examples in () are from Van der Auwera (: ), while negation in English (Horn , Frisch , Wallage , , among others) and Dutch (Hoeksema , Zeijlstra , Van der Auwera , , a, Breitbarth and Haegeman ) has gone through similar transformations, see (). () Il ne peut venir Il ne peut pas venir Il peut pas venir he N can N come ‘He can’t come tonight.’ () STAGE I STAGE II STAGE III

ENGLISH ne … ne … na:ht …. noht

ce ce ce this

soir. STAGE I (Old French) soir. STAGE II (Middle French) soir. STAGE III (Modern colloquial) evening

DUTCH en … en ... niet …. niet

ETYMOLOGY OF THE 2ND WORD na:ht na wiht ‘no creature, thing’ niet ne iet ‘no thing, nothing’

Jespersen’s observation regarding the tendency for regular renewal of the expression of negation in a language—particularly regarding the French ne . . . pas—has been preceded not only by Antoine Meillet’s seminal work on grammaticalization (: –), but as pointed out recently in Van der Auwera , by Alan H. Gardiner (), who discusses the origin of negators in Egyptian and Coptic in relation to the French negator transformations. These words, from the Latin passum and punctum, were originally adverbial accusatives placed at the end of negative sentences for the purpose of emphasis; just like the English “not a jot”,



Renewal and stability

“not a straw” . . . . Pas and point, and like them the Demotic , Coptic ⲁⲛ an, next lose their emphasizing force, and become mere adjuncts of the negative words (French ne, Coptic ⲛ̅ ). Last of all, they come themselves to be looked upon as negative words. (Gardiner : , cited in van der Auwera : )

Jespersen’s phrasing is open to interpretation regarding the terms weakening and strengthening, whether what he had in mind was in fact phonological, or perhaps semantic (see Kiparsky and Condoravdi ). However, the negator renewal path in the languages he studied was morphosyntactic; all three languages (English, French, and Danish) manifested a doubling stage, the addition of a second element (either an indefinite or a minimizer), that after a point was required for the expression of plain sentential negation and eventually assumed the function of plain sentential negation by itself. This had a permanent effect in the understanding of the Jespersen processes. The much-neglected work of Gardiner introduces the notion of emphasis in an explicit way and makes all the relevant generalizations as early as . Yet apart from the fact that Gardiner’s () contribution did not receive nearly the same amount of attention as did the observations and claims Meillet () and Jespersen () made independently, the negator renewal path in Coptic, which is discussed by Gardiner, is morphosyntactic as well and thus could only reinforce the dominant morphosyntactic viewpoint regarding the Jespersen’s Cycle processes. As a result, Jespersen’s Cycle is generally described in the literature as a diachronic multistage process that involves three main stages, this occurring as recently as de Swart (: ), whose description of each stage is given relatively to the position of the verb (). () Preverbal expression of sentential negation. Discontinuous expression of sentential negation. Postverbal expression of sentential negation.

STAGE I STAGE II STAGE III

A similar three-stage description for Jespersen’s Cycle, which is the dominant approach, is given in Burridge (), Bernini and Ramat (), Haspelmath (a), Zanuttini (), Horn (), Hoeksema (), Roberts and Roussou (), Van der Auwera and Neuckermans (), Mazzon (), Roussou (), and Lucas (), among others (see for extensive bibliography Van der Auwera : ). Several intermediate stages have been occasionally postulated, the number of which can vary depending on the level of detail and/or the language(s) under consideration. Intermediate stages are generally the stages of ambiguity or competing forms, in which either one of the two exponents of propositional negation can be optional. To name but a few, Schwegler (), Lenz (), Van Kemenade (), Larrivée (), and Schwenter () posit a total of four stages, while Honda (), Donhauser (), Anderwald (), Zeijlstra (), Willis () identify five stages, Van der Auwera () identifies six stages, while Willmott () posits a

Traditional accounts of Jespersen’s Cycle



total of seven stages.1 An example of a four-stage Jespersen’s Cycle is found in Fischer, van Kemenade, Koopman, and van der Wurff (), as shown in (), while a five-stage expansion of the cycle, which manifests the competition between old and new forms, is given in van der Auwera (a: ) presented in (). ()

STAGE : Negation is expressed by one negative marker. STAGE : Negation is expressed by a negative marker in combination with a negative adverb or noun phrase. STAGE : The second element in stage  takes on the function of expressing negation by itself; the original negative marker becomes optional. STAGE : The original negative marker becomes extinct.

()

STAGE 1 ne …

STAGE 2 ne … ne … pas

STAGE 3 ne … pas

STAGE 4 ne … pas … pas

STAGE 5 … pas

Greek shows evidence for intermediate stages, stages in which the former exponent of propositional negation coexists and is in free variation with the new one, in particular during the Late Medieval stage (see chapter ). However, this is where the similarities of Greek with morphosyntactic accounts of Jespersen’s Cycle stop. As is shown in the following section (.), Greek did not manifest a Jespersen’s Cycle in the traditional sense, regarding neither the developments of N, nor those of N. The notions of emphasis and bleaching that have been extensively discussed within the Jespersen’s Cycle literature and in connection with the expression of negation and polarity in general (Gardiner , Meillet , Jespersen , Wackernagel , Pohl , Borkin , Schmerling , Fauconnier , Dahl , Horn , Heim , Hock , Hopper and Traugott , Krifka , Lenz , Beukema , van Kemenade , Dahl a, Detges and Waltereit , Eckhard , Kiparsky and Condoravdi , Breitbarth and Haegeman , Jäger , Van der Auwera , Israel ) are highly relevant to the diachronic development of the Greek negators. The underlying assumption in most of these studies is that the defining properties of negator renewal regularities as a diachronic tendency are semantic. The goal of section . is to make this assumption more prominent and provide a formal description of the intuitive notions of emphasis and bleaching.

1

See for more typological diversity and variation on the analyses in Vossen ().



Renewal and stability

. Atypical negative cycles in Greek and beyond .. The atypicality of Greek N and N developments: no doubling stage According to the traditional descriptions for Jespersen’s Cycle we have seen, Greek is problematic in that it does not have a Stage II, as observed also in Willmott (), which is the stage of discontinuous negation. The Attic Greek N u:(k[h]) was gradually replaced in Late Medieval by N οὐδέν /udhén/, see () and (), and the same happened to N in its negative functions, although to a lesser extent, see () and (), as N μηδέν /midhén/ did not eventually replace the former N μη /mi/ in any of its uses (a point to which I return in section .). ()

. . . τοῖς φίλοις αὐτῶν οὐ θεωροῦσιν. (Attic Greek) h h tois p ilois auto:n u: t eo:ru:sin the. friends. their. N look... ‘they do not observe (the misfortunes) of their friends.’ (Isocrates, Epist  (Ad filios Jasonis) .–; translation based on Norlin )

() οιμέν αδέλφιν μας καλόν, ουδέν σε θεωρούμεν. oimén adhélfin mas kalón, udhén se theorúmen alas sibling. our. good. N you. see... (Late Medieval Greek) ‘Alas our dear sibling, we do not see you.’ (Digenis Akritis ) () μὴ mε: N

ψεῦσον . . . τῆς ἐπιούσης ἐλπίδος. pseuson tε:s epiu:sε:s elpidos falsify.. the. sum... hope. (Attic Greek) ‘Do not prove wrong the hope that comes from this.’ (Aristophanes, Thesmophoriazusae ; translation based on O’Neill ) (Late Medieval Greek)

() το αδέλφιν μας το άρπαξες, μηδέν μας το to adhélfin mas to árpakses, midhén mas to the sibling our. which abduct.. N our. it στερέψης. sterépsis deprive.. ‘Do not deprive us from our sibling that you took away.’ (Digenis Akritis ) An explanation for the lack of a Stage II comes from the fact that Greek, being a nonstrict negative concord language, according to the Giannakidou () typology, at least until the end of Koine (third c. ), employed two syntactic strategies of negative reinforcement (). On a par with negative concord structures in Italian and Spanish (cf. Zanuttini , Laka , Herburger ), ()

Atypical negative cycles in Greek and beyond



and (), the Greek negators of Classical and Koine Greek were generally required, in the case that the n-word was postverbal, (), () and dropped, if the n-word was preverbal, (), () () Negative reinforcement strategy : N verb N-WORD Negative reinforcement strategy : N-WORD verb. () Non ha telefonato nessuno. N has called n-body ‘Nobody called.’

(Italian)

() Nessuno ha telefonato. n-body has called ‘Nobody called.’

(Italian)

() οὐ πέπονθεν οὐδέν. (Attic Greek) u: peponthen u:den N suffer.-. N-thing ‘Nothing happened to him.’ (Aristophanes, Pax  (fifth century )) () οὐδὲν πέπονθεν. (Attic Greek) u:den peponthen N-thing suffer.-. ‘Nothing happened to him.’ (Isocrates, In Call. . (fifth century )) () μὴ νῦν ἔτι εἴπῃς μηδέν. mε: ny:n eti e:pε:is mε:den N now more say... N-thing ‘Do not say anything more.’ (Sophocles, Elena )

(Attic Greek)

() μηδὲν λέγε . . . mε:den lege N-thing say... ‘Do not say anything . . . ’ (Plato, Symposium .d.)

(Attic Greek)

After the significant decline of negative concord structures during the Hellenistic Greek period (see chapter ) and the strong preference of N- and Ncompounds for preverbal position (a preference that is already noticeable in Classical times), strategy  (N-WORD verb) was further stabilized as the dominant strategy for negative reinforcement. As a result, it was strategy  that gave rise to the negators N οὐδέν /udhén/ and N μηδέν /midhén/ of Late Medieval Greek, through the bleaching and reanalysis of the former negative indefinites, and a discontinuous stage of sentential negation, did not occur. () presents the descriptive stages of the Greek negator system, in which the competing forms of the intermediate Medieval Greek stage are included.



Renewal and stability

()

Attic Greek

Koine Greek

Medieval Greek

Standard Modern

Neg1

u: (k[h]) …

u(k) …

u(k) … udhén … dhen …

dhe(n)…

Neg2

mε: …

mi …

mi … midhén …

mi(n)…

This development, at least regarding the Greek N, for which the cycle was completed, is reminiscent of that of the Latin non, which deviates from prototypical Jespersen’s Cycle patterns in a similar way: Latin did not exemplify a discontinuous negation stage either. Latin non resulted from the merging of the weak negator ně (see van der Auwera a: ) and œnum, which was the word for ‘one’ in Early Latin (later unum, ne + œnum > non), an etymology that has been assigned to non already in Jespersen (: –), as it is among the languages that he discussed regarding the expression of negation, and was later supported also in Wackernagel (). The fact that neither Greek nor Latin manifested a Stage II in the traditional Jespersen’s Cycle description may not be circumstantial. Both Greek and Latin at Stage I of the traditional cycle were primarily SOV languages. The potential correlation of Jespersen’s Cycle with word order in a language—a hypothesis already entertained in Vennemann () and Harris (, )—is further discussed in section ... In the following section two more cases of atypical negator renewal instances (that do not fall from the Jespersen’s Cycle description as it is) are presented. One of them involves cases where negators come from former verbs, existential verbs in particular, manifesting what is known as ‘Croft’s cycle’ (after Croft ), exemplified by Chinese among multiple other languages (cf. Heine and Kuteva , Van der Auwera , Van Gelderen b, Veselinova , , ). The other refers to languages in which a negator tripling stage is attested, that is, the expression of plain sentential negation may require the presence of three distinct negative elements. Examples of such a development are found in some Bantu and Austronesian languages and even in European languages, for example in dialects of Italian and Flemish (cf. van der Auwera and Neuckermans b, Devos and van der Auwera , Devos, Kasombo Tshibanda, and van der Auwera , van der Auwera, Vossen, and Devos ). .. Other atypical Jespersen’s Cycle manifestations Negator renewal paths that deviate in one way or another from the traditional understanding of Jespersen’s Cycle and the prototypical case of French have been identified crosslinguistically. Two are of particular relevance, as the approach for Jespersen’s Cycle I am proposing is inclusive of these as well: (i) the case in which

Atypical negative cycles in Greek and beyond



propositional negation has its origin in a verbal head and as a result, there is no syntactic discontinuity effect, (ii) the case where negation is reinforced for a second time with no loss of phonological material from the previous form, resulting in a tripling negator stage. ... Negator renewal through former verbs: no doubling stage It has been known since Givón () and Payne () and more recently discussed and enriched with data from more languages in Croft (), Heine and Kuteva (), Miestamo (), Van der Auwera (), and Van Gelderen (a/b, ), that one second major strategy for negator renewal to which languages may resort for various reasons is the one where sentential negation originates from former verbs and in particular verbs that have some inherent negative property, e.g. ‘to lack’, ‘to fail’, ‘to not exist’, ‘to refuse’. This is a path attested outside the Indo-European language family. Chinese is one prototypical such language. Sentential negation mei in Modern Chinese, used with predicates of inherent perfectivity, comes from the Old Chinese verb mo ‘to die’ with a predictable change in the vowel (Lin , see also Djamouri ,  and Pulleyblank  for a complete overview of negative elements in Old Chinese and earlier stages of the language and on the verbal origin of most, cited in Van Gelderen : ), while the canonical negator of Modern Chinese bu most probably comes from the archaic Chinese form wu, which meant ‘to lack’ (Heine and Kuteva : ). Examples from the diachrony of the first negator, mei, are given below. () Yao Shun ji mo . . . (Old Chinese) Yao Shun since died ‘Since Yao and Shun died . . . ’ (Mengzi, Tengwengong B, from Lin : ) () Yu de wang ren mei kunan wish  died person not.be suffering ‘If you wish that the deceased one has no suffering . . . ’ (Dunhuang Bianwen, Lin : –). (Early Mandarin: intermediate stage) () Wo mei you shu. I N exist book ‘I don’t have a book.’ (from van Gelderen : )

(Modern Chinese)

Numerous such languages have been identified that manifest this pattern, with degrees of variation in the details. Among them are seven Athabaskan languages, including Ahtna, Koyukon, Lower Tanana, and Chipewyan (Van Gelderen : –), Australian aborigine languages, e.g., Nunggubuyu (Croft : ), some varieties of Berber (see Mettouchi , Chaker ), and the Tagalog huwag prohibitive marker, which most likely derives its etymology from a negative verb of desire (see Croft : ). In most of these cases, there is a correlation of the negator



Renewal and stability

that comes from a verbal head with perfectivity or telicity in general, in that at some point in its history (before becoming a negator) it carried perfective aspect marking, as is the situation in Early Mandarin () and in several Athabaskan languages (see Van Gelderen b on prefixal l- negation in these languages and Givón  on the verbal origin of such perfective prefixes in Athabaskan; also Rice ) or it placed restrictions on the predicates it could cooccur with, again in terms of perfectivity, either lexical or grammatical. ... Recycling without any erosion: a tripling stage Another way in which the emergence of plain sentential negation can deviate from the traditional morphosyntactic description of Jespersen’s Cycle that was presented in section . is the case where a language that already has syntactically discontinuous negation can enter the cycle for the second time and have a third element added, which with time bleaches to plain sentential negation without any loss of phonological material from the previous stage. Such a development results in a compound morphosyntactic structure that can require three elements—at least—for the expression of plain propositional negation. Multiple such examples have been attested crosslinguistically in Bantu, e.g. in Kanincin, a dialect of Ruund (Devos and van der Auwera , Devos, Kasombo Tshibanda, and van der Auwera ; see also Nurse  for negation in Bantu in general) and Austronesian languages, e.g. in Lewo, a Vanuatu language (Early a, b, Van der Auwera , Crowley b, van der Auwera and Vossen ), and it is even found in European languages, e.g., in dialects of Dutch/Flemish (van der Auwera and Neuckermans ), as well as in Dego Italian (Manzini and Savoia ). There are in fact cases even of quadrupling and quintupling attested more rarely with or without some kind of fusion of the negative markers (see Vossen , Van der Auwera et al. ). The more standard and not so rare case is, however, negator tripling. Here is an example of a prohibitive structure attested in Lewo (from Early b: ). The basic negative pattern required the embracing structure ve . . . re. However, the prohibitive toko (which is of verbal origin, ‘to desist’) can be further added, resulting to a ve . . . re toko negator without traces of emphasis. () Ve a-kan re toko! N .eat N N ‘Do not eat it!’

(Lewo (Austronesian))

Similarly in Dego Italian, this time in a declarative negation, a tripling stage is attested. The example below is from Manzini and Savoia ( III: ). () a ŋ m εŋ soŋ nεnta lavɔ I N me N am N wash. ‘I have not washed.’

(Dego Italian)

Jespersens unite: a new definition



In fact, although less known, the tripling negator stage not only is not ‘exotic’ regarding Indo-European languages, but it was actually first attested in a Belgian Brabantic dialect, Aarschot, by Jan Lodewijk Pauwels as early as . The example below is from Pauwels (: ) (cited in van der Auwera a). () Pas op dat ge nie en valt nie! take.care that you N N fall N ‘Take care that you don’t fall!’

(Aarschot (Brabantic))

This is a situation that would have occurred, for instance in French, if the ne . . . pas construction was reinforced through another lexical element prior to the loss of ne from the pattern in actual language. This of course was not the case for French, nor was it for English, Standard Dutch, or any of the other typical Jespersen’s Cycle languages. But clearly, it is a typological possibility that, similarly to the developments of Greek, is not predicted by the text book account for Jespersen’s Cycle. The goal of the following section is to amend this situation by proposing a broader definition for this phenomenon.

. Jespersens unite: a new definition for Jespersen’s Cycle The desire of some sort of unification of such typical and atypical Jespersen’s Cycle manifestations underlies much of the literature as recently as Vossen () and van der Auwera et al. (), perhaps as a way of supporting the intuition that deep down it is the same morphosemantic phenomenon realized in various ways. In this section, as has been argued for in Chatzopoulou (a, a/b, ), a description of Jespersen’s Cycle is spelled out, which is inclusive of both typical and atypical Jespersen’s Cycle instantiations. Overall I support a treatment of the particular diachronic tendency in terms of a typology, as a phenomenon that exhibits typological variation, an assumption that already underlies much of the material of the Jespersen literature and a position explicitly taken in Van der Auwera (), where a distinction is drawn between the Jespersen Cycle—understood as a cycle that involves a doubling stage—and the Negative cycle in general. Van der Auwera notes that the term Jespersen Cycle is preferable both ‘to Jespersen’s cycle, because there is too much variation to warrant the definitizing genitive, and to Negative cycle, to single out the cycles with a doubling stage from all other negative cycles’ (van der Auwera : ).2 Yet the point where most Negative cycles meet—Greek included—is the bleaching of predicate negation to plain propositional and based on this observation, which is far from new, I propose the definition in ().

2 For the use of the term Negative cycle see van Gelderen  and Mithun  among others, cited in van der Auwera et al. ().



Renewal and stability

()

Jespersen’s cycle definition: Formalization I (Universal) Negator renewal through the semantic bleaching and structural elevation of intensified predicate negation to plain propositional, further reintensified by morphological, syntactic, prosodic, or other means. Formalization II (includes traditional Jespersen’s Cycle, Greek, Latin, and tripling stage languages) If X is a negative expression, either syntactically continuous or discontinuous, and α a variable of quantities (as of individuals, amounts, or times) Jespersen’s Cycle goes through the following stages: ⟦X⟧ = λP . λα. [∀d > . ¬P(d)(α)] (intensified predicate negation) STAGE II ⟦X⟧ = λp. ¬p (plain propositional negation) STAGE I

This definition reflects a view that labels what has so far been described as emphatic negation, as intensified negation (see also Mustajoki and Heino , Hammond , Xue-lian  for such a treatment of emphatic negation), that is, plain negation with the addition of an intensifier (morphological or syntactic) with an understanding for intensification in the sense of Clara Romero (): intensity refers to a qualitative or quantitative gap between two states relative to a phenomenon. As can be seen in the proposed definition, and in agreement with Clara Romero’s () approach, intensification involves quantification of degrees. The correlation of negation with intensity agrees with Bolinger’s () observation that gradability can have crosscategorial relevance and is not just a property of adjectives (a view preceded in Sapir , see Kennedy and McNally ). I maintain the term Jespersen’s Cycle for all negative cycles, with or without a doubling stage, given that in his  phrasing Otto Jespersen said that negation is strengthened “generally through some additional word.” He did not say “always.” The proposed definition focuses on the emphatic/intensified form of negation that eventually bleaches to plain sentential negation, and identifies two stages: one in which the element is emphatic and one in which it is not. A two-stage analysis for Jespersen’s Cycle has already been pursued in Furtado da Cunha’s () treatment on the history of negation in Brazilian Portuguese, yet from a cognitive–functional perspective, as her analysis of the Cycle includes two successive stages of iconicity and economy (see also de Cuypere : ). Our account is cast within formal semantics, while it also detects the impact of vertical structure on the exact manifestation of Jespersen’s Cycle in a language through evidence from Greek. Predicate negation in Formalization I of the definition refers to a negative element either (i) from within the verb phrase (VP), which can be either the complement of the verb (as in French, Dutch, German, English, Greek, Latin) or the verb itself (as in Chinese, Athabaskan, Semitic, and Austronesian, and languages in general in which

Jespersens unite: a new definition



the negator is of verbal origin) or (ii) immediately above the VP, as in the case of a negative adverb (e.g., never in African American Vernacular, cf. van Gelderen a, and possibly Homeric and Attic Greek N οὐ(κ) /u:(k)/).3 In the formalization provided in Formalization II of the definition, X is used as a structurally and compositionally opaque element, which can stand either for the French ne . . . pas, or the Greek udhén, the Chinese mei, or the three cooccurring negators attested, e.g. ŋ . . . εŋ nεnta in the Italian dialect in (). ⟦ne . . . pas⟧ = λP . λα. [∀d > Ø. ¬P(d)(α)] (intensified predicate negation) STAGE II ⟦ne . . . pas⟧ = λp. ¬p (plain propositional negation)

() STAGE I

⟦udhén⟧ = λP . λα. [∀d > Ø. ¬P(d)(α)] (intensified predicate negation) STAGE II ⟦udhén⟧ = λp. ¬p (plain propositional negation)

STAGE I

⟦ŋ . . . εŋ nεnta⟧ = λP . λα. [∀d > Ø. ¬P(d)(α)] (intensified predicate negation) STAGE II ⟦ ŋ . . . εŋ nεnta ⟧ = λp. ¬p (plain propositional negation) STAGE I

The two forms of definition Part II are truth-conditionally equivalent, something which may facilitate the eventual generalization of intensified negation as plain propositional. The difference is that in the case of emphatic/intensified negation the X element explicitly negates a whole set of entities or quantities or times that are introduced through a scale denoting lexical property. This point is further elaborated in section ... The exact denotations for the emphatic form of negation in Stage I can vary in the details, especially in languages like Chinese, where the negator is of verbal origin. This kind of negator renewal path is covered by Part I of the new definition for Jespersen’s Cycle, namely the elevation of predicate negation to propositional. Part II is broad enough to be inclusive of traditional Jespersen’s Cycle languages (e.g. French, English, Dutch/Flemish, German, Egyptian, Old Norse, Arabic and Berber, Yiddish), Greek, Latin, and many tripling stage languages, but the precise denotations of intensified negation in each of these languages are expected to vary. Under this definition, Greek has undergone only one complete Jespersen’s Cycle (contra Kiparsky and Condoravdi , who posit four cycles under a different approach and with evidence from Modern Greek dialects as well) in that the expression of propositional negation was renewed only once and only for N. Greek N entered the cycle as well, but the new form was not generalized 3

According to Cowgill’s () etymology the Homeric u:(k) comes from a form that used to mean ‘never’: Proto-Greek *ne oiu kwid (‘not ever in my life’) > Homeric Greek u:(k) N (cf. van Gelderen : ), see section ..



