The Linguistic Cycle: Economy and Renewal in Historical Linguistics [1 ed.] 9781003272564, 9781032224336, 9781032224329

Table of contents :
Cover
Half Title
Title
Copyright
Contents
Preface
List of tables
List of figures
List of Abbreviations and primary sources
1 Introduction
1 What is the linguistic cycle?
2 What kinds of cycles exist?
3 How and why to study cyclical change
3.1 The practical side
3.2 The theoretical side
4 Major questions in the study of cycles
5 Terminology, resources, and glossing
6 Conclusions and outline
Suggestions for further reading
Review questions and exercises
2 A history of cyclical change
1 The eighteenth and early nineteenth centuries
2 The late nineteenth and early twentieth century
3 The mid and late twentieth century
4 Recent generative work
5 Recent non-generative work
6 Conclusions
Suggestions for further reading
Review questions
3 Micro cycles: determiner and verbal cycles
1 Definition of a micro cycle
2 The determiner cycle
3 Copula cycles
4 Tense and aspect cycles
4.1 The imperfective cycle
4.2 The perfective cycle
4.3 Imperfective and perfective renewal in Basque
5 Mood and modal cycles
6 Voice cycles
7 Conclusions
Suggestions for further reading
Review questions and exercises
4 Micro cycles: polarity and discourse cycles
1 Negative cycles
1.1 Jespersen’s negative cycle
1.2 Givón’s negative cycle
1.3 Croft’s negative cycle, also known as the negative existential cycle
2 The interrogative cycle
3 Complementizer cycles
4 Pragmatic cycles
4.1 A definition
4.2 Temporal sources
4.3 Emphatic pronoun cycles
5 Interactions between micro cycles
6 Conclusions
Suggestions for further reading
Review questions and exercises
5 Macro cycles
1 Definition and controversies
2 Analytic and synthetic cycles
3 Agreement cycles or head marking cycles
3.1 Subject agreement cycles
3.2 Object cycles
3.3 Morpheme order
4 Case cycles or dependent marking cycles
5 Interactions involving macro cycles
6 Conclusions
Suggestions for further reading
Review questions and exercises
Appendix
6 Explanations and mechanisms
1 Comfort and clarity
2 External influence
3 Construction grammar
4 Early Minimalism: structural and featural economy
5 Later Minimalism: minimal search
6 Attractor states and linguistic cycles
7 Conclusions
Suggestions for further reading
Review questions and exercises
7 Conclusions and future directions
1 Insights from cycles
2 Criticisms of the cycle
3 Future directions
4 Conclusions
Review questions and exercises
Final advanced exercise
Possible answers to the review questions and exercises
References
Index

Citation preview

THE LINGUISTIC CYCLE

Cyclical language change is a linguistic process by which a word, phrase, or part of the grammar loses its meaning or function and is then replaced by another. This can even happen on the level of an entire language, which can experience a change in the language family it is a part of. This new text is a comprehensive introduction to this phenomenon, the mechanisms underlying it, and the relations between the different types of cycles. Elly van Gelderen reviews the subject widely and holistically, defining key terms and comprehensively presenting diverse theoretical perspectives and empirical findings. With coverage of a variety of micro