Renewal and stability

and it did not survive in that function in Modern Greek. The N indefinite μηδέν /midhén/ (etymologically ‘not-even-one’) that entered the cycle—in that it was both semantically bleached and structurally elevated—in Standard Modern Greek grammaticalized in another direction; a direction that arises from the formal description of the intuitive notion of bleaching, which is presented in the following section as a diachronic lexical phenomenon that results in the lexicalization of a scalar end point. .. Jespersen bleaching as scalar endpoint lexicalization Jespersen’s Cycle has been argued to be the result of two interleaving processes: that of semantic weakening, described as bleaching, of the expression of negation, followed by a syntactic or morphological strengthening. Bleaching is usually defined as loss of semantic features of a lexical element at a certain point in its life and is viewed in connection with grammaticalization in general. But what exactly does it mean for an element to bleach and is there something that can determine which features are eventually lost and which features are preserved? The relation of bleaching to Jespersen’s Cycle is uncontested. Among the concluding remarks of their  paper, Kiparsky and Condoravdi state: Jespersen’s cycle is due to the interleaving of two processes: the strengthening of negation by morphosyntactic means, and its loss of compositionality and weakening (grammaticalization, bleaching). . . . An emphatic negative always weakens by itself (it is “bleached”), never in virtue of being combined with some other element. On the other hand, negation is strengthened only by combining a simple negative with an indefinite. A simple negative, or a simple indefinite, never becomes an emphatic negative on its own. (Kiparsky and Condoravdi : )

Similarly Horn () describes Jespersen’s Cycle as ‘the repeated pattern of successive weakening and restrengthening of the negative marker’ (see Horn : ). Regarding the second part of the Kiparsky and Condoravdi comment, I have shown that it is not the case that negation is strengthened only through overt combination of the former negator with an indefinite or, as a matter of fact, any additional element. Instead, the viewpoint I introduce is one according to which the expression of sentential negation in a language can be replaced by another, which can be from any of various sources. Thus it is not the case that something always gets added to negation to reinforce it, but rather that a former emphatic/intensified negative expression at some point loses its force (bleaches), becomes interchangeable with plain sentential negation, and eventually may replace it. The substitution scenario is the only one that can account for negators of verbal origin as well, where it is clear that no reinforcement took place, either with an additional element (as in traditional Jespersen languages) or without one (as in the case of the Greek udhén and African American vernacular never). Descriptively, we can say that the former intensified form loses at some point the specialization it may have with respect to certain kinds of predicates

Jespersens unite: a new definition



(or arguments, in the case of negators that are of verbal origin) and acquires a broader distribution. This is what happened in the history of the Greek οὐδέν /udhén/ (etymologically ‘not-even-one’). As early as the Classical era (see section ..) and more frequently in the stages to follow, οὐδέν /udhén/ occurred, not only in argument position, but with saturated predicates as well, where it could not receive a theta role. This is the negative adverb stage of οὐδέν /udhén/, where it has a ‘not-at-all’ meaning. At some point during Medieval Greek and definitely by the tenth century , οὐδέν /udhén/ was already further bleached into a plain negator, as which it is found with non-quantifiable predicates, most prominently with accomplishments, which would be incompatible with a scaleintroducing element that would impose a gradable reading (for the relation between boundedness and nongradability see, among others, Paradis ). Below is an example from section ... repeated in (). In () οὐδέν /udhén/ can be interpreted only as a plain negator, because the predicate ὑποστρέψεις /ipostrépsis/ ‘return’ cannot receive a gradable reading. Furthermore, οὐδέν /udhén/ in () occurs in the conditional protasis after the repartitioning of labor between the two negators, N and N in Early Medieval, during which N μη /mi/ gradually became syntactically incompatible with the conditional particle ἄν (see section .). () νὰ μὴ ἴδης τὴν Συρίαν, ἄν οὐδεν ὑποστρέψεις na mi idhís tin Sirían, an udhén ipostrépsis  N see.. the Syria if N return.. ‘You shall not see Syria, if you do not return.’ (Dighenis Akritis ; th century ) (accomplishment, bounded, non-gradable) The situation was the same with the French ne . . . pas (sometimes with the addition of the numerical un, e.g. ne . . . un pas, see Schwegler : ). Originally ne . . . (un) pas used to cooccur with verbs of motion such as aller ‘to go’ and marcher ‘to walk’ and contributed the meaning ‘not one step’ (cf. de Cuypere : ), but gradually it was generalized to all kinds of predicates through a similar process. The observation we can make is that at some point—the point of bleaching— intensified forms of negation of this sort failed to introduce a scale (see also C. Lee  for bleaching as loss of scalarity in English negatives). It is this loss of scalar reference that made Greek οὐδέν /udhén/ compatible with nonquantifiable predicates during Medieval Greek, as in () the predicate /ipostrépsis/ ‘return’. In most popular Jespersen’s Cycle manifestations this scale is the numeric scale (see also Israel ). In a Horn scale representation (after Horn ) this scale would have the following form. ()

< 1 . . . ., , , , , , >



Renewal and stability

What emphatic/intensified negation does is to explicitly negate the rightmost element of this scale and everything to its left, either literally or by implicature. The EVEN element which is etymologically detectable in the Greek οὐδέν /udhén/ (‘not-evenone’) is uncontestably a scale-evoking element (cf. Giannakidou  and section ..). Such a focus particle is present in negators of other languages, such as the Amharic -mm (see van der Auwera et al. ), while in other Jespersen’s Cycle languages the EVEN element can be inferred. The morphology of the Attic Greek N-thing μηδέν /mε:den/ is identical: ‘not-even-one’. As was discussed in section .., μηδέν /midhén/ entered the cycle similarly to οὐδέν /udhén/, in that it began to replace the nonveridical negator N μη /mi/ during Late Medieval, but it was not eventually generalized. It rather bleached in another direction, which provides evidence in a clearer way on what it means for an element to bleach. At some point in early Modern Greek μηδέν /midhén/ began to lexicalize ‘zero’, and this meaning was the only one maintained until today. ()  –  = 

μηδέν /midhén/

(Standard Modern Greek)

According to the Horn scale in () literally introduced by the EVEN feature on both οὐδέν /udhén/ and μηδέν /midhén/, ‘zero’ is the scalar end point and the standard of comparison in our formalization of Jespersen’s Cycle in (). Thus μηδέν /midhén/ ended up lexicalizing the endpoint of the scale it originally introduced in a way parallel to course of the English null ‘zero’, which also derives its etymology from the Latin negative indefinite nullum (etymologically ne + ullum ‘not any’). The fact that Greek N was generalized as lexical negation during the Koine Greek stage may relate to the bleaching of μηδέν /midhén/ toward an element in the lexical domain. οὐδέν /udhén/, on the other hand, lexicalized an end point in the functional domain: plain propositional negation. The lexicalization of a scalar endpoint is a common diachronic development of other scale-evoking elements, such as adjectives, comparatives, diminutives, honorifics, etc. In this view, reinforcing of a negator or a comparative involves the explicit reintroduction of a scalar endpoint. With time the scale may be lost from the meaning and the semantic property maintained will be the notion of the end point (something that may relate to language-acquisitional factors, which will not be discussed here). Highly relevant in its viewpoint and exact phrasing is the comment on the diachronic development of adjectives made in Lightfoot’s () paper ‘Explaining Syntactic Change’. ‘ . . . adjectives are regularly ‘devalued’ by a kind of linguistic inflation: ‘excellent’ comes to mean merely ‘good’. ‘enormous’ to mean ‘big’, and ‘fantastic’, ‘fabulous’, etc. lose their original force. As this happens, so new superlatives must be invented to describe the end-point on some scale.’ (Lightfoot : ).

Jespersens unite: a new definition



Although the relation between negation and scalarity has been extensively discussed in the literature (Jespersen : –, Fauconnier , Horn , Miestamo : –, Giannakidou and Yoon  on the (dis)connection between polarity and scalarity), the exact link between them regarding the Jespersen’s Cycle phenomenon, to our knowledge, has not yet been identified. I have shown that the kind of bleaching that relates to the Jespersen’s Cycle processes is one that involves the lexicalization of a scalar end point, either in the realm of lexical functions or—more often regarding Jespersen’s Cycle—in the realm of grammatical functions, and negation is only one of the processes that diachronically exploit this mechanism (cf. also comparatives, diminutives, honorifics). We have thus seen the semantic transition of a word from an indefinite to ‘zero’ and from an indefinite to sentential negation. For the intermediate stage of the transition, from an indefinite to an adverb (a transformation both οὐδέν /udhén/ and μηδέν /midhén/ experienced) our current explanation is syntactic and immediately follows from the Roberts and Roussou () and van Gelderen () generative outlooks on grammaticalization and syntactic change. .. The representation of the Greek Jespersen’s Cycle and word-order considerations The syntactic representation of the Greek negator transformations (which are actually the transformations of the indefinites, οὐδέν /udhén/ and μηδέν /midhén/, that respectively replaced the former N οὐ(κ) /u(k)/ permanently and N μη /mi/ for a while) can be explained through the general tendencies observed in syntactic change. Most prominently through the Late Merge Principle (‘Merge as late as possible’) as elaborated in van Gelderen (, , b), used also in the Roberts and Roussou () generative approach on syntactic change (see section ...). Descriptively, Late Merge manifests the preference of lexical elements for structurally higher positions every time this is possible. The transformations of οὐδέν /udhén/ and μηδέν /midhén/ in the history of Greek are in accordance with this tendency. Figure . shows the diachronic structural elevation of the N-thing οὐδέν /u: den/ from an indefinite to a negator, while figure . presents the same development for the N-thing μηδέν /mε:den/, with the additional presence of the Nonveridicality projection that licenses it (standard N negation, although a nonveridical operator, can never license any N compound, see also section ..) and the restriction that it did not eventually replace the former N μη /mi/, for reasons that are discussed in the following section. The indefinites οὐδέν /u:den/ and μηδέν /mε:den/ had already grammaticalized to a higher position, even in the Attic Greek stage, but had preserved their former function, as is often the case in grammaticalization phenomena (cf. Greenberg ). The indefinite use was still the most frequent one in Classical times.



Renewal and stability NegP

Neg’

Neg

VP

AdvP

VP

NP

V

(ου)δέν /(u)dhén/ οὐδέν /udhén/ οὐδέν /u:den/

NEGATOR STAGE ‘not’

ADVERB STAGE ‘not-at-all’

INDEFINITE STAGE ‘nothing’

FIGURE . Upward lexical micromovement of the Greek N indefinite

We note that the grammaticalization path of the two Greek indefinites observes vertical structure. Regarding the syntactic change, the formal description we can provide is that the uninterpretable [uN] feature that both οὐδέν /udhén/ and μηδέν /midhén/ used to carry until Hellenistic Koine, according to the analysis for Classical Greek negative concord provided in chapter , in Late Medieval Greek either became interpretable [iN] or in the analysis of Zeijlstra (, ) the feature remained uninterpretable [uN] in agreement with a covert negator. This transformation is parallel in a broader sense to the one of the French pas, the Dutch niet, and the English na:ht, with the restriction that the negator in Stage I was overt in these languages, whereas in Greek it was covert (see ..). This agrees with the Roberts and Roussou () view on grammaticalization as loss of—covert—movement (the movement of οὐδέν /udhén/ and μηδέν /midhén/ in a way similar to that of Modern Greek ούτε /úte/, cf. Giannakidou ) and the elements under discussion have been reanalyzed as base-generated in the position they appear. The Late Merge Principle in van Gelderen () makes the same predictions. As a result, the view on grammaticalization as loss of movement (Roberts and Roussou

Jespersens unite: a new definition



NonVerP

NegP

Neg’

Neg

VP

AdvP

VP

NP

V

(μηδέν /midhén/) μηδέν /midhén/ μηδέν /mε:den/

NEGATOR ADVERB STAGE STAGE (temporary) ‘not’ ‘not-at-all’

INDEFINITE STAGE ‘nothing’

FIGURE . Upward lexical micromovement of the Greek N indefinite

, Roberts ) is supported by the data regarding the diachrony of the Greek negator system. The underived order of major constituents is highly relevant to the exact manifestation of Jespersen’s Cycle in a language, as was implied in section .. in connection with the Latin non. The fact that Classical Greek, similarly to Latin, was an underlying SOV language (Ebeling , Joseph [] , Devine and Stephens , Taylor , Dik , , Deligianni a) means that both οὐδέν /udhén/ and μηδέν /midhén/ at their indefinite stage were preverbal, whether they were subjects or objects of the verb form. This fact and the developments that followed support de Cuypere’s (: ) suggestion that discontinuous negation can arise only in VO languages with preverbal negation or OV languages with postverbal. Greek was neither and, as a result, no discontinuity stage occurred. As was discussed in chapter , οὐδέν /udhén/ and μηδέν /midhén/ could also appear postverbally, but only through focusing of the verb form and after one of the discourse particles, such as δέ /de/, that have been analyzed as focus particles in Arad and Roussou (), see section ... Furthermore, their postverbal occurrences in



Renewal and stability

Negative Concord structures were significantly less frequent and they would decline even more during the Hellenistic Koine stage, as the ground was gradually prepared for their reanalysis as sentential negators. Figures . and . show the preference of οὐδέν /u:den/ and μηδέν /mε:den/ for preverbal position in Attic Greek (which is Stage I of the Greek Jespersen’s Cycle) and in Hellenistic Koine respectively. This preference may also have been in effect of the N-F principle (Horn : , preceded in Jespersen : ) that favors the preverbal positioning

400 350 300 250 200 150 100 50 0 Neg1-thing /u:den/ Postpredicate n-word Negative Concord

Neg2-thing /mε:den/ Prepredicate n-word

FIGURE . Syntactic distribution of N- and N-thing in Attic Greek (Stage I)

450 400 350 300 250 200 150 100 50 0 Neg1-thing /udhén/

Postpredicate n-word Negative Concord

Neg2-thing /midhén/

Prepredicate n-word

FIGURE . Syntactic distribution of N- and N-thing in Koine Greek

Jespersens unite: a new definition



of negative elements. As a result, no discontinuity stage occurred, given also that the shift to underived VSO in Hellenistic Koine (Blass and Debrunner : , Horrocks , Taylor , Aitchison , () : –, –, Haug , Horrocks : –, Deligianni b) only facilitated the upward reanalysis of the negative indefinites οὐδέν /udhén/ and μηδέν /midhén/ (see section .). Figure . repeated as . manifests once more the N indefinite reanalysis in connection to word-order change (which can also be seen as a diachronic upward movement of the verb). The developments of the μηδέν /midhén/ indefinite were up to a point the same and affected by the same development in the underlying word order. It is anticipated that a change in one aspect of the language can have repercussions for other aspects, a fact we already observed in the phonological developments of Koine Greek that caused the syntactic emergence of nonveridicality through the semantic broadening of the ἵνα /ína/ particle. But the regulating role of vertical structure in language change, as we have seen with the occasion of the Greek Jespersen’s Cycle, is something that cannot be overstated. This remark has been preceded in a more general sense in Joseph ([] ). the surface syntactic patterns of a language are what many linguists have turned to in order to find a motivation for syntactic change. However, the changes discussed here show that a motivation for syntactic changes can be found which is internal to the language system itself, i.e. a change in one component can trigger a change in another component, under certain conditions. (Joseph [] : )

As a result, language change provides independent and direct evidence for hierarchical structure and for processes of natural language that are otherwise opaque and abstract. Overall, the twofold definition we proposed for Jespersen’s Cycle highlights the semantic dimension of the phenomenon. The syntactic repercussions of these developments can differ from language to language, given the multiplicity of the parameters that are usually involved. Several schematic representations of the Jespersen’s Cycle processes have been presented in the literature (van der Auwera , a, , van der Auwera and Gybels , van Alsenoy and van der Auwera , Vossen and van der Auwera ). In the final part of this section I introduce one more schematic representation of Jespersen’s Cycle for the French ne . . . pas and for the Greek οὐδέν /udhén/, in a way that manifests these transformations as leftward lexical micromovement (Figures . and .). The meaning associated with lexical items and structures is represented through little balloons attached to the words and structures, in particular regarding what van der Auwera et al. () describe as negative contamination of neighboring lexical elements: red balloons indicate negative meaning, green balloons indicate the meaning of entity , what can be described as ‘nounness’ of the n-word in its first stage, both in the



Renewal and stability (Classical Greek: OV)

NegP

Neg’

VP

Neg Ø NP

V

[uNeg] οὐδέν /u:den/ Neg1-thing εἶδον /e:don/ see.aor.ind.1sg ‘I saw nothing.’ NegP

(Koine/Medieval Greek: VO)

Neg’

Neg

VP

V

(NP)

οὐδέν /udhén/ Neg1 εἶδον /ídhon/ see.aor.ind.1sg

‘I didn’t see.’

FIGURE . N-thing reanalysis from Attic Greek to Koine and Medieval Greek

Greek οὐδέν /udhén/ and the French pas, while the little ladder indicates scalarity that tends to gradually fade. Loss of association with results in the transformation of the n-word to a negative adverb, while loss of association with the scale signals the transformation of the structure to mere propositional negation. The two-tier representation, inspired by viewpoints of Goldsmith (, ) and Sadock (, ) in phonology and syntax respectively (e.g. the notion of mismatch), is intended to provide a visualization of Jespersen’s Cycle procedures as well as what is called the



Jespersens unite: a new definition

Neg

Neg

ne

ne

pas

The French Jespersen's Cycle

Neg

Neg

ne

pas

ne

pas

Neg

ne

ne

Neg

pas

pas

ne

pas

Neg

ne

Neg

ne

Neg

pas

ne

pas

Neg

Neg

pas

pas

FIGURE . Leftward lexical micromovement in the French Jespersen’s Cycle

‘Jespersen’s argument cycle’ in Ladusaw () or ‘Quantifier cycle’ in Willis et al. (), among others.4 The micro- part of our ‘micromovement’ terminology captures the gradualness of the changes described, which is in agreement with outlooks on grammaticalization 4

See also Larrivée () and van der Auwera et al. () for discussion.



Renewal and stability

Neg

Neg

The Greek Jespersen's Cycle uk

Neg

u-dh-én not-even-one Reinforcement strategy 2

Neg

uk udhén Reinforcement strategy 1

udhén not-even-one

Neg

udhén

Negative adverb stage

Neg

udhén

Neg

udhén

Neg

udhén

Neg

Neg

Neg

udhén

udhén

dhen

FIGURE . Leftward lexical micromovement in the Greek Jespersen’s Cycle

and language change that highlight the gradual nature of the phenomenon (Lichtenberk , Haspelmath , Hopper and Traugott , C. Lehmann , Lightfoot ; but see Lightfoot  for language change as ‘catastrophic’), which can easily be captured through the notion of mismatch (Sadock , Sadock and Schiller , Sadock ). Yet even within the generative perspective, the position that a minimalist outlook on syntactic change can only be abrupt (described

The diachronically stable uses of N



also as ‘saltational’) and not incremental, as suggested in Kinsella (: –) has been recently dismissed in Clark (), while Roberts (, ), as well as Traugott and Trousdale (b), discuss such micro-operations and support a view for language change as involving a number of micro-steps that can eventually have a macro effect. This representation also captures the intuition behind Zeijlstra’s (, ) analysis on negation being either [iN] or [uN] as a parameter of crosslinguistic variation, as well as the notion of referential deficiency of negative polarity items (Giannakidou ) as an anticipated stage of semantic change. This much on the aspects of the Greek negator system that changed, most importantly on the transformations of N that resulted to a complete Jespersen’s Cycle in the proposed sense. In the next section I turn to the functions of N that have remained remarkably persistent. Minor changes may have taken place, because of the developments in the overall language structure, but the C-related uses of N have remained the same as those of Attic Greek, attested as early as Homeric (eighth century ).

. The diachronically stable uses of N The Greek N preserved its polarity behavior in each linguistic stage, as an element dependent on nonveridicality in the sense of Giannakidou (), a notion for which a distinct syntactic projection was proposed in chapter . According to this analysis the Greek N, like all negative polarity items, carries an uninterpretable Nonveridicality feature [uNV], which is checked against the overt or covert Nonveridicality head introduced by such operators as the question operator, negation, the conditional, modal propositional attitudes, etc. (see chapter  and chapter ). The exact distribution of the Greek N has not remained the same from Classical Greek to Late Medieval and Modern Greek, mainly because the language system itself has undergone alterations, most prominently regarding the decrease of nonfinite forms. The complete loss of all kinds of infinitival constructions after the sixteenth century (Joseph [] : –), resulted in the complete loss of infinitival negation and the severe reduction of forms in the participial paradigms that took place during Hellenistic and Medieval times had a similar effect. The only genuine alteration in the distribution of N—which nevertheless has no impact on its status as a polarity item—is its exclusion from the conditional protasis and substitution by the unmarked N οὐ(κ) /u(k)/ and later on (οὐ)δέν /(u)dhén/, a change that had already been generalized by Late Medieval times (see section .. for discussion). Other than maintaining its polarity behavior, throughout the history of the Greek language, the N has four diachronically persistent functions, from Homeric Greek to Standard Modern: its uses (i) in negative directives or volitionals in general, either unembedded (most significantly in prohibition) or embedded under directive or volitional predicates, (ii) as an optional question particle, (iii) embedded in the



Renewal and stability

scope of verbs of fearing, and (iv) as lexical negation (DP internal negation). The lexical negation function, although very vibrant in Modern Greek, is actually very poorly attested in the Late Medieval and Early Modern Greek era and only with the two remaining participial forms (active and mediopassive), a use in which it had been generalized from a very early stage, as early as Hellenistic Greek, while it was preferred as DP internal negation even in Classical times (see chapter ). The use of N as lexical negation in Standard Modern Greek was in fact reinforced through semantic borrowing and transliteration of the French and English non, e.g., non smoker (see Efthimiou ), as a result of the extensive borrowing from these two languages that took place at the beginning of the twentieth century (cf. AnastassiadisSymeonidis , ). It was, however, preferred in this use to the Modern Greek N δεν /dhen/, which is completely ungrammatical as lexical negation.5 The reason for this preference seems to have been the diachronic correlation of N with participials, which are, however, CP projecting elements (see chapter ). The generalization we observe is that in all its diachronically stable (and not externally induced) uses, N is located or linked to C, a fact that is established for Standard Modern Greek (see Roberts and Roussou , Roussou ). In the following sections the first three stable uses of N are discussed with data from all stages of Greek. In two of them (the question particle use and in the scope of timendi predicates), N is nonnegative. As a result, it did not experience the renewal pressures that Jespersen’s Cycle predicts. In its negative use in prohibition it did manifest a renewal tendency during Late Medieval times, as discussed in the previous sections. This renewal, though, was not completed, most likely owing to inertial forces applied by the two nonnegative, yet C related, functions of N that did not change. .. Negative directives ... Unembedded: N in prohibition The use of N μη /mi/ in prohibition is the statistically most frequent use in all stages (Figure .). This is in agreement with the crosslinguistic preference regarding the availability of negators that are dedicated to prohibition and often other nonveridical functions (% of a sample of  languages from around the world, see van der Auwera and Lejeune  and the map in chapter ). The origin of this function—and form—goes back to the Proto-Indo-European *meH (Fowler , Joseph , , Fortson ) and Greek is among the few Indo-European languages that preserved it (see section ..). Examples () through () contain instances of the use of N in unembedded directives from all stages of the attested history of Greek, including Homeric Greek (eighth century ), Early Medieval Greek (sixth century ) and Early 5 Cf. also the use of the Albanian N mos as lexical negation, generally as a prefix, instead of the Albanian N s’ or nuk, e.g. mosbarazi ‘inequality’, from N mos and barazi ‘equality’, see Joseph (a).

The diachronically stable uses of N



In conditional protasis

With conditional participle Relative clause In purpose/result clause Selected by veridical predicate

With infinitive in veridical scope Causal/modal participle Particle introducing yes/noquestion and timendi predicates Selected by nonveridical predicate

With infinitive in nonveridical scope Unembedded directives DP internal/lexical 0

20

40

60

80

100

120

Neg2 in Late Medieval Neg2 in Koine Greek

FIGURE . The distribution of N μη /mi/ in Attic Greek, in Koine, and in Late Medieval Greek

Modern Greek (eighteenth century ), which are not, however, part of the quantitative data, as discussed in chapter . Homeric and Early Medieval Greek do not reflect a spoken variety, while Early Modern Greek is too close to Standard Modern to be studied independently, at least for our purposes. Yet the presence of N as the negator of prohibitions is very strong in all (cf. Willmott  for Homeric Greek and Markopoulos  for Early Medieval). () ἐξαύδα, μὴ κεῦθε νόῳ. eksauda, mε: keuthe noo:i speak... N hide.. mind. ‘Speak out, do not hide it in your mind.’ (Iliad .; th century )



Renewal and stability

() μὴ φάθι. mε: phathi N speak.. ‘Do not say (that).’ (Plato, Gorgias d ; th century ) () μὴ πολλῶν ἐπιθύμει. mi polón epithými N many. desire... ‘Do not desire many things.’ (Epictetus, Dissertationes ab Arriano digestae ...; st century ) () μὴ ἀπατάτω mi apatáto N deceive... ‘Do not let Ezekias deceive .–; th century )

ὑμᾶς Ἐζεκίας λόγοις. ymás Ezekías lóghis you.. Ezekias. words. you with words.’ (Malalas, Chronographia

() μὴ φοβηθῆς τὸν θάνατον παρὰ μητρὸς κατάραν. mi fovithís ton thánaton pará mitrós katáran N fear.. the. death. but mother. curse. ‘Do not fear death, but a mother’s curse.’ (Digenis Akritis ; th century ) () μὴ ὀργισθῆς το σφάλμα μου, μὴ τόσον μὲ mi orghisthís to sfálma mu, mi tóson me N be.angry.. the fault my N that me. κακίσης. kakísis be.angry.. ‘Do not be angry at me because of my fault, do not be that angry at me.’ (Livistros and Rodamne ; th century ) () Για σήκω απ’ αύτου Γιώτη μου και μη βαριά Ja síko ap áftu Jóti mu ke mi varjá Let rise... from there Jótis my and N heavily κοιμάσαι. kimáse sleep.. ‘Rise from there, dear Yiotis and don’t sleep deeply.’ (Fauriel, , ; th century ) The only development in the use of N in unembedded directives is that roughly after the twelfth century N can no longer cooccur with morphological imperatives, a shift that in section .. I argued was linked to the change in syntactic status of N, from adverbial to capital. This is the status that N maintains in

The diachronically stable uses of N



Modern Greek as well, at least in negative directives (see Giannakidou ), a fact that again falls from the Head Preference Principle of van Gelderen (), discussed in section ... ... Embedded: N selected by directive predicates The members of this category are grouped in figure . as attestations of N in the scope of a nonveridical predicate, among which are directive and volitional predicates in general. The clause that is introduced by N is selected by the embedding form as an internal argument. () ὅρα μὴ γελοῖον ᾖ σὲ ταῦτα hora mε: geloion ε:i se tauta look... N laughable be... you. these λέγειν lege:n say.. ‘See that it be not absurd for you to say these. (Demosthenes, Pro Formione .–; th century ; translation based on Murray ) () σκόπει οὖν μὴ τὸ φῶς τὸ ἐν σοὶ σκότος skópi un mi to phos to en si skótos think... thus N the light the in you darkness ἐστίν. estín is... ‘Make sure that the light inside you is not darkness.’ (Νovum Testamentum, Secundum Lucam ..–; st century ) () βλέπε μη εμπλέξης εις δεσμόν ερωτικής vlépe mi empléksis is dhesmón erotikís look... N entangle.. in bond amorous αγάπης. aghápis love. ‘Make sure that you do not get involved in an amorous relationship.’ (Livistros and Rodamne ; th century ) () Πρόσεχε μην πέσεις. prosehe min pésis look.out... N fall.. ‘Be careful not to fall. (Be careful lest you fall.)’

(Standard Modern Greek)

In some cases these clauses seem like reduced purpose clauses (cf. in English be careful not to fall and be careful so that you don’t fall). The difference is that clauses of



Renewal and stability

this kind, that are introduced by N, receive a theta role from the matrix verb, whereas purpose clauses are unselected. In this sense such phrases seem to be frozen in an intermediate crossclausal grammaticalization stage (see chapter ); they are clearly selected, although a possibly former (Proto-Greek) unselected reading (as purpose clauses) can still be available. These embedded directive attestations of N are not always easy to distinguish from the embedded question particle uses of N discussed in the next section. The difference, however, is that N in embedded directive clauses contributes negativity—indicated in the translations of the examples above— whereas in the embedded question particle uses N is nonnegative. Yet all of them are C-related. .. The diachronic nonnegative uses of N Among the functions of N that have been uninterruptedly attested and can actually be traced back as early as Homeric Greek (eighth century ) are its nonnegative complementizer-related functions: (i) as a particle introducing yes/no questions, and (ii) as a complementizer selected by timendi predicates (verbs of fearing and the like). Being nonnegative, N in these functions did not undergo a Jespersen’s Cycle, as it was not linked to a lower element, the structural elevation of which drives the Jespersen’s Cycle. Thus, the renewal pressures dictated by the Jespersen’s Cycle had no effect on the nonnegative uses of N. In fact, it was most likely these two functions that hindered the generalization of Late Medieval μηδέν /midhén/ as a negator of nonveridical environments. The loss of negativity in a negator and its structural elevation to C positions, as is the case in these functions of the Greek N, is a crosslinguistically attested development (Heine and Kuteva : , Aldridge , van Gelderen : , –). In Greek both these uses of N tend to have an undesirability flavor, but in the question particle use this is not general. A suggestion for explaining the emergence of the nonnegative functions of N is presented that relies on the Cinque () expanded CP hypothesis, which may be useful in the examination of similar compositionally vacuous negatives, as in the case of expletive/paratactic negation (van der Wouden , ) or evaluative negation (Yoon ). ... Particle introducing yes/no questions This section provides examples of the use of N μη as a nonnegative question particle from Homeric Greek (eighth century ) until Early Modern Greek (eighteenth century ). This function of N was identified and discussed in chapter  in the examination of the different specializations N and N have in Attic Greek, particularly in rhetorical questions (section ..). The situation is the same for the following stages: N can introduce yes/no questions, either in requesting information or as a dubitative. Representative examples from the diachrony of Greek follow.

The diachronically stable uses of N



() μή πού τινα δυσμενέων φάσθ’ ἔμμεναι mε: pu: tina dysmeneo:n phasth emmenai N maybe someone enemy.. say. be. ἀνδρῶν; andro:n man.. ‘Do you think he could be an enemy?’ (Odyssea .; th century ) ()

-Μή οὖν ἐγὼ ληρῶ; mε: u:n ego: lεro:? N thus I speak-nonsense../. ‘Am I speaking nonsense?’ (Plato, Theaetetus d ; th century )

() μὴ πάντες; ἀπόστολοι μὴ πάντες προφῆται; mi pántes apóstoli? mi pántes prophíte? N all apostles. N all prophets. ‘Are all apostles? Are all prophets?’ (Novum Testamentum, Ad Corinthios I ..–.; st century ; translation by Senior et al. ()) () μὴ φύλαξ τοῦ ἀδελφοῦ μού εἰμι ἐγώ; mi fýlaks tu adhelfú mu imí eghó N guard the. brother. my be... I ‘Am I the guardian of my brother?’ (Chronicon paschale .–; th century ) () Μὴ τοῦτος εἶν’ τὸν λέγουσιν ὁ Διγενὴς Mi tútos in’ ton léghusin o Dighénis N he is whom call... the. Dighenis. Ἀκρίτης; Akrítis? Akritis. ‘Is he the one they call Digenis Akritis?’ (Digenis Akritis ; th century ) () Πάντως μὴ ἔπαθες κακόν, πάντως μὴ pántos mi épathes kakón pántos mi somehow N suffer... wrong somehow N ἐφαντάχθης; efantáxthis see.ghost... ‘Did something bad happen to you? Did you happen to see a ghost?’ (Livistros and Rodamne ; th century )



Renewal and stability

() Μην είδατε τον άντρα μου τον Λούκα min ídhate ton ándra mu ton Lúka N see.. the. husband. my the. Lukas. Καλιακούδα; Kaljakúdha Kaljakudhas. ‘Did you happen to see my husband, Lukas Kaliakudas?’ (Fauriel (–), .; th century ) This attestation of N can be analyzed as movement to a higher position on Cinque’s () hierarchy, which was discussed in section .. in relation to the loss of N from the conditional antecedent in Medieval Greek. This development in Greek may be explained through reference to Cinque’s elaborate ordering of functional projections, as an instance of upward reanalysis in the sense of Roberts and Roussou (). The labels of ‘Mood’ and ‘Mod(ality)’ in terms of our viewpoint should be understood merely as higher C positions, since we have established the independence of N both of morphological mood and of modality.6 ()

The upward reanalysis of the Greek N μή /mε:/ on Cinque’s hierarchy

Question particle

Dubitative

MoodSpeechAct MoodEvaluative MoodEvidential ModEpistemic T(Past) T(Future) MoodIrrealis ModNecessity

ModPossibility AspHabitual AspRepetitive(I) AspFrequentative(I) AspCelerative(I) ModVolitional ModObligation ModAbility/Permission AspCelerative(II) T(Anterior) AspTerminative AspContinuative AspPerfect(?) AspRetrospective AspProximative AspDurative AspGeneric/progressive AspProspective AspSgCompletive(I) AspPlCompletive Voice AspCelerative(II) AspSgCompletive(II) AspRepetitive(II) AspFrequentative(II) AspSgCompletive(II)

As noted in chapter , the Cinque hierarchy has already found application in explaining diachronic syntactic/semantic change (see Roberts ). Yet given that we have no evidence from an earlier stage in which N was not already in this position, we cannot strongly argue that this grammaticalization path is the path out of which the question particle use of N emerged. It is a plausible scenario, because based on frequency of use (see figure . for Attic), the negative functions of N must have chronologically preceded the nonnegative. Taking this into account, it must have been the negative uses of N that gave rise to the nonnegative, and not the other way around. However, most frequent among the negative functions of N is its prohibitive function, most likely base-generated at the topmost position (MoodSpeechAct) on Cinque’s hierarchy. Thus the question particle function of N resulted either through micromovement of N from MoodIrrealis to MoodSpeechAct or from loss of negativity in its prohibitive use, which was already in 6 All three (N, mood, and modality) depend on Nonveridicality, but the distribution of N cannot be reduced to mood or modality considerations only.

The diachronically stable uses of N



MoodSpeechAc. In lack of evidence from an earlier stage we can only identify possible pathways for grammatical change, but crosslinguistic considerations point to the first solution: micromovement of from MoodIrrealis to MoodSpeechAc. ... Complementizer in the scope of verbs of fearing (verba timendi) Similar considerations can be made for the diachronic function of N in introducing timendi predicates: they may be the outcome of micromovement of N from MoodIrrealis to MoodSpeechAct or MoodEvaluative that must have taken place at an earlier stage (Proto-Greek or even Pre-Greek). Examples on these uses starting from Homeric Greek follow and the Cinque hierarchy is repeated in () with the relevant indications added. () ἀμφιτρομέω καὶ δείδια μή τι amphitromeo: kai de:dia mε: ti tremble... and fear... N something πάθῃσιν. pathε:isin suffer... ‘I tremble and fear lest something happens to them.’ (Odyssea .; th century ) () ἡμεῖς δέδοικα μὴ τἀναντία πράττοντες hε:me:s dedoika mε: tanantia prattontes we. fear... N the.opposite do... φανῶμεν. phano:men seem.. ‘I fear that we may seem to have pursued the opposite.’ (Isocrates, Archidamus .–; th century ) () σὺ . . . τρέμεις, μή σοι λείπῃ sy trémis mi si lípi you tremble... N you. lack... τὰ ἀναγκαῖα. ta anagéa the necessary ‘You tremble lest you lack the things that are necessary to you.’ (Epictetus, Dissertationes ab Arriano digestae ...; st century ) () δεδοίκαμεν μὴ πάλιν πάθῃ ἡ πόλις dhedhíkamen mi pálin páthi I polis fear... N again suffer... the. city. ἡμῶν. imón our. ‘We feared that our city would suffer again.’ (Malalas, Chronographia .–; th century )



Renewal and stability

() δέδοικα μὴ φονευθῶ πρὸ ὥρας. dhédhika mi fonefthó pro óras fear... N be.killed.. before time ‘I fear that I may be killed prior to my time.’ (Ptohoprodromos I. ; th century ) () Ο Γιάννης φοβάται μην αρρωστήσει. ο Jánis fováte min arostísi the Janis fear.. N get.sick.. (Standard Modern Greek) ‘John is afraid that he may get sick.’ ()

The upward reanalysis of the Greek N μή /mε:/ on Cinque’s hierarchy with verba timendi

Question particle

Dubitative (also QP)

MoodSpeechAct MoodEvaluative MoodEvidential ModEpistemic T(Past) T(Future) MoodIrrealis ModNecessity ModPossibility AspHabitual AspRepetitive(I) AspFrequentative(I) AspCelerative(I) ModVolitional ModObligation ModAbility/Permission AspCelerative(II) T(Anterior) AspTerminative AspContinuative AspPerfect(?) AspRetrospective AspProximative AspDurative AspGeneric/progressive AspProspective AspSgCompletive(I) AspPlCompletive Voice AspCelerative(II) AspSgCompletive(II) AspRepetitive(II) AspFrequentative(II) AspSgCompletive(II)

Although the undesirability flavor in the use of N with timendi predicates arguably comes from the lexical semantics of the selecting verb form and world knowledge (the things we fear are generally undesirable), the MoodEvaluative seems like a plausible landing site for N in this use. The MoodSpeechAct could also be a likely candidate, under a perception for Speech Act merely as C information, otherwise I would be hesitant to assign illocutionary force to N in a function like this one which is always embedded.

. Summary In this final chapter of our investigation the diachronic transformations of the Greek N and N particles have been examined and their individual paths discussed. A connection between the Greek negator developments and Jespersen’s Cycle was made, in a way that required a different understanding for the latter. A new approach for Jespersen’s Cycle was proposed by placing the regularity of the phenomenon in the semantics, in order to properly include Greek in the Jespersen’s Cycle languages, as well as other languages, which, like Greek, did not manifest a doubling stage in their history for the expression of plain sentential negation. The exact manifestation of Jespersen’s Cycle in a language is the outcome of multiple parameters (e.g., underived word order at each stage). The defining property of the phenomenon,

Summary



however, is semantic: Jespersen’s Cycle is a result of semantic change, as is grammaticalization in general (Lightfoot ). It does have morphosyntactic and phonological repercussions, but these are language-particular. In Greek only N underwent Jespersen’s cycle according to the proposed sense. N reached a stage of renewal, but was eventually hindered by its two diachronic nonnegative functions which—being nonnegative—remained inert: the question particle function of N and the use of N in introducing verba timendi complements. These functions go back to Homeric Greek (and possibly earlier). An account that explains the transformations of N to C-related elements was proposed that relies on Cinque’s () cartographic approach and the tendencies of diachronic upward reanalysis of elements in syntactic structures (Roberts and Roussou , van Gelderen ). The unifying property of all environments of N, negative and nonnegative, in all stages of the Greek language, is the semantic property of nonveridicality.

Conclusion We have tracked the history of the Greek negator system through a timespan of over . thousand years, through quantitative evidence from three major stages of spoken Greek: Classical Greek, Koine Greek, and Late Medieval Greek and qualitative evidence from Homeric Greek up to Standard Modern. The presence of a second negative element, N, that is particular to nonveridical environments in the sense of Giannakidou ( et seq.) is diachronically stable. A syntactic account for the attestation of N was proposed, in terms of a Nonveridicality syntactic projection (NVP). This was supported through general developments that took place in the history of Greek, certain phenomena within each stage (e.g. the independence of nonveridical marking from morphological mood), while it is also supported by the presence of nonveridical negators crosslinguistically, within and outside the IndoEuropean language family. The Nonveridicality phrase also relates to evidentiality, while typologically it can be represented as overlapping with mood or overlapping with C (cf. Matthewson, Davis, and Rullmann , for a comparable treatment of evidentiality as parasitic on other functional projections). Yet, in the history of Greek, it emerged as an independent projection in the transition of ἵνα /hina/ from purpose complementizer in Classical Greek to the subjunctive mood particle να /na/ by Late Medieval. Nonveridicality is in the in-between area. Explanations were offered for a number of other facts in the diachrony of Greek negation, in particular those that mark the transition from Koine Greek to Late Medieval and Standard Modern Greek: the transformation of the former complementizer ἴνα /hina/ into a subjunctive mood marker by the Late Medieval stage, the loss of True Negative Imperatives and the ban of N from the conditional antecedent. The explanations proposed agree with generative perspectives of regular diachronic change regarding the tendencies of (i) up-the-tree movement (Roberts and Roussou ; cf. also van Gelderen’s  Late Merge Principle), and (ii) syntactic status shifts from phrases to heads (van Gelderen , , ). Both tendencies agree with the transformations of negative elements in the history of Greek. The examination of the history of the Greek negation system by definition places this study within the Jespersen’s Cycle literature. The correlation of our findings with the Jespersen’s Cycle phenomenon has repercussions for the understanding of Negation and Nonveridicality in the History of Greek. First edition. Katerina Chatzopoulou. © Katerina Chatzopoulou . First published  by Oxford University Press.



Conclusion

the latter, as a diachronic tendency the regularities of which are to be found in the semantics. Given that Greek did not manifest a doubling stage regarding the expression of plain propositional negation, a broader definition was proposed for Jespersen’s Cycle, which accommodates not only Greek, but also other languages that deviate in their developments from what the traditional morphosyntactic account for Jespersen’s Cycle predicts (e.g., languages with negators of verbal origin, languages with tripling stage negators). An account for Jespersen’s Cycle was offered that includes two main stages: one in which an element is intensified predicate negation and another in which the same element is plain propositional negation. A number of microstages can be involved and this change was represented as upward lexical micromovement. The bleaching of intensified negation to plain as a defining property of Jespersen’s Cycle is an intuition that already underlies in-depth viewpoints on negation, diachronic change, and language typology (Horn , Roberts and Roussou , Kiparsky and Condoravdi , van Gelderen , van der Auwera , a, Vossen , van der Auwera et al. ). Our contribution was to state this intuition explicitly and provide a formalization. On the issue of the unidirectionality of diachronic change (for: Hopper and Traugott , Haspelmath , Heine and Kuteva , among others; against: Campbell , Joseph , , Newmeyer , Janda , Campbell and Janda , Frajzyngier , among others), our findings were mostly in support of this view, but one counter-example was also attested: the fact that the indefinite μηδέν /midhén/ of Classical and Koine Greek reassumed a lexical function in Modern Greek (meaning ‘zero’) after first having bleached to plain sentential negation of nonveridical environments in Late Medieval Greek. As for the case of grammaticalization in general, our view on unidirectionality, developed based on the present study, is that unidirectionality can be a tendency and not a principle, agreeing in this sense with Lightfoot (: –) that diachronic change is not in need of a theory, such as grammaticalization, in this understanding. Diachronic change is in need of description through the formal tools that synchronic linguistic theories have made available (cf. Labov ). It is from this description that the relevant regularities emerge. Not as pre-empirical principles, but as conspiratorial forces that can provide some explanation of and some insight into the complex phenomenon of language change.

Appendix of texts Texts examined for N and N distribution Attic Greek (fifth to fourth centuries ) Aristophanes, Lysistrata, Ranae Euripides, Alkistes, Hippolytus Lysias, In Eratosthenem, In Agoratum, De caede Eratosthenes Plato, Respublica IV Koine Greek (first century  to first century ) Strabo, Geographica VI–VIII Greek New Testament (Synoptic Gospels: Secundum Mattheum, Secundum Marcum, Secundum Lucam) Epictetus, Dissertationes ab Arriano digastae III Late Medieval Greek (twelfth to fourteenth centuries ) Digenis Akritis (Escorial manuscript) Livistros and Rodamne (Vatican version) Imperios and Margarona Chronicle of Moreas Ptoholeon

Texts examined for N-thing and N-thing distribution Attic Greek (fifth to fourth centuries ) Aristophanes, Aves, Equites, Lysistrata, Pax, Ranae Euripides, Alkistes, Electra, Hippolytus, Iphigenia Taurica, Medea Lysias, In Eratosthenem, In Agoratum, De caede Eratosthenes, In Diogitonem Plato, Respublica (a–a), Cratylus, Protagoras, Apologia Socratis, Phaedo, Euthydemus, Symposium Demosthenes, De corona, De Chersoneso, De falsa legatione Isocrates, Antidosis, De bigis, In sophistas, Panegyricus Sophocles, Philoctetes, Oedipus Tyrannus Koine Greek (first century  to first century ) Strabo, Geographica Greek New Testament (Synoptic Gospels: Secundum Mattheum, Secundum Marcum, Secundum Lucam, Epistulae Pauli [all but Ad Hebraeos]) Epictetus, Dissertationes ab Arriano digastae I–III



Appendix of texts

Late Medieval Greek (twelfth to fourteenth centuries ) Digenis Akritis (Escorial manuscript) Livistros and Rodamne (Vatican version) Imperios and Margarona Chronicle of Moreas Ptoholeon

References Primary Sources Attic Greek (fifth to fourth centuries ) Aeschines. Martin, V. and G. de Budé. –. Eschine. Discours, vols –. Paris: Les Belles Lettres. Aristophanes. Acharnenses, Equites, Pax, Aves, Lysistrata, Thesmophoriazusae, Ranae, Plutus: Coulon, V. and M. van Daele, (eds). . Aristophane, Paris: Les Belles Lettres; Ecclesiazusae: Ussher, R. G. (ed.) . Aristophanes. Ecclesiazusae. Oxford: Clarendon Press. Nubes: Dover, K. J. (ed.) . Aristophanes. Clouds. Oxford: Clarendon Press. Vespae: MacDowell, D. M. (ed.) . Aristophanes. Wasps. Oxford: Clarendon Press. Euripides. Murray, G. (ed.) . Euripidis fabulae. Oxford: Clarendon Press. Isocrates. Mathieu, G. and E. Bremond. . Isocrate. Discours. Paris: Les Belles Lettres. Lysias. Lamb, W. R. M. . Lysias (with English translation) Cambridge, MA: Harvard University Press; London: William Heinemann. Plato. John Burnet (ed.) . Plato. Platonis Opera. Oxford: Oxford University Press. Sophocles. Dain, A. and P. Mazon. . Sophocle. Paris: Les Belles Lettres. Xenophon. Marchant, E. C. –. Xenophontis opera omnia, vol. –. Oxford: Clarendon Press. Koine Greek (first century  to first century ) Epictetus. Schenkl, H. . Epicteti dissertations ab Arriano digestae. Leipzig: Teubner. Novum Testamentum Graece cum apparatu critico. curavit D. Eberhard Nestle, novis curis elaboravit D. Erwin Nestle adiuvante D. Kurt Aland, nd edition.. Stuttgart: Württembergische Bibelanstalt. P.Oxy. = The Oxyrhynchus Papyri. –, ed. Bernard P. Grenfell and Arthur S. Hunt, vols –. London: Egypt Exploration Fund. P.Oxy.Astr. = The Astronomical Papyri from Oxyrhynchus. , ed. A. Jones. Philadelphia, PA: American Philosophical Society. Strabo. Meineke, A. . Strabonis geographica, vols –. Leipzig: Teubner. Testamentum Abrahae (recensio A). James, M. R. . The testament of Abraham, Cambridge: Cambridge University Press. Early Medieval Greek (sixth– to seventh centuries ) Chronicon Paschalae. Dindorf, L. (ed.) . Chronicon paschal. Bonn: Weber. Joannes Malalas. Thurn, I. . Ioannis Malalae chronographia. Berlin: Mouton de Gruyter. Late Medieval Greek (twelfth– to fourteenth centuries ) Chronicle of Morea. Kalonaros, P. P. . Το Χρονικόν του Μορέως (To Hronikon tou Moreos). Athens: Dimitrakos.



References

Digenis Akritis. Alexiou, S. . Vasileios Digenis Akritis kai to Asma tou Armouri. Athens: Ermis. Jeffreys, E. M. and M. J. Jeffreys. ‘ “Imberios and Margarona”: the manuscripts, sources, and edition of a Byzantine verse romance’. Byzantion : –. (Reprinted in E. M. Jeffreys and M. J. Jeffreys (eds) . Popular literature in Late Byzantium. London: Variorum Reprints). Livistros & Rodamni (version b). Lendari, T. . Livistros and Rodamne: The Vatican version. Athens: ΜΙΕΤ. Ptoholeon. Kehagioglou, G. . Ptoholeon. Thessaloniki: MIET. Ptohoprodromika. Eideneier, H. . Ptochoprodromos (Neograeca Medii Aevi V). Cologne: Romiosini. Early Modern Greek (eighteenth– to nineteenth centuries ) Kleftika. Alexis Politis. . Το Δημοτικό Τραγούδι (To dimotiko tragoudi). Athens: Estia. Listrika. Dimitris Halatsas. . Τα Ληστρικά Τραγούδια (Ta listrika Tragoudia). Athens: Estia. Varlaam kai Ioasaf. Vasiliou, Panagiotis. . Βαρλαάμ και Ιωάσαφ: Η παραλλαγή του Κώδικα Ι  της Μονής της Μεγίστης Λαύρας. Thessaloniki: IMT. Translations Andocides. K. J. Maidment . Minor Attic Orators in two volumes , Antiphon Andocides, with an English translation. Cambridge, MA: Harvard University Press; London: William Heinemann Ltd. Aristophanes. Eugene O’Neill, Jr. . Women at the Thesmophoria. The Complete Greek Drama, vol. . New York: Random House. Aristotle. On interpretation. E. M. Edghill and W. A. Pickard-Cambridge. . Categories, On Interpretation and On Sophistical Refutations by Aristotle. Lawrence, KS: Digireads. Demosthenes. Norman W. DeWitt, and Norman J. DeWitt. Demosthenes, with an English translation. . Cambridge, MA: Harvard University Press; London: William Heinemann Ltd. Demosthenes. A. T. Murray. . Demosthenes, with an English translation. Cambridge, MA: Harvard University Press; London: William Heinemann Ltd. Demosthenes. C. A. Vince and J. H. Vince. . Demosthenes, with an English translation. Cambridge, MA: Harvard University Press; London: William Heinemann Ltd. Epictetus. Thomas Wentworth Higginson. trans. . The Works of Epictetus: His Discourses, in Four Books, the Enchiridion, and Fragments. Epictetus. New York: Thomas Nelson and Sons. Euripides. Whitney J. Oates and Eugene O’Neill, Jr (eds). . The Complete Greek Drama. in two volumes. . Helen, trans. E. P. Coleridge. New York: Random House. Isocrates. George Norlin. . Isocrates with an English Translation in three volumes. Cambridge, MA: Harvard University Press; London: William Heinemann Ltd. Lysias. W. R. M. Lamb. . Lysias, with an English translation. Cambridge, MA: Harvard University Press; London: William Heinemann Ltd. New Testament. Donald Senior, Mary Ann Getty, Carol Stuhlmueller, and John J. Collins. . The Catholic Study Bible: The New American Bible. New York–Oxford: Oxford University Press.

References



Plato. R. G. Bury (trans.).  and . Plato in Twelve Volumes, vols.  and . Cambridge, MA: Harvard University Press; London: William Heinemann Ltd. Plato. Harold N. Fowler (trans.). . Plato in Twelve Volumes, vol. . Cambridge, MA: Harvard University Press; London: William Heinemann Ltd. Plato. Harold N. Fowler (trans.). . Plato in Twelve Volumes, vol. . Cambridge, MA: Harvard University Press; London: William Heinemann Ltd. Plato. W. R. M. Lamb (trans.). . Plato in Twelve Volumes, vol. . Cambridge, MA: Harvard University Press; London: William Heinemann Ltd. Plato. W. R. M. Lamb (trans.). . Plato in Twelve Volumes, vol. . Cambridge, MA: Harvard University Press; London: William Heinemann Ltd. Plato. Paul Shorey (trans.). . Plato in Twelve Volumes, vols  and . Cambridge, MA: Harvard University Press; London: William Heinemann Ltd. Sophocles. Sir Richard Jebb (ed.). . The Antigone of Sophocles. Cambridge: Cambridge University Press. Sophocles. Sir Richard Jebb (ed.). . The Electra of Sophocles. Cambridge: Cambridge University Press. Strabo. H. L. Jones (ed.). . The Geography of Strabo. Cambridge, MA: Harvard University Press; London: William Heinemann, Ltd. Xenophon. Carleton L. Brownson.  (vol. ), , (vol. ). Xenophon in Seven Volumes,  and . Cambridge, MA: Harvard University Press; London: William Heinemann, Ltd. Xenophon. Carleton L. Brownson. . Xenophon in Seven Volumes, . Cambridge, MA: Harvard University Press; London: William Heinemann, Ltd. Xenophon. E. C. Marchant. . Xenophon in Seven Volumes, . Cambridge, MA: Harvard University Press; London: William Heinemann, Ltd. Xenophon. Trans. E. C. Marchant and O. J. Todd. . Xenophon in Seven Volumes, . Cambridge, MA: Harvard University Press; London: William Heinemann, Ltd. General Bibliography Acquaviva, P. . The Logical Form of Negation. A Study of Operator-Variable Structures in Syntax, PhD dissertation, Scuola Noimale Superiore. Pisa: Garland. Adrados, Francisco R. . A History of the Greek Language: from its Origins to the Present. Leiden–Boston: Brill. Aitchison, Jean. . The Order of Word Order Change. Transactions of the Philological Society (), –. Aitchison, Jean.  []. Language Change Progress or Decay? Cambridge: Cambridge University Press. Aken, Adolf Friedrich. . Die Grundzüge der Lehre von Tempus und Modus im Griechischen. Rostock: Stiller. Aldridge, Edith. . Neg-to-Q: The Historical Origin and Development of Question Particles in Chinese. The Linguistic Review , –. Alexiadou, Artemis and Elena Anagnostopoulou. . Parametrizing Agr: Word Order, V-Movement and EPP-Checking. Natural Language and Linguistic Theory , –.



References

Alexiadou, Artemis, Liliane Haegeman, and Melita Stavrou. . Noun Phrase in the Generative Perspective. Berlin: Mouton de Gruyter. Allan, Keith. . Linguistic Meaning. London and New York: Routledge & Kegan Paul. Allen, W. Sidney.  []. Vox Graeca. Α Guide to the Pronunciation of Classical Greek, Cambridge: Cambridge University Press. Alonso-Ovalle, Luis and Elena Guerzoni. . Double Negatives, Negative Concord and Metalinguistic Negation. Chicago Linguistics Society  (). The Main Session –. Anastassiadis-Symeonidis, Anna. . Η νεολογία στην κοινή νεοελληνική. (Neology in Standard Greek). PhD dissertation, Aristotle University of Thessaloniki. Anastassiadis-Symeonidis, Anna. . Νεολογικός δανεισμός-Άμεσα δάνεια από τη γαλλική και την αγγλοαμερικανική-Μορφοφωνολογική ανάλυση. (Neological loanwords – direct loans from French and American English – Morphophonological analysis.) Thessaloniki: selfpublished. Anastassiadis-Symeonidis, Anna and Katerina Chatzopoulou. . Changement de frontières morphémiques et création suffixale. XXe Colloque International des Neo-hellenistes des universites francophones, –. Université Charles de Gaulle—Lille , Villeneuve d’Ascq. Anastassiadis-Symeonidis, Anna and Katerina Chatzopoulou. . Why does a Language Borrow Suffixes? The Case of Greek and Latin. . In Proceedings of the th International Conference on Greek Linguistics, ed. Katerina Chatzopoulou, Alexandra Ioannidou, and Suwon Yoon, –. e-publication. Columbus, OH: Ohio State University. [Online.] Available at http://www.ling.ohio-state.edu/ICGL/proceedings.html. Anderwald, Liselotte. . Negation in Non-Standard British English. Gaps, Regularizations, and Asymmetries. London: Routledge. Andrews, Edna. . Markedness Theory: The Union of Asymmetry and Semiosis in Language. Durham, NC: Duke. Andriotis, Nikolaos. . Die Ausdrucksmittel für “gar nichts”, “ein wenig” und “sehr viel.” in Alt-, Mittel und Neugriechischen. Byzantinisch-neugriechische Jahrbücher , –. Andriotis, Nikolaos. . Ετυμολογικό Λεξικό της Κοινής Νεοελληνικής (Etimologiko lekiko tis Koinis Neoellinikis). Thessaloniki: Institouto Neoellinikon Spoudon. Anton, C. T. . De discrimine particularum οὔ et μή brevis dissertatio. Archiv für Philologie und Pädagogik , –. Arad, M. and Anna Roussou. . Particles and C-positions in Classical Greek. Unpublished ms., University College London and University of Wales, Bangor. Arndt, William F. and F. Wilbur Gingrich. . A Greek-English Lexicon of the New Testament and Other Early Christian Literature. Chicago, IL: The University of Chicago Press. Ashby, W. J. . The Loss of the Negative Morpheme, ne, in Parisian French. Lingua , –. Ashby, W. J. . The Loss of the Negative Particle, ne, in French: A Syntactic Change in Progress. Language , –. Babiniotis, Georgios. . Συνοπτική ιστορία της ελληνικής γλώσσας. Athens: Ellinika Grammata. Bach, Kent and Robert M. Harnish. . Linguistic Communication and Speech Acts. Cambridge, MA: MIT Press.

References



Baker, Carl L. . Double Negatives. Linguistic Inquiry (), –. Bakker, Dik. . Formal and Computational Aspects of Functional Grammar and Language Typology. Studies in Language and Language Use . Amsterdam: IFOTT. Barr, James. . The Semantics of Biblical Language. Oxford: Oxford University Press. Bardovi-Harlig, Kathleen. . Markedness and Salience in Second-Language Acquisition. Language Learning , –. Barsalou Lawrence. . The Instability of Graded Structure in Concepts. In Concepts and Conceptual Development: Ecological and Intellectual Factors in Categorization, ed. Neisser Ulric, –. New York: Cambridge University Press. Basset, Louis. . La syntaxe de l’imaginaire : Étude de modes et des négations dans l’Iliade et l’Odyssée. Lyon: Maison de l’ Orient. Battistella, Edwin. . Markedness: the evaluative superstructure of language. Albany, NY: SUNY Press. Bauer, W., W. F. Arndt, F. W. Gingrich, and F. W. Danker. . A Greek-English Lexicon of the New Testament and other Early Christian Literature, nd edition. Chicago, IL: University of Chicago Press. Bäumlein, Wilhelm Friedrich Ludwig von. . Untersuchungen über die griechischen Modi und die Partikeln κέν und ἄν. Heilbronn: Landherr. Bäumlein, Wilhelm Friedrich Ludwig von. . Untersuchungen über griechische Partikeln. Stuttgart: J. B. Metzler. Bech, Gunnar.  []. Studien über das deutsche Verbum Infinitivum. Tübingen: Niemeyer. Bech, Kristin. . Word order patterns in Old and Middle English: A syntactic and pragmatic study. Doctoral dissertation, University of Bergen. Beck, Hans-Georg. . Geschichte der byzantinischen Volksliteratur (Byzantinisches Handbuch II. [HdAW XII..]), Munich: Beck. Beekes, Robert Stephen Paul. . Comparative Indo-European Linguistics. An Introduction. Amsterdam-Philadelphia, PA: John Benjamins. Beekes, Robert Stephen Paul. . Etymological Dictionary of Greek. Leiden: Brill. Benveniste, Emile. . Mutations of linguistic categories. In Directions for Historical Linguistics, ed. Winfred Lehmann and Yakov Malkiel, –. Austin, TX: University of Texas Press. Berg, Wolfgang. . Uneigentliches Sprechen: Zur Pragmatik und Semantik von Metapher, Metonymie, Ironie, Litotes und Rhetorischer Frage. PhD dissertation, Tubingen: TBL. Bernardi, Raffaella. . Reasoning with Polarity in Categorial Type Logic. PhD dissertation, University of Utrecht. Bernini, Giuliano and Paolo Ramat. . La frase negativa nelle lingue d’Europa. Bologna: Il Mulino. Bernini, Giuliano and Paolo Ramat. . Negative Sentences in the Languages of Europe: A Typological Approach. Berlin: Mouton de Gruyter. Beukema, Frits. . Five Ways of Saying No: The Development of Sentential Negation in English in a Government and Binding Perspective. In Negation in the history of English, ed. Ingrid Tieken-Boon van Ostade, Gunnel Tottie, and Wim van der Wurff, –. Berlin: Mouton de Gruyter.



References

Biberauer, Theresa. . Doubling and Omission: Insights from Afrikaans Negation. In Microvariation in Syntactic Doubling, ed. Sjef Barbiers, Margreet van der Ham, Olaf Koeneman, and Marika Lekakou, –. Bingley: Emerald. Biberauer, Theresa. . Jespersen Off Course? The Case of Contemporary Afrikaans Negation. In Cyclical change, ed. Elly van Gelderen, –. Amsterdam: John Benjamins. Biberauer, Theresa and Ian Roberts. . Negative Words and Related Expressions: A New Perspective on Some Familiar Puzzles. In Evolution of Negation. Beyond the Jespersen Cycle ed. Richard Ingham and Pierre Larrivée, –. Berlin: Mouton de Gruyter. Biberauer, Theresa and Hedde Zeijlstra. . Negative Concord in Afrikaans: Filling the Typological Gap. Journal of Semantics (), –. Blass, Friedrich and Albert Debrunner. . A Greek Grammar of the New Testament and Other Early Christian Literature, trans. R. W. Funk. Chicago, IL: University of Chicago Press. Bolinger, Dwight. . Degree Words. The Hague: Mouton de Gruyter. Bonifazi, Anna, Annemieke Drummen, and Mark de Kreij. . Particles in Ancient Greek Discourse: Five Volumes Exploring Particle Use across Genres. Washington DC: Center for Hellenic Studies. Borkin, Ann. . Polarity Items in Questions, Chicago Linguistic Society , –. Borschev, Vladimir, Elena V. Paducheva, Barbara H. Partee, Yakov Testelets, and Igor Yanovich. . Russian Genitives, Non-Referentiality, and the Property-Type Hypothesis. In Formal Approaches to Slavic Linguistics: The Stony Brook Meeting  (FASL ), ed. Andrei Antonenko et al., –. Ann Arbor, MI: Michigan Slavic Publishers. Brandtler, Johan. . The Evaluability Hypothesis: The Syntax and Semantics of Polarity Item Licensing in Swedish. Amsterdam: John Benjamins. Bréal, Michel. . Essai de Sémantique. Paris: Hachette. Bréal, Michel. . Semantics. New York: Henry Holt & Co. [Bréal  trans. Mrs Henry Cust]. Breitbarth, Anne. . A Hybrid Approach to Jespersen’s Cycle in West-Germanic. The Journal of Comparative Germanic Linguistics , –. Breitbarth, Anne and Liliane Haegeman. . Not continuity, but change: stable stage II in Jespersen’s cycle. Ms, University of Cambridge and Université de Lille III. Brixhe, Claude. . Le déclin du dialecte crétois: essai de phénoménologie. In Dialectologica Graeca. II coloquio internacional de dialectología griega, ed. E. Crespo, J. L. García Ramón, and A. Striano, –. Madrid. Brixhe, Claude. . Phonétique et phonologie du grec ancien. Louvain: Peeters. Brixhe, Claude. . Linguistic Diversity in Asia Minor during the Empire: Koine and NonGreek Languages. In A Companion to the Ancient Greek Language, ed. Egbert J. Bakker, –. Oxford: Blackwell. Brown, Roger and Albert Gilman. . The Pronouns of Power and Solidarity. In American Anthropologist (), –. Also found in Language and Social Context: Selected Readings, ed. P. Giglioli (), –. Browning, Robert. . Medieval and Modern Greek. London: Hutchinson University Library. Browning, Robert. . Medieval and Modern Greek. Cambridge: Cambridge University Press.

References



Buchholz, Oda and Wilfried Fiedler. . Albanische Grammatik. Leipzig: Verlag Enzyklopädie. Buck, Carl Darling. . Comparative Grammar of Greek and Latin. Chicago, IL: University of Chicago. Buck, Carl Darling. . The Greek Dialects. Chicago, IL: University of Chicago. Buijs, Michel. . Clause Combining in Ancient Greek Narrative Discourse. Leiden–Boston: Brill. Burguière, Paul. . Histoire de l’infinitif en grec. Paris: Klincksieck. Burridge, Kate. . On the Development of Dutch Negation. In Linguistics in the Netherlands , ed. Hans Bennis and W. U. S. van Lessen Kloeke, –. Dordrecht: Foris. Burton, Ernest de Witt. . Syntax of the Moods and Tenses in New Testament Greek. Edinburgh: T. & T. Clark. Bury, Robert G.  and . Plato. Plato in Twelve Volumes, Cambridge, MA: Harvard University Press; London: William Heinemann Ltd. Busch, Peter. . Das Testament Salomos. Die älteste christliche Dämonologie, kommentiert und in deutscher Erstübersetzung. Berlin: Mouton de Gruyter. Buyssens, Edward. . Negative Contexts, English Studies , –. Bybee, Joan . Mechanisms of Change in Grammaticization: The Role of Frequency. In The Handbook of Historical Linguistics, ed. Brian D. Joseph and Richard D. Janda, –. Oxford: Blackwell. Bybee, Joan. . From Usage to Grammar: The Mind’s Response to Repetition. Language (), –. Cadbury, Henry J. a. National Ideals in the Old Testament. New York: Scribner’s. Cadbury, Henry J. b. Luke—Translator or Author? The American Journal of Theology , –. Cadoux, Cecil John. . The Imperatival Use of ἵνα in the New Testament. Journal of Theological Studies , –. Campbell, Constantine R. . Verbal Aspect, the Indicative Mood and Narrative: Soundings in the Greek of the New Testament. Studies in Biblical Greek . New York: Peter Lang. Campbell, Constantine R. a. Basics of Verbal Aspect in Biblical Greek. Grand Rapids, MI: Zondervan. Campbell, Constantine R. b. Verbal Aspect and Non-Indicative Verbs: Further Soundings in the Greek of the New Testament. New York: Peter Lang. Campbell, Lyle. . What’s Wrong with Grammaticalization? Language Sciences , –. Campbell, Lyle and Richard Janda. . Introduction: Conceptions of Grammaticalization and their Problems. Language Sciences , –. Caragounis, Chrys C. . The Development of Greek and the New Testament. Tübingen: Mohr Siebeck. Cervin, Richard S. . Word order in ancient Greek: VSO, SVO, SOV or all of the above? PhD dissertation, University of Illinois. Chafe, Wallace. . Givenness, Contrastiveness, Definiteness, Subjects, Topics, and Point of View. In Subject and Topic, ed. Charles N. Li, –. New York: Academic Press. Chafe, Wallace. . Cognitive Constraints on Information Flow. In Coherence and Grounding in Discourse, ed. Russell Tomlin, –. Amsterdam: John Benjamins.



References

Chaker, Salem. . Quelques remarques préliminaires sur la négation en berbère. In La négation en berbère et en Arabe Maghrébin, ed. Salem Chaker and Dominique Caubet, –. Paris: Harmattan. Chantraine, Paul. . Les recherches sur l’ordre des mots en grec, Anales de Filologia Clásica , . Chantraine, Pierre. –. Dictionnaire etymologique de la langue grecque. Paris: Klincksieck. Chantraine, Paul. . Morphologie historique du grec. Paris: Klincksieck. Chatelain, Emile. . Du pluriel de respect en latin. Revue de Philologie IV (April), –. Chatzopoulou, Katerina. . Negation beyond fate. Presented at the th International Conference on Greek Linguistics, Komotini, September –. Chatzopoulou, Katerina. a. Negation and Nonveridicality in the history of Greek. PhD dissertation, University of Chicago. Chatzopoulou, Katerina. b. The second negation of Greek: mi as a polarity item in diachrony. [in Greek: Η δεύτερη άρνηση της Ελληνικής: το μη ως πολικό στοιχείο στη διαχρονία]. In Proceedings of the th International Conference on Greek Linguistics—ICGL , –. September , Komotini, Greece. Chatzopoulou, Katerina. c. The Nonveridicality projection as an indicator of the unreal. Presented at Languagetalks : Graduate Conference at Ludwig-Maximilians-Universität, Alles Mögliche: Sprechen, Denken und Schreiben des (Un)Möglichen, Munich, November –. Chatzopoulou, Katerina. a. The History of the Greek NEG: Two Parameter Resets Linked to a Syntactic Status Shift. Journal of Historical Syntax (), –. Chatzopoulou, Katerina. b. Re(de)fining Jespersen’s Cycle. In Proceedings of the th Penn Linguistics Colloquium—PLC , University of Pennsylvania Working Papers in Linguistics— PWPL (), –. Chatzopoulou, Katerina. . The Greek Jespersen’s Cycle: Renewal, Stability and Structural Microelevation. In: Language Change at the Syntax-Semantics Interface, ed. Chiara Gianollo, Agnes Jäger, Doris Penka, –. Berlin: Mouton De Gruyter. Chatzopoulou, Katerina. . Language evolution and technological wisdom: from Arthur () to Protagoras. [In Greek: Γλωσσική εξέλιξη και ἔντεχνος σοφία: από τον Arthur () στον Πρωταγόρα]. In Proceedings of the st Panhellenic Conference of Epistemology, Association of Ancient Greek Philosophy ‘syn Athena’, Philosophia Ancilla Concilium II, ed. Konstantina Katsarou and Maria Papanikou, –, Kavala: Saita. Chatzopoulou, Katerina. . Language and eschatology: the stability of nonveridical marking and the consolation of linguistics [In Greek: Γλώσσα και εσχατολογία: η σταθερότητα της μη αληθειακής σήμανσης και η παραμυθία της Γλωσσολογίας]. In Proceedings of the th AMDL, ed. Kiki Tsalakanidou, –. Aristotle University of Thessaloniki. Chatzopoulou, Katerina and Anastasia Giannakidou. . Negator selection in Attic Greek is a polarity phenomenon. Poster presentation at DiGS XIII, UPenn, Philadelphia, June –. Chatzopoulou, Katerina and Ioanna Sitaridou. . Negator selection in Romeyka conditionals: Jespersen’s cycle for NEG and conditional inversion. Presented at th International Diachronic Generative Syntax Conference—DiGS XVI, Research Institute for Linguistics, Hungarian Academy of Sciences, Budapest, July –, .

References



Chomsky, Noam. . Discussion. In Third Texas Conference on Problems of Linguistic Analysis, ed. Archibald Hill, –. Austin, TX: University of Texas. Chomsky, Noam. . Lectures on Government and Binding. Dordrecht: Foris. Chomsky, Noam. . The Minimalist Program. Cambridge, MA: MIT Press. Chomsky, Noam. . Minimalist Inquiries: The Framework. In Step by Step: Essays on Minimalist Syntax in Honor of Howard Lasnik, ed. R. Martin, D.Michaels, and J. Uriagereka, –. Cambridge, MA: MIT Press. Chomsky, Noam. a. Derivation by Phase. In Ken Hale: A Life in Language, ed. M. Kenstowitz, –. Cambridge, MA: MIT Press. Chomsky, Noam. b. Beyond explanatory adequacy. Cambridge, MA: MIT. Chomsky, Noam and Morris Halle. . The Sound Pattern of English. New York: Harper & Row. Christidis, Anastasios-Phoebos, (ed.) . A History of Ancient Greek: From the Beginnings to Late Antiquity. Cambridge: Cambridge University Press. Cinque, Guglielmo. . Adverbs and Functional Heads: A Cross-Linguistic Perspective. New York: Oxford University Press. Cinque, Guglielmo. . ‘Restructuring’ and Functional Structure. Structures and Beyond , –. Clackson, James. . Indo-European Linguistics. Cambridge: Cambridge University Press. Clark, Brady. . Syntactic Theory and the Evolution of Syntax. Presentation at the Language Variation and Change Workshop, University of Chicago, May . Clarke, Charles C. . The Actual Force of the French ne. Modern Philology , –. Colvin, Stephen. . A Historical Greek Reader: Mycenaean to the Koine. Oxford: Oxford University Press. Conrad, Robert and Kepas Wogiga. . An Outline of Bukiyip Grammar. Canberra: Australian National University. Cook-Wilson, John. –. On some apparent anomalies in the use of μή. Transactions of the Oxford Philological Society, –. Cooper, Guy L. . (after K. W. Krüger) Attic Greek Prose Syntax. Revised and Expanded in English. Ann Arbor, MI: University of Michigan Press. Cowgill, Warren. . Greek ou and Armenian oč. Language (), –. Cremer, Hermann. . Biblico-theological Lexicon of New Testament Greek. Trans. William Urwick. th edition. Edinburgh: T. & T. Clark. Crisma, Paola and Giuseppe Longobardi. . Change, Relatedness and Inertia in Historical Syntax. In Historical Syntax and Linguistic Theory, ed. Paola Crisma and Giuseppe Longobardi. Oxford–New York: Oxford University Press. Croft, William. . The evolution of negation. Journal of Linguistics , –. Crosby, Alpheus. . A Grammar of the Greek Language. Boston, MA: Crosby, Nichols, Lee and Co. Crowley, Terry. . The Paamese Language of Vanuatu: Pacific Linguistics, Series B, No. . Canberra: ANU Dept. of Linguistics–Research Unit of Pacific Studies. Crowley, Terry (ed. John Lynch). a. The Avava Language of Central Malakula (Vanuatu). Canberra: Australian National University.



References

Crowley, Terry. b. Nese: A Diminishing Speech Variety of Northwest Malakula (Vanuatu). Canberra: Australian National University. Culicover, Peter. . Polarity, Inversion, and Focus in English. In Proceedings of ESCOL , –. Dagitsis, Konstantinos. . Ετυμολογικό Λεξικό της Νεοελληνικής (Εtimologiko leksiko tis Neoellinikis). Athens: I. G. Basileiou. Dahl, Eystein. . Time, Tense and Aspect in Early Vedic Grammar: Exploring Inflectional Semantics in Rigveda. Leiden–Boston: Brill. Dahl, Östen. . Typology of Sentence Negation. Linguistics , –. Dahl, Östen. a. Grammaticalization and the Life Cycles of Constructions. RASK , –. Dahl Östen. b. Inflationary Effects in Language and Esewhere. In Frequency and the Emergence of Linguistic Structure, ed. Joan Bybee and Paul Hopper, –. Amsterdam: John Benjamins. Dahl, Östen. . The Growth and Maintenance of Linguistic Complexity. Amsterdam: John Benjamins. Danckaert, Lieven and Liliane Haegeman. . Conditional Clauses, Main Clause Phenomena and the Syntax of Polarity Emphasis. In Advances in Comparative Germanic Syntax, ed. Caroline Heycock, Guido Vanden Wyngaerd, and Robert Truswell, –. Amsterdam: Benjamins. Dangitsis, Konstantinos. . Ετυμολογικό Λεξικό της Νεοελληνικής, vol. , Λ-Π. Athens: Vasiliou. Davies, Anna Morpurgo. . Negation and Disjunction in Anatolian—and Elsewhere. Anatolian Studies , –. De Clercq, Karen. . A Unified Syntax of Negation. PhD dissertation, Ghent University. De Clercq, Karen. . The nanosyntax of French negation: A diachronic perspective. In Studies on Negation: Syntax, Semantics and Variation, ed. Silvio Cruschina, Katharina Hartmann, and Eva-Maria Remberger et al., –. Vienna: V&R unipress GmbH. De Cuypere, Ludovic. . Limiting the Iconic: From the Metatheoretical Foundations to the Creative Possibilities of Iconicity in Language. Amsterdam–Philadelphia, PA: John Benjamins. De Cuypere, Ludovic, Johan van der Auwera, and Klaas Willems. . Double Negation and Iconicity. In Insistent Images, ed. Elżbieta Tabakowska, Christina Ljungberg, and Olga Fischer, –. Amsterdam: Benjamins. De Swart, Henriëtte. . Marking and Interpretation of Negation. In Crosslinguistic Research in Syntax and Semantics: Negation, Tense and Clausal Architecture, ed. Raffaella Zanuttini, Héctor Campos, Elena Herburger, and Paul H. Portner, –. Washington, DC: Georgetown University Press. De Swart, Henrietta. . Expression and Interpretation of Negation. Studies in Natural Language and Linguistic Theory, volume . Dordrecht: Springer. De Swart, Henrietta and Ivan Sag. . Negation and Negative Concord in Romance. Linguistics and Philosophy , –. Deissmann, G. Adolph. . Neue Bibelstudien. Marburg: Elwert.

References



Deissmann, G. Adolph. . The Philology of the Greek Bible: Its Present and Future, trans. L. R. M. Strachan. London: Hodder and Stoughton. Deissmann, G. Adolph. . Licht vom Osten. Tübingen: Mohr Siebeck. Delboeuf, Joseph. . Théorie de la négation dans la langue grecque. Revue de l’instruction publique en Belgique , –. Delbrück, Berthold. . Der Gebrauch des Konjuktivs und Optativs in Sanskrit und Griechisch. Syntaktische Forschungen . Halle: Waisenhaus. Delbrück, Berthold. . Vergleichende Syntax der indogermanischen Sprachen. Strasbourg: Karl Trübner. Deligianni, Efrosini. a. Modern Greek word order in the process of syntacticization: preliminary evidence from Late Byzantine and Early Modern Greek. In Proceedings of the th International Conference on Greek Linguistics (ICGL ), ed. Katerina Chatzopoulou, Alexandra Ioannidou, and Suwon Yoon, –. Deligianni, Efrosini. b. The grammaticalization of negation in Greek. Ms. Den Besten, Hans. . Double negation and the genesis of Afrikaans. In Substrata Universals in Creole Genesis, ed. Pieter Muysken and Norval Smith, –. Amsterdam: John Benjamins. Den Besten, Hans. . Studies in West Germanic Syntax. PhD dissertation, University of Tilburg. Den Dikken, Marcel. . Parasitism, Secondary Triggering and Depth of Embedding. In Crosslinguistic Research in Syntax and Semantics: Negation, Tense and Clausal Architecture, ed. Raffaella Zanuttini, Héctor Campos, Elena Herburger, and Paul H. Portner, –. Washington, DC: Georgetown University Press. Denniston, John D. . Greek Prose Style. Oxford: Clarendon Press. Denniston, John D. . The Greek Particles. Oxford: Clarendon Press. Detges, Ulrich and Richard Waltereit. . Grammaticalization vs. Reanalysis: A SemanticPragmatic Account of Functional Change in Grammar. Zeitschrift für Sprachwissenschaft , –. Devine, Andrew and Lawrence Stephens. . The Prosody of Greek Speech. Oxford–New York: Oxford University Press. Devine, Andrew and Lawrence Stephens. . Discontinuous syntax: Hyperbaton in Greek. Oxford: Oxford University Press. Devos, Maud, Michael Kasombo Tshibanda, and Johan van der Auwera. . Jespersen Cycles in Kanincin: Double, Triple and Maybe Even Quadruple Negation, Africana Linguistica , –. Devos, Maud and Johan Van der Auwera. . Jespersen cycles in Bantu: double and triple negation. Journal of African Languages and Linguistics  (), –. Dik, Helma. . Word Order in Ancient Greek: A Pragmatic Account of Word Order Variation in Herodotus. Amsterdam: Gieben. Dik, Helma. . Word Order in Greek Tragic Dialogue. Oxford: Oxford University Press. Dindorf, L. ed. . Chronicon Pascale, vols  and  of Reinhold Niebuhr, Corpus scriptorum hist. byzantinae, de Cange’s preface and commentary. Bonn: Weber. Djamouri, Redouane. . Particule de negation dans les inscriptions sur bronze de la dynastie des Zhou. Cahiers de Linguistique—Asie Orientale (), –.



References

Djamouri, Redouane. . Comptes rendus-Pulleyblank. Cahiers de Linguistique—Asie Orientale (), –. Doetjes, Jenny. . Quantifers and selection. PhD dissertation, Rijksuniversiteit. Donhauser, Karin. . Negationssyntax in der deutschen Sprachgeschichte. In Deutsch -typologisch, ed. Ewald Lang and Gisela Zifonun, –. Berlin: Mouton de Gruyter. Dover, Kenneth. . The Evolution of Greek Prose Style. Oxford: Clarendon Press. Dover, Kenneth.  []. Greek Word Order. London: Cambridge University Press. Dowty, David. . Word Meaning and Montague Grammar. Dordrecht: Reidel. Dowty, David. . The Role of Negative Polarity and Concord Marking in Natural Language. From Semantics and Linguistic Theory (SALT) IV, –. Ithaca, NY: Cornell University. Dryer, Matthew and Martin Haspelmath (eds). . The World Atlas of Language Structures Online. Munich: Max Planck Digital Library. [Online.] Available at http://wals.info/ Accessed May , . Dueck, Daniela. . Strabo of Amasia: Greek Man of Letters in Augustan Rome. New York: Routledge. Duff, Jeremy. . Elements of New Testament Greek. Cambridge: Cambridge University Press. Dum-Tragut, Jasmine. . Armenian. Amsterdam–Philadelphia, PA: John Benjamins. Dummett, Michael. . Frege: Philosophy of Language. New York: Harper and Row. Early, Robert. a. Lewo. In Typological Studies in Negation, ed. Peter Kahrel and René van den Berg, –. Amsterdam: Benjamins. Early, Robert. b. A grammar of Lewo, Vanuatu. PhD dissertation, Australian National University. Ebeling, Heinrich L. . Some Statistics on the Order of Words in Greek. In Studies in Honor of Basil L. Gildersleeve, ed. B. L. Gildersleeve, –. Baltimore, MD: Johns Hopkins University Press. Eckardt, Regine. . Eine Runde im Jespersen-Zyklus. Negation, emphatische Negation und negativ-polare Elemente im Altfranzösischen. Konstanzer Online-Publikations-System. [Online.] Available at http://kops.ub.uni-konstanz.de/bitstream/handle/urn:nbn:de: bsz:-opus-/Eckardt.pdf?sequence=> Accessed April , . Eckardt, Regine. . Meaning Change in Grammaticalization. Oxford: Oxford University Press. Efthimiou, Angeliki. . Negative Morphemes in Modern Greek: The Case of a- and mi. In La raison morphologique: Hommage à la mémoire de Danielle Corbin, ed. Bernard Fradin, –. Amsterdam–Philadelphia, PA: John Benjamins. Eichner, Heiner. . Urindogermanisch *kwe ‘wenn’ im Hethitischen. Münchener Studien zur Sprachwissenschaft , –. Elšik, Viktor and Yaron Matras. . Markedness & Language Change. Berlin: Mouton de Gruyter. Errington, R. Malcolm. . A History of the Hellenistic World. Oxford: Blackwell. Falaus, Anamaria. . Polarity Items and Dependent Indefinites in Romanian. PhD dissertation, Université de Nantes. Faller, Martina. . Semantics and Pragmatics of Evidentials in Cuzco Quechua. PhD dissertation, Stanford.

References



Fanning, Buist M. . Verbal Aspect in New Testament Greek. Oxford: Clarendon Press. Fanning, Buist M. . Approaches to Verbal Aspect in New Testament Greek: Issues in Definition and Method. In Biblical Greek Language and Linguistics, ed. Stanley E. Porter and D. A. Carson, –. Sheffield: JSOT Press. Farkas, Donka F. . Intensional Descriptions and the Romance Subjunctive Mood. New York: Garland. Farkas, Donka F. a. On the Semantics of Subjunctive Complements. In Romance Languages and Modern Linguistic Theory, ed. Hirschbühler, P., et al., –. Amsterdam and Philadelphia, PA: John Benjamins. Farkas, Donka F. b. Mood Choice in Complement Clauses. In Approaches to Hungarian, vol. : The Structure of Hungarian, ed. I. Kenesei, and E. Pleh, –. Szeged: JATE. Farkas, Donka F. . Extreme non-specificity in Romanian. In Romance Languages and Linguistic Theory , ed. Claire Beyssade, Reineke Bok-Bennema, Frank Drijkoningen, and Paola Monachesi, –. Amsterdam: John Benjamins. Fauconnier, Gilles. . Polarity and the Scale Principle. Chicago Linguistics Society . –. Fauconnier, Gilles. . Etude de certains aspects logiques et grammaticaux de la quantification et de l’anaphore en Français et en Anglais. PhD dissertation, University of Paris VII. Fauriel, Claude Charles. –. Chants populaires de la Grèce modern, recueillis et publiés avec une traduction française, des éclaircissements et des notes. V. –. Paris: Didot. Fernald, Theodore B. and Ellavina Perkins. . Negative polarity items in Navajo. In Proceedings of the Dene (Athabaskan) Languages Conference (= ALC) (Alaska Native Language Center Working Papers ), ed. A. Berez, S. Gessner and S. Tuttle, –. Fairbanks, AK: ANLC Publications. Fintel, von, Kai. . NPI-licensing, Strawson-entailment, and Context Dependency. Journal of Semantics , –. Fischer, Olga, Ans van Kemenade, Willem Koopman, and Wim van der Wurff. . The Syntax of Early English. Cambridge: Cambridge University Press. Fischer, Susann. . Grammaticalization within the IP Domain. In The Paradox of Grammatical Change, ed. Ulrich Detges and Richard Waltereit, –. Amsterdam– Philadelphia, PA: John Benjamins. Fletcher, Judith. . Performing Oaths in Classical Greek Drama. Cambridge: Cambridge University Press. Fortson, Benjamin. W.  []. Indo-European Language and Culture: An Introduction. Malden, MA: Blackwell Textbooks in Linguistics. Fowler, Frank Hamilton. . The Negatives of the Indo-European Languages. Chicago, IL: The University of Chicago Press. Frajzyngier, Zygmunt. . Grammaticalization, Typology, and Semantics: Expanding the Agenda. In Rethinking Grammaticalization. New Perspectives. ed. Maria Jose Lopez-Couso and Elena Seoane, –. Amsterdam: Benjamins. Frank, Jane. . You Call That a Rhetorical Question?: Forms and Functions of Rhetorical Questions in Conversation. In Journal of Pragmatics , –. Franke, Friedrich. . De particulis negantibus linguae Graecae commentario. Part . In Annalium scholasticorum particula undetricesima qua publicata ad examina in Gymnasio



References

Electorali Hasso-Schaumburgo DD. IX, X, XII, XIII Aprilis instituenda, ed. Augustus Schiek, –. Rinteln: Gymnasium Electorali Hasso-Schaumburgo. Franke, Michael, Gergard Jäger, and Robert van Rooij. . Vagueness, Signaling & Bounded Rationality. In JSAI-isAI , ed. T. Onoda, D. Bekki, and E. McCready, –. Berlin: Springer. Frantz, Donald G. . Blackfoot Grammar. Toronto: University of Toronto Press. Frazier, Lyn, Jason Merchant, Thomas Weskott, and Charles Clifton. . Fragment answers to questions: A case of inaudible syntax. Ms. Frege, Gottlob.  [], translated by J. L. Austin. The Foundations of Arithmetic: A Logico-mathematical Enquiry into the Concept of Number. New York: Harper. Friberg, Timothy. . New Testament Greek Word Order in Light of Discourse Considerations. PhD dissertation, University of Minnesota. Friedman, Victor. . A Problem in Grammatical Invariance. In The CLS Book of Squibs, ed. Samuel E. Fox, Woodford E. Beach, and Shulamith Philosoph, –. Chicago: Chicago Linguistic Society. Friedman, Victor. . Verbal Categories in the Languages of the Balkans (Report). In The American Philosophical Society Yearbook , . Philadelphia, PA: American Philosophical Society. Friedman, Victor. . The Balkans as a Linguistic Area. In Elsevier Encyclopedia of Language and Linguistics, vol. , editor-in-chief Keith Brown, –. Oxford: Elsevier. Friedrich, Johannes. . Zum hethitischen Lexicon. Journal of Cuneiform Studies , –. Frisch, Stefan A. . The Change in Negation in Middle English: A NegP Licensing Account. Lingua , –. Frösén, Jaakko. . Prolegomena to a study of the Greek language in the first centuries ..: the problem of Koiné and Atticism. PhD dissertation, University of Helsinki. Furtado da Cunha, Maria Angélica. . O modelo das motivações competidoras no domínio funcional da negação. D.E.L.T.A. , –. Gaatone, David. . Etude descriptive du système de la négation en français contemporain. Geneva: Libraire Droz. Gallaway, William Francis. . On the Use of Mē with the Participle in Classical Greek. PhD dissertation, Johns Hopkins University. Baltimore, MD: John Murphy & Co. Gardiner, Alan H. . The word . Zeitschrift für Ägyptische Sprache und Altertumskunde , –. Garzonio, Jacopo & Cecilia Poletto. . Minimizers and quantifiers: a window on the development of negative markers. STiL – Studies in Linguistics, CISCL Working Papers, Vol. . Gass, Susan M. and Larry Selinker. . Second Language Acquisition: An Introductory Course, rd edition. New York: Routledge. Gayler, Christoph. . Particularum Graeci sermonis negativarum οὐ et μή, οὐ μή et μή οὐ accurata disputatio. Tubingen: C. F. Osiander. George, A. R. . The imperative use of hina in the New Testament. Journal of Theological Studies , –. Gerö, Eva-Carin. . Negatives and Noun Phrases in Classical Greek: An Investigation Based on the Corpus Platonicum. Frankfurt am Main–Berlin: Peter Lang.

References



Gerö, Eva-Carin. . The usage of AN and KE in Ancient Greek: Towards a Unified Description. Glotta , –. Gerö, Eva-Carin. a. Negatives and Noun Phrases in Classical Greek: A Reconsideration. Glotta , –. Gerö, Eva-Carin. b. Irrealis and Past Tense in Ancient Greek. Glotta , –. Geurts, Bart. . Explaining Grammaticalization (the Standard Way). Linguistics , –. Giannakidou, Anastasia. . The Semantic Licensing of NPIs and the Modern Greek Subjunctive. In Language and Cognition , Yearbook of the Research Group for Theoretical and Experimental Linguistics, ed. Ale de Boer, Helen de Hoop, and Henriette de Swart, –. Groningen: University of Groningen. Giannakidou, Anastasia. . Subjunctive, habituality and negative polarity. In Mandy Simons and Teresa Galloway (eds), Semantics and Linguistic Theory (SALT) V, –. Ithaca, NY: CLC Publications. Giannakidou, Anastasia. . The Landscape of Polarity Items. PhD dissertation, University of Groningen. Giannakidou, Anastasia. . Polarity Sensitivity as (Non)veridical Dependency. Amsterdam: John Benjamins. Giannakidou, Anastasia. . Affective dependencies. Linguistics and Philosophy , –. Giannakidou, Anastasia. . Negative . . . Concord? Natural Language and Linguistic Theory , –. Giannakidou, Anastasia. . The Meaning of Free Choice. Linguistics and Philosophy , –. Giannakidou, Anastasia. . Licensing and Sensitivity in Polarity Items: From Downward Entailment to Nonveridicality. In CLS : Papers from the th Annual Meeting of the Chicago Linguistic Society, ed. Maria Andronis, Anne Pycha, and Keiko Yoshimura, –. Chicago, IL: Chicago Linguistic Society. Giannakidou, Anastasia. . N-Words and Negative Concord. In The Linguistics Companion, Volume , ed. M. Everaert, H. van Riemsdijk, R. Goedemans, and B. Hollebrandse, –. Oxford: Blackwell Publishing. Giannakidou, Anastasia. . Only, Emotive Factives, and the Dual Nature of Polarity Dependency. Language, (), –. Giannakidou, Anastasia. . The Landscape of EVEN. Natural Language and Linguistic Theory , –. Giannakidou, Anastasia. . The Dependency of the Subjunctive Revisited: Temporal Semantics and Polarity. Lingua , –. Giannakidou, Anastasia. . Negative and Positive Polarity Items: Variation, Licensing, and Compositionality. In Semantics: An International Handbook of Natural Language Meaning, ed. Claudia Maienborn, Klaus von Heusinger, and Paul Portner, –. Berlin: Mouton de Gruyter. Giannakidou, Anastasia. . The Greek future as an epistemic modal. In Proceedings of the International Conference on Greek Linguistics—ICGL , –. Komotini: University of Thrace. Giannakidou, Anastasia. a. The Futurity of the Present and the Modality of the Future: A Commentary on Broekhuis and Verkuyl. Natural Language and Linguistic Theory (), –.



References

Giannakidou, Anastasia. b. The Prospective as Nonveridical: Polarity Items, Speaker Commitment and Projected Truth. In The Black Book. Feestschrift for Frans Zwarts, ed. D. Gilberts and J. Hoeksema, –. Groningen: CLCG. Giannakidou, Anastasia. . The subjunctive as evaluation and nonveridicality. In Mood, Aspect and Modality: What is a Linguistic Category?, ed. J. Blaszack et al. Chicago, IL: University of Chicago Press. Giannakidou, Anastasia and L. Cheng. . (In)definiteness, polarity and the role of whmorphology in free choice. Journal of Semantics , –. Giannakidou, Anastasia and Alda Mari. . Mixed non-verdicaility and mood choice with emotives. In Proceedings of CLS  (), –. Giannakidou, Anastasia and Alda Mari. . A Unified Analysis of the Future as Epistemic Modality: The View from Greek and Italian. Natural Language and Linguistic Theory , –. Giannakidou, Anastasia and Suwon Yoon. . Scalar Marking without Scalar Meaning: Nonscalar, Nonexhaustive Even-Marked NPIs in Greek and Korean. Language (), –. Giannakidou, Anastasia and Hedde Zeijlstra. . The Landscape of Negative Dependencies: Negative Concord and n-words. In The Wiley Blackwell Companion to Syntax, ed. Martin Everaert and Henk C. Van Riemsdijk, nd edition. New York: Blackwell. Giannakidou, Anastasia and Frans Zwarts. . Aspectual properties of temporal connectives. In Greek Linguistics: Proceedings of the rd International Conference on Greek Linguistics, ed. Amalia Mozer, –. Athens: Ellinika Grammata. Gignac, Francis T. –. A Grammar of the Greek Papyri of the Roman and Byzantine periods. Milan: Istituto Editoriale Cisalpino—La Goliardica. Gildersleeve, Basil L. . Encroachments of μή on οὐ in Later Greek. American Journal of Philology , –. Ginzburg, Jonathan. . Interrogatives: Questions, Facts, and Dialogue. In The Handbook of Contemporary Semantic Theory, ed. S. Lappin, –. Oxford: Blackwell. Giorgi, Alessandra. . Toward a Syntax of the Subjunctive Mood. Lingua, (), –. Giorgi, Alessandra and Fabio Pianesi. . Tense and Aspect: From Semantics to Morphosyntax. Oxford: Oxford University Press. Givón, Talmy. . Historical Syntax and Synchronic Morphology: An Archaeologist’s Field Trip. Chicago Linguistic Society , –. Givón, Talmy. . Serial Verbs and Syntactic Change: Niger-Congo. In Word Order and Word Order Change, ed. Charles N. Li, –. Austin, TX–London: University of Texas Press. Givón, Talmy. . Negation in Language: Pragmatics, Function, Ontology. Syntax and Semantics : Pragmatics, ed. Peter Cole, –. New York: Academic Press. Givón, Talmy. a. On Understanding Grammar. New York–London: Academic Press. Givón, Talmy. b. Discourse and Syntax. New York–London: Academic Press. Givón, Talmy. . Internal Reconstruction: As Method, as Theory. In Reconstructing Grammar, ed. Spike Gildea, –. Amsterdam: Benjamins. Godard, D. . French Negative Dependency. In Handbook of French Semantics, ed. F. Corblin and H. de Swart, –. Stanford, CA: CSLI Publications. Goetze, Albrecht. . Indefinites and Negations, Prohibitives and Imperatives in the Urartean Language. Revue hittite et asianique , –.

References



Goldbach, Maria. . Morphological Developments Affecting Syntactic Change. In The Paradox of Grammatical Change, ed. Ulrich Detges and Richard Waltereit, –. Amsterdam–Philadelphia, PA: John Benjamins. Goldsmith, John. . Autosegmental Phonology. MIT dissertation. New York: Garland Press. Goldsmith, John. . Autosegmental and Metrical Phonology. Oxford: Basil Blackwell. Goldstein, J. A. . II Maccabees. New York: Doubleday. Gonda, Jan. . The Character of the Indo-European Moods. Wiesbaden: Harrassowitz. Goodman, Nelson. . The Problem of Counterfactual Conditionals. In Semantics and the Philosophy of Language, ed. L. Linsky, –. Urbana, IL: University of Illinois. Goodwin, William Watson. . Syntax of the Moods and Tenses of the Greek Verb. New York: Macmillan. Grant, Michael. . From Alexander to Cleopatra: the Hellenistic World. New York: Scribner. Green, Peter. . The Hellenistic Age. New York: Modern Library. Greenberg, Joseph H. (ed.) . Universals of Language. Cambridge, MA: MIT Press. Greenberg, Joseph H. . Language Universals. The Hague: Mouton de Gruyter. Greenberg, Joseph H. . The Realis-Irrealis Continuum in the Classical Greek Conditional. In On Conditionals, ed. Elizabeth Closs Traugott et al., –. Cambridge: Cambridge University Press. Greenberg, Joseph H. . The Last Stages of Grammatical Elements: Contractive and Expansive Desemanticization. In Approaches to Grammaticalization,  vols, ed. Elizabeth Closs Traugott and B. Heine, vol. , –. Amsterdam: Benjamins. Grice, Paul. . Logic and Conversation. In Syntax and Semantics : Speech Acts, ed. P. Cole and J. Morgan, –. New York: Academic Press. Groenendijk, Jeroen. . The Logic of Interrogation. In Proceedings of SALT IX, ed. t. Matthews and D. Strolovitch, –. Ithaca, NY: CLC Publications. Groenendijk, Jeroen. . Inquisitive Semantics: Assertions, Questions, and Hybrids. Ms. Amsterdam. Groenendijk, Jeroen and Martin Stokhof. . Studies in the Semantics of Questions and the Pragmatics of Answers. PhD thesis. University of Amsterdam. Gundel, Jeanette. . Universals of Topic-Comment Structure. In Studies in Syntactic Typology, ed. Michael Hammond, Edith Moravcsik, and Jessica Wirth, –. Amsterdam: John Benjamins. Gunlogson, Christine. . True to Form: Rising and Falling Declaratives as Questions in English. PhD dissertation, UCLA. Haberland, Hartmut. . Mood in Greek. In Moods in the languages of Europe, ed. Björn Rothstein and Rolf Thieroff, –. Amsterdam: Benjamins. Haegeman, Liliane. . The Syntax of Negation. Cambridge: Cambridge University Press. Haegeman, Liliane and Terjje Lohndal. . Negative Concord and Multiple Agree: A Case Study of West-Flemish. Linguistic Inquiry , –. Haegeman, Liliane and Raffaella Zanuttini. . Negative Heads and the Neg-criterion. The Linguistic Review , –. Haegeman, Liliane and Raffaella Zanuttini. . Negative Concord in West Flemish. In Parameters and Functional Heads, ed. Adriana Belleti and Luigi Rizzi, –. Oxford: Oxford University Press.



References

Hahn, E. Adelaide. . Some Hittite Words in ta-. Language (), –. Hahn, E. Adelaide. . Subjunctive and Optative: Their Origin as Futures. New York: American Philological Association. Hale, William. G. . A Century of Metaphysical Syntax. St. Louis Congress of Arts and Science, Universal Exposition , –. Boston, MA: Houghton Mifflin. Hamari, Arja. . Continuity and Change in the Negation System of Mordvin Languages. Ms. Hammond, Lila. . Serbian: An Essential Grammar. London–New York: Routledge. Han, Chung-Hye. . The syntax and semantics of Imperatives and Related Constructions. PhD dissertation, University of Pennsylvania. Han, Chung-Hye. . The Structure and Interpretation of Imperatives: Mood and Force in Universal Grammar (Outstanding Dissertations in Linguistics). New York: Garland Publishing. Han, Chung-Hye. . Force, Negation and Imperatives. The Linguistic Review , –. Han, Chung-Hye. . Interpreting Interrogatives as Rhetorical Questions. Lingua , –. Hansen, M.-B. M. . The Grammaticalization of Negative Reinforcers in Old and Middle French: A Discourse-Functional Approach. In Current Trends in Diachronic Semantics and Pragmatics, ed. Maj-Britt Mosegaard Hansen and Jacqueline Visconti, –. Bingley: Emerald. Harris, Alice C. and Lyle Campbell. . Historical Syntax in Cross-linguistic Perspective. Cambridge: Cambridge University Press. Harris, Martin. . A Typological Approach to Word Order Change in French. In Romance Syntax: Synchronic and Diachronic Perspectives, ed. Martin Harris, –. Salford: University of Salford. Harris, Martin. . The Evolution of French Syntax. London: Longman. Hartung, Johann Adam. . Lehre von den Partikeln der Griechischen Sprache, Two volumes. Erlangen: Palm & Enke. Haspelmath, Martin. a. Indefinite Pronouns. Oxford: Oxford University Press. Haspelmath, Martin. b. From Space to Time: Temporal Adverbials in the World’s Languages. (Lincom Studies in Theoretical Linguistics, ). Münich–Newcastle: Lincom Europa. Haspelmath, Martin. . On Directionality in Language Change with Particular Reference to Grammaticalization. In Up and Down the Cline: The Nature of Grammaticalization, ed. Olga Fischer, Muriel Norde, and Harry Perridon, –. Amsterdam: John Benjamins. Hatzidakis, George N. . Einleitung in die neugriechische Grammatik. Leipzig: Breitkopf & Härtel. Hatzidakis, George N. . Σύντομος ἱστορία τῆς Ἑλληνικῆς γλώσσης. Athens: ΣΥΛΛΟΓΟΣ ΠΡΟΣ ΔΙΑΔΟΣΙΝ ΩΦΕΛΙΜΩΝ ΒΙΒΛΙΩΝ (Silogos pros diadosin ofelimon vivlion). Hatzidakis, George N. . Γλωσσικαί παρατηρήσεις [glōssikaí paratīrī́seis: language observations]. Athens: Sakellariou. Hatzidakis, George N. . Ετυμολογικά και Συντακτικά. Λεξικογραφικόν αρχείον της Μέσης και της Νέας Ελληνικής , –. Hatzidakis, George N.  []. Glōssologikai ereunai,  vols. Athens: Academy of Athens.

References



Haug, Dag. . Word order change and stability in ancient Greek. Ms, University of Oslo [Online.] Available at http://www.hf.uio.no/ifikk/english/research/projects/proiel/Activ ities/proiel/publications/haug-nijmegen.pdf. Accessed May , . Hawkins, J. D. . The Negatives in Hieroglyphic Luwian. Anatolian Studies , –. Hay, Jen, Christopher Kennedy, and Beth Levin. . Scale Structure Underlies Telicity in `Degree Achievements’. In Proceedings of SALT IX, ed. Tanya Matthews and Devon Strolovitch, –. Ithaca, NY: CLC Publications. Heim, Irene. . A Note on Negative Polarity and Downward Entailingness. In Proceedings of NELS , ed. Charles Jones and Peter Sells, –. Amherst, MA: University of Massachusetts. Heim, Irene. . Presupposition Projection and the Semantics of Attitude Verbs. Journal of Semantics , –. Heine, Bernd and Tania Kuteva. . World Lexicon of Grammaticalization. Cambridge: Cambridge University Press. Heine, Bernd and Tania Kuteva. . The Genesis of Grammar: A Reconstruction. Oxford: Oxford University Press. Heine, Bernd, U. Claudi, and F. Huennemeyer. . Grammaticalization: A Conceptual Framework. Chicago, IL: University of Chicago Press. Helbing, Robert. . Grammatik der Septuaginta. Laut- und Wortlehre, Göttingen: Vandenhoeck and Ruprecht. Hengeveld, Kees. . The hierarchical structure of utterances. In Layers and levels of representation in language theory, ed. J. Nuyts, A. M. Bolkestein, and C. Vet, –. Amsterdam: Benjamins. Herburger, Elena. . The Negative Concord Puzzle Revisited. Natural Language Semantics , –. Hewson, John and Vit Bubenik. . Tense and Aspect in Indo-European Languages. Amsterdam/ Philadelphia, PA: John Benjamins. Hinrichs, E. . A Compositional Semantics for Aktionsarten and NP Reference. PhD dissertation, Ohio State University. Hock, Hans Henrich. . Principles of Historical Linguistics. Berlin: Mouton de Gruyter. Hock, Hans Henrich and Brian D. Joseph. . Language History, Language Change and Language Relationship, nd edition. Berlin–New York: Mouton de Gruyter. Hoeksema, J. . Monotonie en superlatieven. In Proeven van taalwetenschap, ed. C. Hoppenbrouwers, Joop Houtman, Inneke Schuurman, and Frans Zwarts. Groningen: TABU, Nederlands Instituut. Hoeksema, J. . Negation and Negative Concord in Middle Dutch. In Negation and Polarity: Syntax and Semantics, ed. D. Forget, P. Hirschbühler, F. Martineau, and M.-L. Rivero, –. Amsterdam: John Benjamins. Hoeksema, Jack. . Negation and polarity. In Negation and Polarity: Syntactic and Semantic Perspectives, ed. L. Horn and Y. Kato, –. Oxford–New York: Oxford University Press. Hoeksema, Jack. . Negative and Positive Polarity Items: An investigation of the interplay of lexical meaning and global conditions on expression. In The Expression of Negation, ed. L. R. Horn, –. Berlin: Mouton de Gruyter.



References

Hoffmann, Karl. . Der Injunktiv im Veda: Eine synchronische Functionsuntersuchung. Heidelberg: Winter. Hoffner, Harry A. Jr and H. Craig Melchert. . A Grammar of the Hittite Language. Winona Lake, IN: Eisenbrauns. Holm, John A. . Pidgins and Creoles: Theory and Structure, vol. . Cambridge: Cambridge University Press. Holton, David, Peter Mackridge, and Irene Philippaki-Warburton. . Greek: A Comprehensive Grammar of the Modern Language. London: Routledge. Honda, Isao. . Negation: A cross-linguistic study. PhD dissertation, Buffalo: State University of New York at Buffalo. Hoogeveen, Hendrik. . Doctrina particularum linguae graecae. Two volumes. Leiden: Etypograph Dammeano. Hoogeveen, Hendrik.  []. Doctrina Particularum Graecarum, revised by Christian G. Schütz. Leipzig. Hopper, Paul J. . On Some Principles of Grammaticization. In Approaches to Grammaticalization, vol. I, ed. Elizabeth Closs Traugott and Bernd Heine, –. Amsterdam: John Benjamins. Hopper, Paul J. and Elizabeth Closs Traugott.  []. Grammaticalization. Cambridge: Cambridge University Press. Hörmann, Hans. . Semantic Factors in Negation. Psychological Research , –. Horn, Larry R. . A presuppositional approach to Only and Even. In Proceedings of the Chicago Linguistic Society , –. Horn, Larry R. . On the Semantic Properties of Logical Operators in English, PhD thesis, UCLA. Horn, Larry R. . Metalinguistic negation and pragmatic ambiguity. Language , –. Horn, Larry R. . A Natural History of Negation. Stanford, CA: CSLI Publications. (nd edition , Chicago, IL: University of Chicago Press.) Horn, Larry R. . Multiple Negation in English and Other Languages. In The Expression of Negation, ed. L. R. Horn, –. Berlin–New York: Mouton de Gruyter. Horn, Robert Chisolm. . The use of the subjunctive and optative moods in the non-literary papyri. PhD dissertation, University of Pennsylvania. Horrocks, Geoffrey. . Clitics in Greek: A Diachronic Review. In Greek Outside Greece II, ed. M. Roussou and S. Panteli, –. Athens: Diaspora Books. Horrocks, Geoffrey.  []. Greek: A History of the Language and its Speakers. . London: Longman. . Oxford: Wiley-Blackwell. Horrocks, Geoffrey. . Ouk Ísmen Oudén: Negative Concord and Negative Polarity in the History of Greek. Journal of Greek Linguistics , –. Horsley, G. H. R. . New Documents Illustrating Early Christianity, vol. . Sydney: Macquarie University. Horsley, Richard H. . The Fiction of Jewish Greek. In New Documents Illustrating Early Christianity, ed. G. H. R. Horsley, vol. , –. Sydney: Macquarie University, The Ancient History Documentary Research Centre.

References



Hovdhaugen, Even and Mosel, Ulrich. . Negation in Oceanic Languages. Typological Studies. Lincom Studies in Austronesian Languages, vol. . Munich and Newcastle: Lincom Europa. Housholder, Fred W. . The Syntax of Apollonius Dyscolus. Studies in the History of Linguistics . Amsterdam: John Benjamins. Humbert, Jean. . Syntaxe Grecque. Paris: Librairie Klincksieck. Hummel, Pascale. . De lingua Graeca. Bern: Peter Lang. Hunger, Herbert. . Die hochsprachliche profane Literatur der Byzantiner,  vols,. Munich: Beck. Ifantidou, Elly. . Evidentials and Relevance. Amsterdam: John Benjamins. Ioannidou, Alexandra and Marcel den Dikken. . P-drop, D-drop, D-spread. Talk given at Syracuse University. Israel, Michael. . The Grammar of Polarity. Cambridge: Cambridge University Press. Izvorski, Roumyana. . The Present Perfect as an Epistemic Modal. In Proceedings from Semantics and Linguistic Theory VII, ed. A. Lawson, –. Ithaca, NY: Cornell University. Jackendoff, Ray. . An Interpretive Theory of Negation. Foundations of Language , –. Jackson, Williams, A. V. . An Avesta Grammar in comparison with Sanskrit. Stuttgart: Kohlhammer. Jäger, Agnes. . History of German Negation. Amsterdam–Philadelphia, PA: John Benjamins. Jakobson, Roman. . The Structure of the Russian Verb. Reprint in: L. R. Waugh and M. Halle (eds.). , Russian and Slavic Grammar Studies, –, –. Berlin: Mouton de Gruyter. Jakobson, Roman. . Beitrag zur allgemeinen Kasuslehre: Gesamtbedeutungen der russischen Kasus. Travaux du Cercle Linguistique de Prague , –. Prague : Cercle Linguistique de Prague. Janda, Richard D. . Beyond ‘Pathways’ and ‘Unidirectionality’: On the Discontinuity of Language Transmission and the Counterability of Grammaticalization. Language Sciences (Special Issue—Grammaticalization: A Critical Assessment, ed. Lyle Campbell) , –. Jannaris, Antonius N. . An Historical Greek Grammar. London: MacMillan. Jespersen, Otto. . Negation in English and Other Languages. Copenhagen: A. F. Høst. Jespersen, Otto. . The Philosophy of Grammar. London: Allen & Unwin. Joseph, Brian D. . A new convergence involving the Balkan loss of the infinitive. Indogermanischen Forschungen , –. Joseph, Brian D. . On the Synchrony and Diachrony of Modern Greek na. In Byzantine and Modern Greek Studies , –. Joseph, Brian D. . The Synchrony and Diachrony of the Balkan Infinitive: A Study in Areal, General and Historical Linguistics. Cambridge: Cambridge University Press. Joseph, Brian D. . Complementizers, Particles and Finiteness in Greek and the Balkans. Folia Slavica (), –. Joseph, Brian D.  []a. Morphology and universals in syntactic change: evidence from Medieval and Modern Greek. PhD dissertation, Harvard University.



References

Joseph, Brian D.  []b. Greek. In The World’s Major Languages, ed. Bernard Comrie, –. New York–Oxford: Oxford University Press. Joseph, Brian D. . A Diachronic Phonological Solution to the Syntax of Vedic Negative Particles. In Studies in Sanskrit Syntax: A Volume in Honor of the Centennial of Speijer’s Sanskrit Syntax, ed. Hans Henrich Hock, –. Delhi: Motilal Banarsidass Publishers. Joseph, Brian D. a. Textual Authenticity: Evidence from Medieval Greek. In Textual Parameters in Ancient Languages, ed. S. Herring, P. van Reenen, and L. Schoesler, –. Amsterdam–Philadelphia, PA: John Benjamins. Joseph, Brian D. b. Is there such a thing as ‘grammaticalization’? Language Sciences (), –. Joseph, Brian D. a. Language Contact and the Development of Negation in Greek—and How Balkan Slavic Helps to Illuminate the Situation. A Festschrift for Leon Twarog. Working Papers in Slavic Studies, vol. , ed. I. Masing-Delic & M. Matejic, –. Columbus, OH: The Ohio State University: Department of Slavic and East European Languages and Literatures. Joseph, Brian D. b. Language Contact and the Development of Negation in Greek and the Balkans. In Greek Linguistics ‘. Proceedings of the th International Conference on Greek Linguistics, Nicosia, September , –. Thessaloniki: University Studio Press. Joseph, Brian D. . Balkan insights into the Syntax of *me: in Indo-European. In IndoEuropean Perspectives (Journal of Indo-European Studies Monograph Series ), ed. M. Southern, –. Washington, DC: Institute for the Study of Man. Joseph, Brian D. . Morphologization from Syntax. In The Handbook of Historical Linguistics, ed. Brian D. Joseph and Richard D. Janda, –. Malden, MA: Blackwell. Joseph, Brian D. . Rescuing Traditional (Historical) Linguistics from Grammaticalization “Theory.” In Up and Down the Cline—The Nature of Grammaticalization, ed. O. Fischer, M. Norde, and H. Perridon, –. Amsterdam: John Benjamins. Joseph, Brian D. . Some Ancient Shared Metaphors in the Balkans. Acta Studia Albanica (), –. Joseph, Brian D. . How Accommodating of Change is Grammaticalization? The Case of “Lateral Shifts.” Logos and Language. Journal of General Linguistics and Language Theory (), –. Joseph, Brian D. . Review Article. Theodore Markopoulos. The Future in Modern Greek: From Ancient to Medieval. In Journal of Greek Linguistics , –. Joseph, Brian D. . Grammaticalization: a general critique. In Handbook of Grammaticalization, ed. H. Narrog and B. Heine, –. Oxford: Oxford University Press. Joseph, Brian D. and Richard D. Janda. . The Modern Greek Negator mi(n)(-) as a Morphological Constellation. In Greek Linguistics: Proceedings of the rd International Conference on Greek Linguistics, ed. G. Babiniotis, –. Athens: Ellinika Gramata. Joseph, Brian D. and Richard D. Janda. a. The Handbook of Historical Linguistics. Malden, MA: Blackwell. Joseph, Brian D. and Richard D. Janda. b. On Language, Change and Language Change— or, Of History, Linguistics and Historical Linguistics. In Handbook of Historical Linguistics, ed. B. Joseph and R. Janda, –. Oxford: Blackwell. Joseph, Brian D. and Panayiotis Pappas. . On Some Recent Views Concerning the Development of the Greek Future System. Byzantine and Modern Greek Studies. , –.

References



Joseph, Brian D. and Irene Philippaki-Warburton. . Modern Greek. London: Croom Helm. Joseph, Brian D. and Carol Smirniotopoulos. . The Morphosyntax of the Modern Greek Verb as Morphology and not as Syntax. Linguistic Inquiry , –. Josephson, Folke. . Pronominal Adverbs of Anatolian: Formation and Function. Revue Hittite et Asianique , –. Kahrel, Peter. . Aspects of negation. PhD dissertation, University of Amsterdam. Kahrel, Peter and René van den Berg (eds). . Typological Studies in Negation. Amsterdam: John Benjamins. Kallel, Amel. . The Loss of Negative Concord in Standard English, Language Variation and Change, , –. Kambitsis, Jean. . L’ Antiope d’Euripide. Athens: Editions Elie Hourzamanis. Kannicht, Richard (ed.). . Tragicorum Graecorum Fragmenta, vol. V. Euripides. Gottingen: Vandenhoeck & Ruprecht. Karali, Maria. . Aspects of Delphic Word Order. DPhil thesis, University of Oxford. Karttunen, Lauri. . Presupposition and Linguistic Context. Theoretical Linguistics , –. Karttunen, Lauri and Stanley Peters. . Conventional implicature. In Syntax and Semantics, Volume : Presupposition, ed. Choon-Kyu Oh and David A. Dinneen, –. New York: Academic Press. Kas, Mark. . Essays on Boolean Functions and Negative Polarity. PhD dissertation, University of Groningen. Kay, Paul. . EVEN. Linguistics and Philosophy , –. Kean, Mary-Louise. . The theory of markedness in generative grammar. PhD dissertation, MIT. Keenan, Edward L. . Creating anaphors: an historical study of the English reflexive pronouns. Ms., University of California at Los Angeles. Keenan, Edward L. . Explaining the Creation of Reflexive Pronouns in English. In Studies in the History of the English Language: A Millennial Perspective, ed. D. Minkova and R. Stockwell, –. Berlin–New York: Mouton de Gruyter. Keenan, Edward L. . Historical Syntax and Linguistic Theory. Oxford–New York: Oxford University Press. Kennedy, Christopher. . Projecting the Adjective: The Syntax and Semantics of Gradability and Comparison. PhD dissertation, University of California, Santa Cruz. Kennedy, Christopher. . Comparative Deletion and Optimality in Syntax. Natural Language & Linguistic Theory , –. Kennedy, Christopher. . Comparative Deletion and Optimality in Syntax. Natural Language and Linguistic Theory (), –. Kennedy, Christopher. . Vagueness and Grammar: The Semantics of Relative and Absolute Gradable Adjective. Linguistics and Philosophy (), –. Kennedy, Christopher and Beth Levin. . Measure of Change: The Adjectival Core of Degree Achievements. In Adjectives and Adverbs: Syntax, Semantics, and Discourse, ed. Louise McNally and Christopher Kennedy, –. Oxford: Oxford University Press.



References

Kennedy, Christopher and Louise McNally. . From Event Structure to Scale Structure: Degree Modification in Deverbal Adjectives. In Semantics and Linguistic Theory , ed. Tanya Matthews and Devon Strolovitch, –. Ithaca, NY: CLC Publications. Kennedy, Christopher and Louise McNally. . Scale Structure, Degree Modification and the Semantics of Gradable Predicates. Language , –. Kenny, Anthony. . A Stylometric Study of the New Testament. Oxford: Clarendon. Kent, Roland G. . Old Persian. New Haven, CT: American Oriental Society. Kiecker, Ernst. . Die Stellung des Verbs im Griechischen und in den verwandten Sprachen. Strasbourg: Trübner. Kinsella, Anna. . Language Evolution and Syntactic Theory. Cambridge: Cambridge University Press. Kiparsky, Paul. . Tense and Mood in Indo-European Syntax. Foundations of Language , –. Kiparsky, Paul. . Grammaticalization as Optimization. Ms. [Online.] Available at http:// www.stanford.edu/~kiparsky/Papers/yalegrammaticalization.pdf. Accessed May , . Kiparsky, Paul and Cleo Condoravdi. . Tracking Jespersen’s Cycle. In Proceedings of the nd International Conference of Modern Greek Dialects and Linguistic Theory, ed. M. Janse, B. D. Joseph, and A. Ralli, –. Mytilene: Doukas. Kiss, Katalin E. . Discourse Configurational Languages. New York–Oxford: Oxford University Press. Klein, Jared. . On the Idiomatic Nature of the Gothic New Testament: A Comparative Study of Prepositional Usage in Gothic and New Testament Greek. Transactions of the Philological Society , –. Klein, Jared. . Negation and Polarity in the Greek, Gothic, Classical Armenian, and Old Church Slavic Gospels. Ms. Klein, Jared. . Negation and Polarity in the Greek, Gothic, Classical Armenian, and Old Church Slavic Gospels: A Preliminary Study. In Indo-European Syntax and Pragmatics: Contrastive Approaches, ed. Eric Welo. Oslo Studies in Language (), –. Oslo: University of Oslo. Klima, Edward. . Negation in English. In The Structure of Language, ed. J. A. Fodor and J. J. Katz –. Englewood Cliffs, NJ: Prentice-Hall. Koeneman, Olaf and Hedde Zeijlstra. . The Rich Agreement Hypothesis Rehabilitated. Linguistic Inquiry , –. Koeneman, Olaf and Hedde Zeijlstra. . Introducing Syntax. Cambridge: Cambridge University Press. König, Ekkehard. . The meaning of focus particles. London: Routledge. Koppers, Bertha. . Negative conditional sentences in Greek and some other Indo-European languages. The Hague: Westerbaan. Kovacs, David. . Euripides. (with an English translation by David Kovacs). Cambridge, MA: Harvard University Press. Kratzer, Angelika. . More structural analogies between pronouns and tense. In SALT VIII: Proceedings of the Second Conference on Semantics and Linguistic Theory , ed. Devon Strolovitch and Aaron Lawson, –. Ithaca, NY: CLC Publications.

References



Kretschmer, Paul. . Die Entstehung der Koine. In Sitzungsberichte der Wiener Akademie der Wissenschaften. B. , . Krevans, Nita and Alexander Sens, . Language and Literature. In The Cambridge Companion to the Hellenistic World, ed. Glenn R. Bugh, –. Cambridge: Cambridge University Press. Krifka, Manfred. . The Semantics and Pragmatics of Polarity Items. Linguistic Analysis , –. Krifka, Manfred. . The Origins of Telicity. In Events and Grammar, ed. S. Rothstein, –. Dordrecht: Kluwer. Kroch, Anthony. . Reflexes of Grammar in Patterns of Language Change. Language Variation and Change , –. Kroch, Anthony. . Syntactic Change. In The Handbook of Contemporary Syntactic TheorY, ed. Mark Baltin and Chris Collins, –. Oxford: Blackwell. Krumbacher, Karl. . Beiträge zu einer Geschichte der griechischen Sprach. Weimar: Hof-Buchdruckerei. Krumbacher, Karl. . Byzantinische Zeitschrift. Leipzig: B. G. Teubner. Krumbacher, Karl. . Geschichte der byzantinischen Litteratur von Justinian bis zum Ende des oströmischen Reiches (–), vol. . New York: Burt Franklin. Kühner, Raphael. . Ausführliche Grammatik der griechischen Sprache. Two volumes. Hanover: Hahnsche. Kühner, Raphael. . Grammar of the Greek Language. English translation  by B. B. Edwards and S. H. Taylor. New York: Nabu Press. Kühner, Raphael and Bernhard Gerth. –. Ausführliche Grammatik der griechischen Sprache. Satzlehre. Hannover: Hahn. Labov, William. . Negative Attraction and Negative Concord in English Grammar. Language , –. Labov, William. . Principles of Linguistic Change, Volume. Internal Factors. Oxford: Blackwell. LaCara, Nicholas. . Discourse Inversion and Deletion in as-parentheticals. In Parenthesis and Ellipsis: Cross-Linguistic and Theoretical Perspectives, ed. Marlies Kluck, Dennis Ott, and Mark de Vries, . Berlin: Mouton de Gruyter. Ladusaw, William A. . Polarity sensitivity as inherent scope relations. PhD thesis, University of Texas. Ladusaw, William A. . Expressing Negation. In Semantics and Linguistic Theory (SALT) II, ed. Chris Barker and David Dowty, –, Ithaca, NY: Cornell University Press. Ladusaw, William A. . Negation, Indefinites, and the Jespersen Cycle. BLS , –. Ladusaw, William A. . Negation and Polarity Items. In Handbook of Contemporary Semantic Theory, ed. S. Lappin, –. Oxford: Blackwell. Laird, J. M. . When is Generic μή Particular? The American Journal of Philology , –. Laka, Itziar. . Negation in Syntax: On the Nature of Functional Categories and Projections. PhD dissertation, MIT. Larrivée, Pierre. . L’association négative. Depuis la syntaxe jusqu’à l’interprétation. Geneva: Droz.



References

Larrivée, Pierre. . Is There a Jespersen Cycle? In Evolution of Negation. Beyond the Jespersen Cycle, ed. Richard Ingham and Pierre Larrivée, –. Berlin: Mouton de Gruyter. Lauffenburger, Olivier. . Hittite Grammar. [Online.] Available at http://www.pre miumwanadoo.com/cuneiform.languages/hittite_grammar.pdf. Accessed November , . Lee, Chungmin. . Types of NPIs and nonverdicality in Korean and other languages. In: G. Storto (ed.), Syntax at Sunset  (UCLA Working Papers in Linguistics ). Los Angeles, CA: University of California, –. Lee, Chungmin. . Negative polarity items and free choice in Korean and Japanese: A contrastive study. Korean Society of Bilingualism , –. Lee, Chungmin. . Middle English Negative Concord: How it Competed with Negative Polarity. Presented at the th Medieval English Studies Symposium—MESS , Poznań, Poland. Lee, Jung-Hyuck . Nonveridical dependency: Korean and Japanese focus particles. Japanese/ Korean Linguistics , –. Lee, Young-Suk and Laurence Horn. . Any as Indefinite plus Even. Unpublished ms,Yale University. Legrand, Émile. –. Bibliographie Hellénique ou description raisonnée des ouvrages publiés par des grecs au XVIIe siècle, I–V. Paris: Publications de l’Ecole des langues orientales vivantes. Lehmann, Christian. . Thoughts on Grammaticalization. Munich: Lincom Europa. Lehmann, Christian. . Theory and Method in Grammaticalization. In Grammatikalisierung, Special issue of Zeitschrift für Germanistische Linguistik , ed. G. Diewald, –. Lehmann, Winfred P. . Theoretical Bases of Indo-European Linguistics. London–New York: Routledge. Lendari, Tina. . Livistros and Rodamne. The Vatican version. Athens: ΜΙΕΤ (National Bank of Greece Cultural Foundation). Lenz, Barbara. . Negationsverstärkung und Jespersens Zyklus im Deutschen und in anderen europäischen Sprachen. In Deutsch—typologisch, ed. Ewald Lang and Gisela Zifonun, –. Berlin: Mouton de Gruyter. Lesky, Albin. . Geschichte der Griechischen Literatur. Munich: Saur. Lichtenberk, Frantisek. . On the Gradualness of Grammaticalization. In Approaches to Grammaticalization, vol. , ed. Elizabeth Closs Traugott and B. Heine, –. Amsterdam: John Benjamins. Liddell, Henry George and Robert Scott. . A Greek-English Lexicon. (Revised and augmented throughout by Sir Henry Stuart Jones with the assistance of Roderick McKenzie.) Oxford: Clarendon Press. Lightfoot, David W. . Natural Logic and the Greek Moods: The Nature of the Subjunctive and Optative in Classical Greek. The Hague: Mouton de Gruyter. Lightfoot, David W. . Principles of Diachronic Syntax. Cambridge: Cambridge University Press. Lightfoot, David W. . Explaining Syntactic Change. In Explanation in Linguistics: The Logical Problem of Language Acquisition, ed. Norbert Hornstein and David Lightfoot, –. London–New York: Longman.

References



Lightfoot, David W. . How to set parameters: arguments from language change. Cambridge, MA: MIT. Lightfoot, David W. . Why UG Needs a Learning Theory: Triggering Verb Movement. In Clause Structure and Language Change, ed. Adrian Battye and Ian Roberts, –. New York: Oxford University Press. Lightfoot, David W. . The Development of Language: Acquisition, Change, and Evolution. Oxford: Blackwell. Lightfoot, Douglas J. . Can the Lexicalization/Grammaticalization Distinction be Reconciled? Studies in Language , –. Lin, Jo-Wang. . Polarity Licensing and Wh-Phrase Quantifi cation in Chinese. PhD dissertation, University of Massachusetts, Amherst, MA. Lin, Nina Yuhsun. . A Corpus-Based Study on the Development of Mandarin Negative Marker mei. Paper presented at th International Conference on Sino-Tibetan Languages and Linguistics, Arizona State University, Tempe, November –. Lindberg, David C. . The Beginnings of Western Science. The European Scientific Tradition in Philosophical, Religious, and Institutional Context, Prehistory A.D. , nd edition Chicago, IL: University of Chicago Press. Linebarger, Marcia C. . The Grammar of Negative Polarity. PhD dissertation, MIT. [Reproduced by the Indiana University Linguistics Club, Indiana, ]. Linebarger, Marcia C. . Negative Polarity and Grammatical Representation. Linguistics and Philosophy , –. Ljungvik, Herman. . Beitrage zur Syntax der spätgriechischen Volkssprache. Uppsala: Almqvist och Wiksell. Lobeck, Christian August. . Phrynichi Eclogae Nominum et Verborum Atticorum. Leipzig: Weimann. Longobardi, Giuseppe. . Formal Syntax, Diachronic Minimalism, and Etymology: The History of French ‘chez’. Linguistic Inquiry (), –. Lopez-Couso, M. J. and E. Seoane. . New Perspectives on Grammaticalization. In Rethinking Grammaticalization: New Perspectives, ed. María José López-Couso and Elena Seoane, –. Amsterdam: John Benjamins. Louie, Megan. . Atemporal Anchoring of Individuals, Events and Sub-Events in Blackfoot. MA thesis. University of Toronto. Lucas, Christopher. . Jespersen’s Cycle in Arabic and Berber. Transactions of the Philological Society (), –. Madvig, Johan Nikolai. . Syntax der griechíschen Sprache, besonders der attíschen Sprachform, für Schulen. Braunschweig: Bieweg. Mandilaras, B. G. . The Verb in the Greek Non-Literary Papyri. Athens: Hellenic Ministry of Culture and Sciences. Mann, Stuart E. . An Armenian Historical Grammar in Latin Characters. London: Luzac and Company. Manzini, Maria Ritta and Leonardo M. Savoia. . I dialetti italiani e romanci. Morfosintassi generativa.  vols. Alexandria: Orso. Margariti, Anna Maria. . Quantification at the Syntax-Semantics interface: Greek every NPIs. PhD dissertation, University of Patras, Greece.



References

Margariti, Anna Maria. . Quantification at the Syntax-Semantics Interface: Greek every NPIs. Journal of Greek Linguistics (), –. Markopoulos, Theodore. . The Future in Greek. From Ancient to Medieval. Oxford: Oxford University Press. Markopoulos, Theodore. . Case Overlap in Medieval Cypriot Greek: A Socio-Historical Perspective. Folia Linguistica Historica , –. Markopoulos, Theodore. . Language Contact in the Middle Ages: Rhodes under the Hospitallers. Presented at the International Conference on Greek Linguistics—ICGL , Democritus University of Thrace, Komotini, Greece. Martin, Raymond A. . Syntax Criticism of the Synoptic Gospels. Lewiston, NY: The Edwin Mellen Press. Master, Alfred. . The Zero Negative in Dravidian. Transactions of the Philological Society , –. Mathesius, Vilém. . Double Negation and Grammatical Concord. Mélanges de linguistique et de philology offert à Jacq. van Ginneken à l’occasion du soixantième annìversaire de sa naissance ( avril ), –. Paris: Klincksieck. Matthewson, Lisa. . Determiner Systems and Quantificational Strategies. Evidence from Salish. The Hague: Holland Academic Graphics. Matthewson, Lisa, Henry Davis, and Hotze Rullmann. . Evidentials as Epistemic Modals: Evidence from St’´at’imcets. In Linguistic Variation Yearbook , vol. , ed. J. van. Craenenbroeck, –. Amsterdam: John Benjamins. Maynard, Senko. . Linguistic Emotivity. Amsterdam: John Benjamins. Mazzon, Gabriella. . A History of English Negation. Harlow: Longman. McCawley, James D. . Contrastive negation and metalinguistic negation. In Proceedings of the Chicago Linguistic Society : The Parasession on Negation, –. Meier-Brügger, M. a. Griechische Sprachwissenschaft. I: Bibliographie. Einleitung. Syntax. Berlin/New York: Walter de Gruyter. Meier-Brügger, M. b. Griechische Sprachwissenschaft. II: Wortschatz. Formenlehre. Lautlehre. Indizes. Berlin/New York: Walter de Gruyter. Meillet, Antoine. . L’évolution des formes grammaticales. Scientia (Rivista di Scienza) (), –. Meillet, Antoine. . Geschichte des Griechischen. Heidelberg: Carl Winter’s Universitätabuchhandlung. Meillet, Antoine. . Grammaire du Vieux-Perse, nd edition. Paris: Librairie Ancienne Honoré Champion. Meillet, Antoine.  []. Apercu d’une histoire de la langue grecque, th edition. Paris: Klincksieck. Mendez-Dosuna, Julian. . La valeur de l’optativ grec: un regard fonctionel-typologique. In Les completeives en Grec ancient, ed. B. Jacquinod, –. Saint Etienne: University of Saint Etienne. Merchant, Jason. . The Syntax of Silence. Oxford: Oxford University Press. Merchant, Jason. . Fragments and Ellipsis. Linguistics and Philosophy (), –. Merchant, Jason. . ‘Why No(t)?’. Style (/), –. (Special issue edited by William Salmon and Charalabos Kalpakidis as a Festschrift for Haj Ross).

References



Merchant, Jason. . Polarity items under ellipsis. In Diagnosing Syntax, ed. Lisa Lai-Shen Cheng and Norbert Corver, –. Oxford: Oxford University Press. Mertyris, Dionysios. . Case Syncretism in the Personal Pronouns of Medieval Greek: The Loss of the Genitive Plural. In Proceedings of the International Conference on Greek Linguistics ICGL , ed. Katerina Chatzopoulou, Alexandra Ioannidou, and Suwon Yoon, –. [Online.] Available at http://www.ling.ohio-state.edu/ICGL/proceedings.html. Mesthrie, Rajend. . Koinés. In Concise Encyclopedia of Sociolinguistics, ed. R. Mesthrie, –. Amsterdam: Elsevier. Mettouchi, Amina. . La négation dans les langues du Maghreb: Synthèse. In La négation en berbère et en Arabe Maghrébin, ed. Salem Chaker and Dominique Caubet, –. Paris: Harmattan. Michael, Lev. . The Nanti Reality Status System: Implication for the Typological Validity of the Realis/Irrealis Contrast. Linguistic Typology (), –. Miestamo, Matti. . Towards a Typology of Standard Negation. Nordic Journal of Linguistics. , –. Miestamo, Matti. . Clausal negation: A typological study. PhD dissertation, Helsinki: University of Helsinki. Miestamo, Matti. a. Standard Negation: The Negation of Declarative Verbal Main Clauses in a Typological Perspective. Berlin: Mouton de Gruyer. Miestamo, Matti. b. Symmetric and Asymmetric Standard Negation. In The World Atlas of Language Structures, ed. Martin Haspelmath, Matthew Dryer, David Gil, and Bernard Comrie, –. Oxford: Oxford University Press. Miestamo, Matti. c. Subtypes of Asymmetric Standard Negation. In The World Atlas of Language Structures, ed. Martin Haspelmath, Matthew Dryer, David Gil, and Bernard Comrie, –. Oxford: Oxford University Press. Miestamo, Matti. . Negation. In Grammar, Meaning and Pragmatics (Handbook of Pragmatics Highlights ), ed. Frank Brisard, Jan-Ola Östman, and Jef Verschueren, –. Amsterdam: John Benjamins. Miestamo, Matti. . Negatives Without Negators. In Rethinking Universals: How Rarities Affect Linguistic Theory, ed. Jan Wohlgemuth and Michael Cysouw, –. Berlin: Mouton de Gruyter. Miestamo, Matti and Johan van der Auwera. . Negative declaratives and negative imperatives: similarities and differences. In Linguistics festival. May , Bremen, ed. Andreas Ammann, –. Bochum: Brockmeyer. Mirambel, André. . Négation et mode en grec modern. Bulletin de la société linguistique de Paris , –. Mirambel, André. . La Langue Grecque Moderne: Description et Analyse. Paris: Librairie Klincksieck. Mirambel, André. . Participe et gérondif en grec médiéval et moderne. Bulletin de la Société de Linguistique de Paris , –. Mithun, Marianne. . Is Basic Word Order Universal? In Coherence and Grounding in Discourse, ed. Russell S. Tomlin, –. Amsterdam–Philadelphia, PA: John Benjamins. Mithun, Marianne. . What Cycles When and Why? In Cyclical Change Continued, ed. Elly van Gelderen, –. Amsterdam: John Benjamins.



References

Montague, Richard. . On the Nature of Certain Philosophical Entities. The Monist , –, Reprinted  in Formal Philosophy: Selected Papers of Richard Montague, edited and with an introduction by R. H. Thomason, –. New Haven, CT: Yale University Press. Moorhouse, Alfred C. . Studies in the Greek Negatives. Cardiff: University of Wales Press. Moorhouse, Alfred C. . A Use of oὐδείς and μηδείς. Classical Quarterly, n.s. , –. Moule, C. F. D. . An Idiom Book of New Testament Greek. repr. Oxford: Oxford University Press. Moulton, J. H. . A Grammar of New Testament Greek. Volumes I–IV. Edinburgh: T. & T. Clark. Moulton, J. H. and W. F. Howard. . A Grammar of New Testament Greek. II. Accidence and Word-Formation. Edinburgh: T. & T. Clark. Mufwene, Salikoko. . The Ecology of Language Evolution. Cambridge: Cambridge University Press. Muller, Claude. . La négation en français. Syntaxe, sémantique et éléments de comparaison avec les autres langues romanes. Geneva: Droz. Murray, Sarah E. . Evidentiality and Questions in Cheyenne. In Proceedings of SULA , ed. S. Lima, –. Amherst, MA: GLSA Publications. Mustajoki, Arto and Hannes Heino. . Case Selection for the Direct Object in Russian Negative Clauses. Part II: Report on a Statistical Analysis. (Slavica Helsingiensia ). Helsinki: University of Helsinki. Muysken, Pieter. . Functional Categories. Cambridge: Cambridge University Press. Nevala, Minna. . Family First: Address and Subscription Formulae in English Family Correspondence from the Fifteenth to the Seventeenth Century. In Diachronic Perspectives on Address Term Systems, ed. Andreas Jucker and Irma Taavitsainen, –. Amsterdam: John Benjamins. Newmeyer, Frederick J. . Language Form and Language Function. Cambridge, MA: MIT Press. Newmeyer, Frederick J. . Deconstructing Grammaticalization. Language Sciences , –. Noel Aziz Hanna, Patrizia. . Jespersen’s Cycle and the Issue of Prosodic Weakness. In Advances in Comparative Germanic Syntax, ed. Artemis Alexiadou, Jorge Hankamer, Thomas McFadden, Justin Nuger, and Florian Schäfer (Proceedings of CGSW  and ), –. Amsterdam–Philadelphia, PA: John Benjamins. Norde, Muriel. . Deflexion as a counterdirectional factor in grammatical change. Language Sciences , –. Norlin, G. . Isocrates, with an English Translation (three volumes). Cambridge, MA: Harvard University Press; London: William Heinemann Ltd; New York: CLC Publications. Nurse, Derek. . Tense and Aspect in Bantu. Oxford: Oxford University Press. Ofen, N., S. Whitfield-Gabrieli, X. J. Chai, R. F. Schwarzlose, and J. D. E. Gabrieli. . Neural Correlates of Deception: Lying about Past Events and Personal Beliefs. Social Cognitive and Affective Neuroscience (), –. Ouhalla, Jamal. . Sentential Negation, Relativized Minimality and the Aspectual Status of Auxiliaries. The Linguistic Review , –.

References



Ouhalla, Jamal. . Functional Categories and Parametric Variation. London and New York: Routledge. Ouhalla, Jamal. . Subject-Extraction, Negation and the Anti-Agreement Effect. Natural Language & Linguistic Theory , –. Palmer, Frank R.  []. Mood and Modality. Cambridge: Cambridge University Press. Papadopoulos, Theodoros and Antonios Ioannidis. . Η αναγκαιότητα κατάργησης των τόνων και των πνευμάτων επί τη βάσει της διατριβής του ιατροφιλοσόφου Ν. Β. Φαρδύ: «Περι ατονου και απνευματιστου γραφης της νεωτερας ελληνικης γλωσσης. Μαρτυριαι και αποδειξεις.». In Proceedings of the th Panhellenic Historical Conference (ΚΔ’ Πανελλήνιο Ιστορικό Συνέδριο), –, Thessaloniki: Ελληνική Ιστορική Εταιρεία (Eliniki istoriki eteria, = Greek historical company). Papafragou, Anna. . Epistemic Modality and Truth Conditions. Lingua , –. Pappas, Panayiotis A. . The Microcosm of a Morphological Change: Variation in thelō + Infinitive Futures and ēthela + Infinitive Counterfactuals in Early Modern Greek . Diachronica , – . Pappas, Panayiotis A. . Variation and Morphosyntactic Change in Greek: from Clitics to Affixes. New York: Palgrave Macmillan. Pappas, Panayiotis A. and Brian D. Joseph. . The Development of the Greek Future System: Setting the Record Straight. In Greek Linguistics ’. Proceedings of the th International Conference on Greek Linguistics, Nicosia, September , –. Thessaloniki: University Studio Press. Paradis, Carita. . Adjectives and Boundedness. Cognitive Linguistics (), –. Paradis, Carita. . Between Epistemic Modality and Degree: The Case of really. In Modality in Contemporary English, ed. Roberta Facchinetti, Manfred Krug, and Frank Palmer, –. Berlin and New York: Mouton de Gruyter. Partee, B. . The syntax and semantics of quotation. In A festschrift for Morris Halle, ed. S. Anderson and P. Kiparsky, –. New York: Holt, Reinhart and Winston. Partee, Barbara H. . Nominal and temporal anaphora. Linguistics and Philosophy , –. Partee, Barbara H. . Negation, Intensionality, and Aspect: Interaction with NP Semantics. In Theoretical and Crosslinguistic Approaches to the Semantics of Aspect, ed. S. Rothstein, –. Amsterdam: John Benjamins. Pauwels, Jan Lodewijk. . Het dialect van Aarschot en omstreken. Brussel: Belgisch Interuniversitair Centrum voor Neerlandistiek. Payne, John R. . Negation. In Language Typology and Syntactic Description, Volume I, Clause Structure, ed. Timothy Shopen, –. Cambridge: Cambridge University Press. Pederson, Eric. . Zero Negation in Dravidian. Chicago Linguistic Society (The Parasession on Negation) ., –. Philippaki–Warburton, Irene. . The theory of empty categories and the pro-drop parameter in Modern Greek. Journal of Linguistics , –. Philippaki-Warburton, Irene. . Subjects in English and in Greek. Ms, University of Reading. Philippaki-Warburton, Irene. . On mood and complementizers in Modern Greek. Reading University Working Papers in Linguistics , –.



References

Philippaki-Warburton, Irene. . The subjunctive mood and the syntactic status of the particle na in Modern Greek. Folia Linguistica  (–), –. Philippaki-Warburton, Irene. . Functional Categories in Modern Greek Syntax. The Linguistic Review (–), –. Philippaki-Warburton, Irene and Vassilis Spyropoulos. . A Change of Mood: The Development of the Greek Mood System. Linguistics –, –. Philippaki-Warburton, Irene and Ioannis Veloudis. . Η υποτακτική στις συμπληρωματικές προτάσεις. Studies in Greek Linguistics , –. Pilot-Raichoor, Christiane. . Le zéro negative dans les conjugaisons dravidiennes. Fait de langues , –. Pilot-Raichoor, Christiane. . Enjeux théoriques et éthodiques du zéro negative dravidien. Actances , –. Pilot-Raichoor, Christiane. . The Dravidian Zero Negative: Diachronic Context of its Morphogenesis and Conceptualization. In Rara & Rarissima: Documenting the Fringes of Linguistic Diversity, ed. Jan Wohlgemuth and Michael Cysouw, –. Berlin: Mouton de Gruyter. Pinker, S. and M. Ullman. . The Past and Future of the Past Tense. Trends in Cognitive Science, , –. Platzack, Christer. . The Loss of Verb Second in English and in French. In Clause Structure and Language Change, ed. Adrian Battye and Ian Roberts, –. New York: Oxford University Press. Pohl, Jacques. . Ne dans le français parlé contemporain: les modalités de son abandon. Actas del XI congresso internacional de linguistica y filología Romanicas, –. Madrid. Poletto, Cecilia and Raffaella Zanuttini. . Emphasis as Reduplication: Evidence from sì che/no che Sentences. Lingua , –. Politis, Nikolaos. . Eklogai apo ta Tragoudia tou Ellinikou Laou [Selections from the Songs of the Greek People]. Athens: Estia. Pollock, Jean-Yves. . Verb Movement, Universal Grammar and the Structure of IP. Linguistic Inquiry (), –. Porter, Stanley E. . Verbal Aspect in the Greek of the New Testament, with Reference to Tense and Mood. New York–Bern: Peter Lang. Porter, Stanley E. . Idioms of the Greek New Testament. Sheffield: Sheffield Academic Press. Portner, Paul. . The semantics of imperatives within a theory of clause types. In Proceedings of Semantics and Linguistic Theory , ed. K. Watanabe and R. B. Young. Ithaca, NY: CLC Publications, Cornell University Linguistics Department. Posner, Rebecca. . Post-Verbal Negation in Non-Standard French: A Historical and Comparative View. Romance Philology , –. Postgate, John P. . Grammatical Annotations upon the Oedipus Rex. Transactions of the Cambridge Philological Society  (–), –. Potts, Christopher. . The Logic of Conventional Implicatures. Oxford: Oxford University Press. Potts, Christopher and Shigeto Kawahara. . Japanese Honorifics as Emotive Definite Descriptions. In Proceedings of SALT , ed. K. Watanabe and R. B. Young, –. Ithaca, NY: CLC.

References



Poulos, George and Bosch, Sonja E. . Zulu. (Languages of the World/Materials, ) München: Lincom Europa. Probert, Philomen. . Ancient Greek Accentuation. New York: Oxford University Press. Progovac, Ljiljana. . A Binding Approach to Polarity Sensitivity, PhD dissertation, University of Southern California, Los Angeles. Progovac, Ljiljiana. . Negative Polarity: A Semantico-Syntactic Approach. Lingua , –. Progovac, Ljiljiana. . A Syntax of Serbian: Clausal Architecture. Bloomington, IN: Slavica Publishers. Prüfer, Ernst. . De μή et οe particulis epitome. Breslau. Pulleyblank, Edwin. . Outline of Classical Chinese Grammar. Vancouver: University of British Columbia Press. Quer, Josep. . Mood at the Interface. PhD dissertation, University of Utrecht. Quer, Josep. . Mood Management: An Updated Toolkit. Lingua , –. Rice, Keren. . A Grammar of Slav. Berlin: Mouton de Gruyter. Rijksbaron, Albert.  []. The Syntax and Semantics of the Verb in Classical Greek. Chicago, IL: University of Chicago Press. Rijksbaron, Albert. . Does Ancient Greek Have a Word for ‘No’? Journal of Greek Linguistics (), –. Rivero, Maria Luisa. . The Location of Non-Active Voice in Albanian and Modern Greek. Linguistic Inquiry , –. Rivero, María Luisa. . Negation, Imperatives and Wackernagel Effects. Rivista di Linguistica (), –. Rivero, María Luisa and Arhonto Terzi. . Imperative, V-movement and Logical Mood. Journal of Linguistics , –. Rizzi, Luigi. . A Restructuring Rule in Italian Syntax. In Recent Transformational Studies in European Languages, ed. S. J. Keyser, –. Cambridge, MA: MIT Press. Rizzi, Luigi. . The Fine Structure of the Left Periphery. In Elements of Grammar: Handbook of Generative Syntax, ed. Liliane Haegeman, –. Dordrecht–Boston, MA: Kluwer Academic Publishers. Rizzi, Luigi. a. The Structure of CP and IP: The Cartography of Syntactic Structures, vol. . New York: Oxford University Press. Rizzi, Luigi. b. Locality and the Left Periphery. In Structures and Beyond. The Cartography of Syntactic Structures, ed. A. Belletti, vol. , –. New York: Oxford University Press. Roberts, Craige. . Information Structure in Discourse: Towards an Integrated Formal Theory of Pragmatics, Technical report, Linguistics Department, The Ohio State University, Columbus, Ohio. OSU Working Papers in Linguistics, volume , ed. Jae-Hak Yoon and Andreas Kathol. Roberts, Ian. . Verbs and Diachronic Syntax: A Comparative History of English and French. Dordrecht: Kluwer. Roberts, Ian. . Grammaticalization, the Clausal Hierarchy and Semantic Bleaching. In Gradience, Gradualness and Grammaticalization, ed. Elizabeth Closs Traugott and Graeme Trousdale, –. Cambridge: Cambridge University Press.



References

Roberts, Ian. . Towards a parameter hierarchy for verb-movement: diachronic considerations. Presentation at Deutsche Gesellschaft für Sprachwissenschaft—DGfS , Frankfurt, March . Roberts, Ian and Anna Roussou. . Syntactic Change. A Minimalist Approach to Grammaticalization. Cambridge: Cambridge University Press. Robertson, A. T. . A Grammar of the Greek New Testament in Light of Historical Research. New York: Hodder & Stoughton. Romero, Clara. . Pour une définition générale de l’intensité dans le langage. Travaux de linguistique, , La scalarité, –. Rooth, Mats. . Association with Focus. PhD dissertation, University of Massachusetts, Amherst, MA. Roszko, F. Casimir. . A Classical Armenian Grammar. Indianapolis, IN: Catholic Seminary Foundation of Indianapolis. Rouchota, Vassiliki. . Relevance Theory and na-interrogatives in Modern Greek. Lingua , –. Rouchota, Vassiliki. . The Semantics and Pragmatics of Subjunctive in Modern Greek: A Relevance-Theoretic Approach. PhD dissertation, University College London. Roussou, Anna. . Minimalism and Diachronic Syntax: The Development of Negative Expressions. In Festschrift for Prof E Panagopoulos, th International Symposium on Theoretical and Applied Linguistics, ed. E. Agathopoulou, M. Dimitrakopoulou, and D. Papadopoulou, –. Ruipérez, Martin. S. . Esquisse d’une histoire vocalisme grec. Word , –. Sadock, Jerry M. . Queclaratives. In Papers from the Seventh Regional Meeting of the Chicago Linguistic Society , –. Sadock, Jerry M. . Towards a Linguistic Theory of Speech Acts, New York, San Francisco, CA, London: Academic Press. Sadock, Jerry M. . Autolexical Syntax: A Theory of Parallel Grammatical Representations. Chicago, IL: University of Chicago Press. Sadock, Jerry M. . The Modular Architecture of Grammar. Cambridge: Cambridge University Press. Sadock, Jerry M. and Eric Schiller. . The Generalized Interface Principle. CLS . –. Sampanis, Konstantinos. . A semantic features analysis for the Subjunctive mood. Wiener Linguistische Gazette Institut für Sprachwissenschaft Universität Wien A.–. Sadock, Jerry M. and Arnold M. Zwicky. . Speech Act Distinctions in Syntax. In Language Typology and Syntactic Description I: Clause Structure, –. Cambridge: Cambridge University Press. Sapir, Edward. . Grading: A Study in Semantics. Philosophy of Science , –. Savickiene, Ineta. . The Acquisition of Diminutives in Lithuanian. In Studies in the Acquisition of Number and Diminutive Marking, ed. S. Gillis, –. Antwerp: Universiteit Anwerpen. Schaffer, Deborah. . Can Rhetorical Questions Function as Retorts? Is the Pope Catholic? Journal of Pragmatics, , –. Schlicher, John J. . The Moods of Indirect Quotation. American Journal of Philology (), –.

References



Schmerling, Susan. . A Note on Negative Polarity. Papers in Linguistics (), –. Schmitt, Rüdiger. . Grammatik des Klassisch-Armenischen. mit sprachvergleichenden Erläuterungen. Innsbruck: Innsbrucker Beiträge zur Sprachwissenschaft. Schwarz, Bernhard and Rajesh Bhatt. . Light Negation and Polarity. In Crosslinguistic Research in Syntax and Semantics: Negation, Tense and Clausal Architecture, ed. Raffaella Zanuttini, Héctor Campos, Elena Herburger, and Paul H. Portner, –. Washington, DC: Georgetown University Press. Schwegler, Armin. . Predicate Negation and Word-Order Change: A Problem of Multiple Causation. Lingua (), –. Schwegler, Arnim. . Word-Order Changes in Predicate Negation Strategies in Romance Languages. Diachronica , –. Schwegler, Arnim. . Analyticity and Syntheticity: A Diachronic Perspective with Special Reference to Romance Languages. Berlin: Mouton de Gruyter. Schwenter, S. A. . Fine-tuning Jespersen’s Cycle. In Drawing the Boundaries of Meaning: Neo-Gricean Studies in Pragmatics and Semantics in Honor of Laurence R. Horn, ed. B. J. Birner and G. Ward, –. Amsterdam: John Benjamins. Schwyzer, Eduard and Albert Debrunner.  []. Griechische Grammatik,  vol. Munich: Beck. Scott-Phillips, Thom C. . Pragmatics and the Aims of Language Evolution. Psychonomic Bulletin and Review (), –. Semenov, Anatol F. . The Greek Language and its Evolution. London: George Allen and Unwin. Sevdali, Christina. . Tense and Ancient Greek Infinitives. In On Space and Time in Language, ed. M. Coene and L. Tasmowski, –. Cluj-Napoca: Clusium. Sevdali, Christina. . Infinitival Clauses in Ancient Greek: Overt and Null Subjects, the Role of Case and Focus. PhD dissertation, Cambridge University. Sevdali, Christina. . On the Various Verbs ‘seem’ in Ancient Greek and the Role of Datives. In Current Issues in Mediterranean Syntax, ed. A. Sumru Ozsoy and A. Gurel, –. Leiden: Brill Academic Publishing. Sharp, D. S. . Epictetus and the New Testament. London: Kelly. Shawcross, Teresa, M. . The Chronicle of Morea: Historiography in Crusader Greece. Oxford–New York: Oxford University Press. Sicking, Christian. M. . Particles in Questions in Plato. In New Approaches to Greek Particles, ed. A. Rijksbaron, –. Amsterdam: Gieben. Siegel, Jeff. . Koines and Koineization. Language in Society (), –. Sim, Margaret G. . Underdeterminacy in Greek Particles. In Bible Translator , –. Sim, Margaret G. . Marking Thought and Talk in New Testament Greek: New Light from Linguistics on the Particles ἵνα and ὅτι. Eugene, OR: Pickwick Publications. Sitaridou, Ioanna. . Greek-Speaking Enclaves in Pontus Today: The Documentation and Revitalization of Romeyka. In Keeping Languages Alive. Language Endangerment: Documentation, Pedagogy and Revitalization, ed. M. Jones and S. Ogilvie, –. Cambridge: Cambridge University Press. Sitaridou, Ioanna. a. Modality, Antiveridicality, and Complementation: The Romeyka Infinitive as a Negative Polarity Item. Lingua , –.



References

Sitaridou, Ioanna. b. The Romeyka Infinitive: Continuity, Contact and Change in the Hellenic Varieties of Pontus. Diachronica (), –. Smith, Neil V. . Consistence, Markedness and Language Change: On the Notion ‘Consistent Language’. Journal of Linguistics (), –. Smyth, Herbert W. . Greek Grammar. Cambridge, MA: Harvard University Press. Smyth, Herbert W. . Greek Grammar. Revised by Gordon M. Messing. Harvard: Harvard University Press. Sophocles, E. A. . A Glossary of Later and Byzantine Greek. Boston, MA: Welch, Bigelow. Speas, Margaret. a. Evidentiality, logophoricity, and the syntactic representation of pragmatic features. Lingua , –. Speas, Margaret. b. Evidential paradigms, world variables and person agreement features. The Italian Journal of Linguistics , –. Speranza, J. L. and Lawrence Horn. . A Brief History of Negation. In Logic: A History of its Central Concepts. Vol. , Handbook of the History of Logic, –. Amsterdam: Elsevier. Sperber, Dan and Gloria Origgi. . A Pragmatic Perspective on the Evolution of Language. In The Evolution of Human Language, ed. R. K. Larson, V. Déprez, and H. Yamakido, –. Cambridge: Cambridge University Press. Stahl, Johann. . Kritisch-historisch Syntax des griechischen Verbums der klassischen Zeit. Heidelberg: C. Winter. Stalnaker, Robert C. . Indicative Conditionals. Philosophia , –. Stalnaker, Robert C. . Assertion. In Syntax and Semantics : Pragmatics, ed. P. Cole, –. New York: Academic Press. Starke, Michail. . Nanosyntax: A Short Primer to a New Approach to Language. Nordlyd , –. Sturz, Friedrich W. . De Dialecto Macedonica et Alexandrina. Leipzig: Weigel. Swarup Misra, Satya. . The Avestan: A Historical and Comparative Grammar. Varanasi: Chaukhambha Orientalia. Tanaka, Shin. . Die Mehrfachnegaton: Ein Sprachgut der Unraffinienten. In Sprache— Sprechen—Handeln: Akten des . Linguistischen Kolloquiums, ed. Dieter W. Halwachs and Irmgard Stütz, –. Tübingen: Niemeyer. Tancredi, Christopher. a. A Multi-Model Modal Theory of I-semantics. Part I: Modals. Tokyo: University of Tokyo. Tancredi, Christopher. b. A Multi-Model Modal Theory of I-semantics. Part II: Identity and Attitudes. Tokyo: University of Tokyo. Taylor, Ann. . The Change from SOV to SVO in Ancient Greek. Language Variation and Change , –. Teodorsson, Sven-Tage. . The Phonemic System of the Attic Dialect – . Göteborg–Lund: Berlingska Boktryckeriet. Teodorsson, Sven-Tage. . The Phonology of Ptolemaic Koiné. Lund: Berlingska Boktryckeriet. Teodorsson, Sven-Tage. . The Phonology of Attic in the Hellenistic Period. Uppsala: Almqvist & Wiksell. Terkourafi, Marina. . Understanding the Present through the Past: Processes of Koineisation on Cyprus. Diachronica , –.

References



Tesnière, Lucien. . Eléments de syntaxe structural. Paris: Klincksieck. Thomason, Olga A. . Prepositional systems in Biblical Greek, Gothic, Classical Armenian, and Old Church Slavic, PhD dissertation, University of Georgia. Thomason, S. G. . Language Contact: An Introduction. Washington, DC: Georgetown University Press. Threatte, Leslie. . The Grammar of Attic Inscriptions. Berlin and New York: Walter de Gruyter. Tomić, Olga. M. . Negation and Imperatives. Catalan Working Papers in Linguistics , –. Tomić, Olga M. . The Balkan Sprachbund Properties. In Balkan Syntax and Semantics, ed. Olga Tomić, –. Amsterdam: John Benjamins. Torrey, Charles C. . Our Translated Gospels. Some of the Evidence. New York: Harper. Toufexis, Notis. . Diglossia and Register Variation in Medieval Greek. Byzantine and Modern Greek Studies (), –. Traugott, Elizabeth Closs. . Legitimate counterexamples to unidirectionality. Paper presented at Freiburg University, October . Traugott, Elizabeth Closs. . The status of onset contexts in analysis of micro-changes. In English Corpus Linguistics: Crossing Paths, ed. M. Kytö, –. Amsterdam: Rodopi. Traugott, Elizabeth Closs and Richard B. Dasher. . Regularity in Semantic Change. Cambridge: University of Cambridge Press. Traugott, Elizabeth Closs and B. Heine (eds). . Approaches to Grammaticalization I and II. Amsterdam: John Benjamins. Traugott, Elizabeth and Graeme Trousdale (eds). a. Gradience, Gradualness and Grammaticalization. Amsterdam–New York: John Benjamins. Traugott, Elizabeth Closs and Graeme Trousdale. b. Gradience, Gradualness and Grammaticalization: How do they Intersect? In Gradience, Gradualness and Grammaticalization, ed. Elizabeth Closs Traugott and Graeme Trousdale, –. Amsterdam: John Benjamins. Trips, Carola. . From OV to VO in Early Middle English. Amsterdam: John Benjamins. Trubetzkoy, Nikolai. . Phonologie et géographie linguistique. Travaux du Cercle Linguistique de Prague , –. Trudgill, Peter. . Dialects in Contact. Oxford: Blackwell Publishing. Trypanis, C. A. . Early Medieval Greek ἵνα. Glotta , –. Tsangalidis, A. . Will and Tha: A Comparative Study of the Category Future. Thessaloniki: University Studio Press. Tsimpli, Ianthi M. . The Clause Structure and Word Order in Modern Greek. UCL Working papers in Linguistics , –. Tsimpli, Ianthi M. and Anna Roussou. . Negation and Polarity Items in Greek. The Linguistic Review , –. Tsurska, Olena. . The Negative Cycle in Early and Modern Russian. In Cyclical Change, ed. Elly van Gelderen, –. Amsterdam–Philadelphia, PA: John Benjamins. Turner, Nigel. . A Grammar of New Testament Greek III. Syntax. Edinburgh: T. & T. Clark. Turner, Nigel. . A Grammar of New Testament Greek IV. Style. Edinburgh: T. & T. Clark. Tzartzanos, Achilleas.  []. Syntaktikon tis Arxaias Ellinikis Glossis. Thessaloniki: Kiriakidis Bros.



References

Uljas, Sami. . The Modal System of Earlier Egyptian Complement Clauses. Leiden: Brill. Uribe-Etxebarria, Miriam. . Interface licensing conditions on negative polarity items: a theory of polarity and tense interactions. PhD dissertation, University of Connecticut, Storrs. Van Alsenoy, Lauren and Johan van der Auwera. . On the relation between double clausal negation and negative concord. Ms.van Craenenbroek, Jeroen. . About something. Towards a syntactic decomposition of polarity. Paper presented at the CUNY Syntax Supper, CUNY Graduate Center at the City University of New York, November , . Van den Wyngaerd, Guido. . Measuring Events. Language , –. Van der Auwera, Johan. . Prohibitives: Why Two Thirds of the World’s Languages are Unlike Dutch. In Proceedings of the Fifteenth Amsterdam Colloquium. December –, , ed. P. Dekker and M. Francke, –. Amsterdam: University of Amsterdam, ILLC. Van der Auwera, Johan. . Why Languages Prefer Prohibitives. 外国语 [Wai guo yu— Journal of Foreign Languages] , –. Van der Auwera, Johan. . Gardiner’s cycle. Unpublished manuscript, Universiteit Antwerpen. Van der Auwera, Johan. . The Jespersen Cycles. In Cyclical Change, ed. E. van Gelderen, –. Amsterdam: John Benjamins. Van der Auwera, Johan. a. On the Diachrony of Negation. In The Expression of Negation, ed. L. R. Horn. Berlin–New York: Mouton de Gruyter. Van der Auwera, Johan. b. Prohibition: Constructions and Markers. In Contrasting Meaning in Languages of the East and West. ed. D. Shu and K. Turner, –. Tübingen: Narr. Van der Auwera, Johan. . The Jespersen Cycles. In Cyclical Change, ed. Elly van Gelderen, –. Amsterdam: John Benjamins. Van der Auwera, Johan and D. N. S. Bhat. . Modality and Negation in Havyaka Kanada. Indian Linguistics , –. Van der Auwera, Johan and M. Devos. . Irrealis in Positive Imperatives and Prohibitives. Language Sciences , –. Van der Auwera, Johan and Paul Gybels. . On Negation, Indefinites and Negative Indefinites in Yiddish. In Yiddish Language Structures, ed. Björn Hansen and Marion Aptroot, –, Berlin: Walter de Gruyter. Van der Auwera, Johan and Ludo Lejeune. . The Prohibitive. In World Atlas of language Structures, ed. Martin Haspelmath, Matthew Dryer, D. Gil, and B. Comrie, –. Oxford: Oxford University Press. Van der Auwera, Johan and Annemie Neuckermans. a. Jespersen’s Cycle and the Interaction of Predicate and Quantifier Negation. In Dialectology Meets Typology. Dialect Grammar from a Cross-Linguistic Perspective, ed. Bernd Kortmann, –. Berlin: Mouton de Gruyter. Van der Auwera, Johan and Annemie Neuckermans. b. Een Oost-Vlaamse driedubbele negatie [An East Flemish triple negation]. In Dialectsyntaxis in bloei [Dialectsyntax in bloom], ed. S. Barbiers, M. Devos, and G. de Schutter, –. Gent: Koninklijke Academie voor Nederlandse Taal-en Letterkunde.

References



Van der Auwera, Johan and Lauren van Alsenoy. . Indefinite Pronouns, Synchrony and Diachrony: Comments on Willis. In Evolution of Negation. Beyond the Jespersen Cycle ed. Richard Ingham and Pierre Larrivée, –. Berlin: Mouton de Gruyter. Van der Auwera, Johan and Frens Vossen. . Negatives between Chamic and Bahnaric. Journal of the Southeast Asian Linguistics Society , –. Van der Auwera, Johan, Ludovic de Cuypere, and Annemie Neuckermans. . Negative Indefinites: A Typological and Diachronic Perspective on a Brabantic Construction. In Types of Variation: Diachronic, Dialectal and Typological Interfaces, ed. Terttu Nevalainen, Juhani Klemola, and Mikko Laitinen, –. Amsterdam: John Benjamins. Van der Auwera, Johan, Olga Krasnoukhova, and Frens Vossen. . Interwining the Negative Cycles. Presented at The Stockholm Workshop on the Negative Existential Cycle. May –, , Department of Linguistics, Stockholm University. Van der Auwera, Johan, Frens Vossen, and Maud Devos. . Le cycle de Jespersen à trois ou quatre negations. Actes du Colloque International ‘A Contrario’, ed. Jacques François, Isabelle Haïk, Dominique Legallois, Michel Morel, and Franck Neveu. Paris: Honoré Champion. Van der Wouden, Ton. . Polarity and ‘Illogical Negation’. In Dymanics, Polarity, and Quantification, ed. M. Kanazawa and C. J. Pinon, –. Stanford, CA: CSLI. Van der Wouden, Ton.  []. Negative Contexts Collocation, Polarity and Multiple Negation. Doctoral disseration, University of Groningen. (Published by Routledge). Van der Wouden, Ton and Frans Zwarts. . A Semantic Analysis of Negative Concord. In SALT III: Proceedings of the Third Conference on Semantics and Linguistic Theory, ed. U. Lahiri and A. Wyner, –. Ithaca, NY: Cornell University Department of Modern Languages and Linguistics. Van Emde Boas, Evert. . Ποίον τον μύθον έειπες; Rhetorical Questions in Ancient Greek. Institute for Logic, Language and Computation, Universiteit van Amsterdam. [Online.] Available at http://www.illc.uva.nl/Publications/ResearchReports/PP--.text.pdf. Van Gelderen, Elly. . Phrases, Heads, Grammaticalization and Economy. International Conference on Historical Linguistics, published in Journal of Comparative Germanic Linguistics . (), –. Van Gelderen, Elly. . Grammaticalization as Economy. Amsterdam–Philadelphia, PA: John Benjamins. Van Gelderen, Elly. . A History of the English Language. Amsterdam–Philadelphia, PA: John Benjamins. Van Gelderen, Elly. a. The Negative Cycle. Linguistic Typology (), –. Van Gelderen, Elly. b. Cycles of Negation in Athabaskan. Working Papers in Athabaskan Languages , –. Van Gelderen, Elly. a. Feature Economy in the Linguistic Cycle. In Historical Syntax and Linguistic Theory, ed. Paola Crisma and Giuseppe. Longobardi, . Oxford–New York: Oxford University Press. Van Gelderen, Elly. b. Cyclical Change, an Introduction. In Cyclical Change, ed. Elly van Gelderen, –. Amsterdam–Philadelphia, PA: John Benjamins. Van Gelderen, Elly. . The Linguistic Cycle: Language Change and the Language Faculty. Oxford: Oxford University Press.



References

Van Helten, W. . Vondel’s taal. Grammatica van het Nederlandsch der Zeventiende Eeuw. Groningen: J. B. Wolters. Van Kemenade, Ans. . Jespersen’s Cycle Revisited: Formal Properties of Grammaticalization. In Diachronic Syntax: Models and Mechanisms, ed. Susan Pintzuk, George Tsoulas, and Anthony Warner, –. Oxford: Oxford University Press. Veloudis, Jannis. . Negation in Modern Greek. PhD dissertation, University of Reading. Veloudis, Jannis. . I arnisi [Negation]. Athens: Patakis. Veloudis, Jannis. . Από τη σημασιολογία της Ελληνικής Γλώσσας: όψεις της επιστημικής τροπικότητας [Concerning the Semantics of Greek Language: Aspects of Epistemic Modality]. Thessaloniki: Institute of Modern Greek Studies (Manolis Triantafyllidis’ Foundation). Veloudis, Jannis. . Ταπεινή Γλώσσα, Τυπική Λογική και Ανθρώπινο βίωμα [Humble Language, Formal Logic and Human Experience]. Athens: Nissos. Veloudis, Jannis. . How Languages Acquire New Grammatical Elements: The Story of ka(n)Series in Modern Greek Retold. Linguistics (), –. Vendler, Zeno. . Verbs and Times. The Philosophical Review , –. Vendler, Zeno. . Linguistics and Philosophy. Ithaca, NY: Cornell University Press. Vennemann, Theo. . Topics, Subjects, and Word Order. From SXV to SVX via TVX. In Historical Linguistics, ed. John M. Anderson and C. Jones, –. Oxford: North-Holland Publishing. Vennemann, Theo. . Verb-Second, Verb-Late and the Brace Construction. In Historical Syntax, ed. Jacek Fisiak, –. Berlin: Mouton de Gruyter. Verkuyl, Henk J. . A Theory of Aspectuality: The Interaction between Temporal and Atemporal Structure. Cambridge: Cambridge University Press. Veselinova, Ljuba. . Negative Existentials: A Cross-Linguistic Study. Italian Journal of Linguistics , –. Veselinova, Ljuba. . The Negative Existential Cycle Revisited. Linguistics , –. Veselinova, Ljuba. . The Negative Existential Cycle Viewed through the Lens of Comparative Data. In Cyclical Change Continued, ed. Elly van Gelderen, –. Amsterdam: John Benjamins. Vessela, Carlo. . La prosa. In Storia delle lingue letterarie greche, ed. A. C. Cassio, –. Florence: Mondadori Education. Viger, François. . De praecipuis graecae dictionis idiotismis liber. Cum animadversionibus Henrici Hoogeveeni, Ioannis Caroli Zeunii et Godofredi Hermanni. Leipzig: Hahn. Viteau, . Étude sur le Grec du Nouveau Testament. Voelz, James W. . The Greek of the New Testament: its Place within the Context of Hellenistic Greek. In Greek: A Language in Evolution, Essays in Honour of Antonios N. Jannaris, ed. Chrys C. Caragounis, –. Hildesheim–Zürich: Georg Olms Verlag. Von Fintel, Kai. . Exceptive Conditionals: The Meaning of Unless. NELS , –. Von Stechow, Arnim. . Comparing Semantic Theories of Comparison. Journal of Semantics , –. von Stechow, Arnim. . Current issues in the theory of focus. In Semantik. Ein internationales Handbuch der zeitgenössischen Forschung ed. A. von Stechow and D. Wunderlich, –. Berlin: de Gruyter. Vossen, Frens. . On the typology of the Jespersen Cycles. Doctoral dissertation, University of Antwerp.

References



Vossen, Frens and Johan van der Auwera. . Multiple Exponence of Standard Negation in Austronesian. th International Conference of Austronesian Linguistics, Bali, July  . Wackernagel, Jacob.  []. Über ein Gesetz der indogermanischen Wortstellung, Indogermanische Forschungen , –. Wackernagel, Jacob. . Vorlesungen über Syntax mit besonderer Berücksichtigung von Griechisch, Lateinisch und Deutsch. Zweite Reihe. Basel: Birkhäuser Verlag. Wallage, Phillip. . Negation in Early English: parametrization and grammatical competition. PhD dissertation, University of York. Wallage, Phillip. . Jespersen’s Cycle in Middle English: Parametric Variation and Grammatical Competition. Lingua , –. Wallage, Phillip W. . Negation in Early English: Grammatical and Functional Change. Cambridge: Cambridge University Press. Waltereit, Richard and Ulrich Detges. . Syntactic Change: From Within and From Without Syntax. In The Paradox of Grammatical Change, ed. Ulrich Detges and Richard Waltereit, –. Amsterdam–Philadelphia, PA: John Benjamins. Weil, Henri. . De l’ordre des mots dans les langues anciennes compares aux langues moderns. Paris: Vieweg. Weiß, Helmut. . Duplex Negation Non Semper Affirmat: A Theory of Double Negation in Bavarian. Linguistics (), –. Westphal, Ernst Oswald Johannes. . An Introductory Comparative Study in Bantu. Mitteilungen des Instituts für Orientforschung. , –. Whitney, William D.  []. Sanskrit Grammar. Cambridge, MA: Harvard University Press. Wildgen, Wolfgang. . Evolutionary Pragmatics. In Handbook of Pragmatics, ed. Jan-Ola Östman and Jef Verschueren, .–. Amsterdam: John Benjamins. Wilkins, Charles. . A Grammar of the Sanskrita Language. London: Bulmer and Co. Wilkinson, Karina. . The scope of even. Natural Language Semantics , –. Willis, David. . A minimalist approach to Jespersen’s Cycle in Welsh. Ms, University of Cambridge. Willis, David, Christopher Lucas, and Anne Breitbarth. . Comparing Diachronies of Negation. In The History of Negation in the Languages of Europe and the Mediterranean. Volume I: Case Studies, ed. David Willis, Christopher Lucas, and Anne Breitbarth, –. Cambridge: Cambridge University Press. Willmott, Jo. . The Moods of Homeric Greek. Cambridge–New York: Cambridge University Press. Willmott, Jo. . Not in the Mood: Modality and Negation in the History of Greek. In Proceedings from the th Annual Meeting of the Department of Linguistics, Aristotle University of Thessaloniki. Willmott, Jo. . Review of Markopoulos The Future in Greek: From Ancient to Medieval. Journal of Hellenic Studies , . Willmott, Jo. . Negation in the History of Greek. In The History of Negation in the Languages of Europe and the Mediterranean. Volume I: Case Studies, ed. David Willis, Christopher Lucas, and Anne Breitbarth, –. Cambridge: Cambridge University Press. Winer, Georg B. . Grammatik des neutestamentlichen Sprachidioms als sichere Gundlage der neutestamentlichen Exegese, th edition. Leipzig: Vogel.



References

Winer, Georg B. . A Treatise on the Grammar of New Testament Greek, trans. W. F. Moulton. Edinburgh: T. & T. Clark. Wisse, Jakob. . Greeks, Romans, and the Rise of Atticism. In Greek Literature in the Roman Period and in Late Antiquity (Greek Literature vol. ), ed. Nagy, Gregory, –. London: Routledge. Wood, Johanna. . Definiteness and number. PhD dissertation, Arizona State University. Wurmbrand, Susanne. . Infinitives: Restructuring the Clause Structure. Studies in Generative Grammar . Berlin: Mouton de Gruyter. Xrakovskij, Victor S. (ed.) . Typology of Imperative Constructions. Münich: Lincom Europa. Xue-lian, Yang. . Again on Semantic Expressions of Intensified Negation in Written English. Ms. Yoon, Suwon. . NOT in the Mood: the syntax and semantics of evaluative negation. PhD dissertation, University of Chicago. Yoshimura, Keiko. . Focus and Polarity in Japanese. PhD dissertation, University of Chicago, Chicago, IL. Young, Richard A. . Intermediate New Testament Greek: A Linguistic and Exegetical Approach. Nashville, TN: Broadman and Holman Publishers. Zanuttini, Raffaella. . Syntactic Properties of Sentential Negation: a comparative study of romance languages. PhD dissertation, University of Pennsylvania. Zanuttini, Raffaella. a. Speculations on negative imperatives. Rivista di Linguistica , –. Zanuttini, Raffaella. b. Re-examining Negative Clauses. In Paths Towards Universal Grammar: Studies in Honor of Richard S. Kayne, ed. Guglielmo Cinque, Jan Koster, JeanYves Pollock, Luigi Rizzi, and Raffaella Zanuttini, –. Washington DC: Georgetown University Press. Zanuttini, Raffaella. . Negation and Verb Movement. In The New Comparative Syntax, ed. Liliane Haegeman, –. Harlow: Longman. Zanuttini, Raffaella.  []. Negation and Clausal Structure. Oxford: Oxford University Press. Zeijlstra, Hedde H. . What the Dutch Jespersen Cycle may Reveal about Negative Concord. Linguistics in Potsdam, ed. A. Alexiadou, S. Fischer, and M. Stavrou, vol. , –. Potsdam: University of Potsdam. Zeijlstra, Hedde H. . Sentential Negation and Negative Concord. PhD dissertation, Universiteit van Amsterdam. Zeijlstra, Hedde H. . The Ban on True Negative Imperatives. In Empirical Issues in Syntax and Semantics , ed O. Bonami and P. Cabredo Hofherr, –. [Online.] Available at http:// www.cssp.cnrs.fr/eiss/eiss.pdf, or https://halshs.archives-ouvertes.fr/halshs-. Zeijlstra, Hedde H. . Emphatic multiple negative expressions in Dutch. Amsterdam: ACLC working papers and lingBuzz/. Zeijlstra, Hedde H. . On the Syntactic Flexibility of Formal Features. In The Limits of Syntactic Variation, ed. Theresa Biberauer, –. Amsterdam: John Benjamins. Zeijlstra, Hedde H. . Not in the First Place. Natural Language & Linguistic Theory , –.

References



Zeijlstra, Hedde H. . Diachronic Developments in the Domain of Negation. Language & Linguistics Compass , –. Zerwick, Maximilian. . Biblical Greek. Rome: Pontifical Biblical Institute. Zwarts, Frans. . Categoriale Grammatica en Algebraische Semantiek. Een studie naar negatie en polariteit in het Nederlands. Doctoral dissertation, University of Groningen. Zwarts, Frans. . Three types of polarity. Ms. University of Groningen. Also appeared as Zwarts  in Plurality and Quantification, ed. F. Hamn and E. Hinrichs, –. Dordrecht: Kluwer. Zwarts, Frans. . Nonveridical Contexts. Linguistic Analysis , –.

Index accusativus cum infinitivo – agreement morphological agreement ,  syntactic agreement –, –, , , ,  Albanian –, –, , n. Algonquian ,  Analyticity  Armenian –, , ,  Attic Greek –, –, , , , , –, –, , , , , –, –, –, –, –, –, –, –, , , –, , , , , –, , –, , –, , , , , , – Bukiyip  conditionals , , –, , –, , , , –, , , –, –, –, –, –, , , , –, , , , –, , , , , ,  consecutio modorum –, –,  Doric , – Early Medieval Greek –, , , , , , –, , , , –,  Ellipsis , ,  factive –, , –, , –, –, –, ,  grammaticalization , –, , , , , –, , , –, , , , –, , , , ,  Hittite –, –,  Homeric Greek , , –, , , –, , , , , n., –, , , ,  homotics – imperatives , , , –, , –, , , –, –, , –, , , , , –, , , , , –, 

true negative imperatives , , –, , , , –, ,  indicative , , , –, –, , , , –, –, , –, –, , –, , , –, ,  intensional , , ,  Ionic ,  irrealis , , , , , , , , , , – irrealis/final infinitives , , ,  Italian , , , , – Dego Italian ,  Modern Italian  Old Italian  Koine Greek –, , , , , , –, , –, –, –, n., –, –, –, , –, –, , –, , , –, , –,  Late Medieval Greek , , , –, –, , , , , , –, –, –, –, , , –, , – Latin –, , , , –, ,  Lewo (Vanuatu)  Luwian  Macedonian ,  marked–unmarked, markedness –, –, , , , , , , , , –, , , , –, , , , ,  unmarked word order , ,  Modern Greek , , –, –, –, –, –, , , , , –, , n., –, , –, , n., –, , –, –, –, –, –, n., , –, –, , –, –, , , –, , , –; Early Modern Greek , , –, , –, ,  nonveridicality , –, , –, , –, –, , , –, –, , , , , , , , –, , , –, –, , , , , , , , , , , , , 



Index

nonveridicality (cont.) nonveridicality projection –, , , , , , , –, , , –, ,  Old Avestan –,  Old Persian –, ,  optative , , , , , –, –, , , –, ,  parameter , , , , ,  parameter resetting , , ,  perfective nonpast , –, –, , –, –, –, , –,  prohibition , , , –, –, –, , , , , , , – Proto-Indo-European , , , , –, , n., ,  questions –, , , , , , –, , , , , , , , , –, , , , , –, –, , 

question particle , –, , , –, , , –, , –, , –, , , – rhetorical questions , , n., , ,  reanalysis –, , , , –, –, –, , , – upward reanalysis , , , , , –, , , , , – Romeyka , – Sanskrit –, –,  subjunctive , , , –, –, , , –, , –, –, , , –, –, –, –, , , ,  volitional, volitionality , , , , –, , , , , –, , , , ,  Zulu , 

OXFORD STUDIES IN DIACHRONIC AND HISTORICAL LINGUISTICS   Adam Ledgeway and Ian Roberts, University of Cambridge   Cynthia Allen, Australian National University; Ricardo Bermúdez-Otero, University of Manchester; Theresa Biberauer, University of Cambridge; Charlotte Galves, University of Campinas; Geoff Horrocks, University of Cambridge; Paul Kiparsky, Stanford University; Anthony Kroch, University of Pennsylvania; David Lightfoot, Georgetown University; Giuseppe Longobardi, University of York; George Walkden, University of Konstanz; David Willis, University of Cambridge   From Latin to Romance Morphosyntactic Typology and Change Adam Ledgeway  Parameter Theory and Linguistic Change Edited by Charlotte Galves, Sonia Cyrino, Ruth Lopes, Filomena Sandalo, and Juanito Avelar  Case in Semitic Roles, Relations, and Reconstruction Rebecca Hasselbach  The Boundaries of Pure Morphology Diachronic and Synchronic Perspectives Edited by Silvio Cruschina, Martin Maiden, and John Charles Smith  The History of Negation in the Languages of Europe and the Mediterranean Volume I: Case Studies Edited by David Willis, Christopher Lucas, and Anne Breitbarth  Constructionalization and Constructional Changes Elizabeth Traugott and Graeme Trousdale  Word Order in Old Italian Cecilia Poletto  Diachrony and Dialects Grammatical Change in the Dialects of Italy Edited by Paola Benincà, Adam Ledgeway, and Nigel Vincent  Discourse and Pragmatic Markers from Latin to the Romance Languages Edited by Chiara Ghezzi and Piera Molinelli

 Vowel Length from Latin to Romance Michele Loporcaro  The Evolution of Functional Left Peripheries in Hungarian Syntax Edited by Katalin É. Kiss  Syntactic Reconstruction and Proto-Germanic George Walkden  The History of Low German Negation Anne Breitbarth  Arabic Indefinites, Interrogatives, and Negators A Linguistic History of Western Dialects David Wilmsen  Syntax over Time Lexical, Morphological, and Information-Structural Interactions Edited by Theresa Biberauer and George Walkden  Syllable and Segment in Latin Ranjan Sen  Participles in Rigvedic Sanskrit The Syntax and Semantics of Adjectival Verb Forms John J. Lowe  Verb Movement and Clause Structure in Old Romanian Virginia Hill and Gabriela Alboiu  The Syntax of Old Romanian Edited by Gabriela Pană Dindelegan  Grammaticalization and the Rise of Configurationality in Indo-Aryan Uta Reinöhl  The Rise and Fall of Ergativity in Aramaic Cycles of Alignment Change Eleanor Coghill  Portuguese Relative Clauses in Synchrony and Diachrony Adriana Cardoso

 Micro-change and Macro-change in Diachronic Syntax Edited by Éric Mathieu and Robert Truswell  The Development of Latin Clause Structure A Study of the Extended Verb Phrase Lieven Danckaert  Transitive Nouns and Adjectives Evidence from Early Indo-Aryan John J. Lowe  Quantitative Historical Linguistics A Corpus Framework Gard B. Jenset and Barbara McGillivray  Gender from Latin to Romance History, Geography, Typology Michele Loporcaro  Clause Structure and Word Order in the History of German Edited by Agnes Jäger, Gisella Ferraresi, and Helmut Weiß  Word Order Change Edited by Ana Maria Martins and Adriana Cardoso  Arabic Historical Dialectology Linguistic and Sociolinguistic Approaches Edited by Clive Holes  Grammaticalization from a Typological Perspective Edited by Heiko Narrog and Bernd Heine  Negation and Nonveridicality in the History of Greek Katerina Chatzopoulou I  Cycles in Language Change Edited by Anne Breitbarth, Elisabeth Witzenhausen, Miriam Bouzouita, and Lieven Danckaert Morphological Borrowing Francesco Gardani Indefinites between Latin and Romance Chiara Gianollo Nominal Expressions and Language Change From Early Latin to Modern Romance Giuliana Giusti

Syntactic Features and the Limits of Syntactic Change Edited by Jóhannes Gísli Jónsson and Thórhallur Eythórsson A Study in Grammatical Change The Modern Greek Weak Subject Pronoun τος and its Implications for Language Change and Structure Brian D. Joseph Reconstructing Pre-Islamic Arabic Dialects Alexander Magidow Word Order and Parameter Change in Romanian Alexandru Nicolae Referential Null Subjects in Early English Kristian A. Rusten The History of Negation in the Languages of Europe and the Mediterranean Volume II: Patterns and Processes Edited by David Willis, Christopher Lucas, and Anne Breitbarth Verb Second in Medieval Romance Sam Wolfe Palatal Sound Change in the Romance Languages Diachronic and Synchronic Perspectives André Zampaulo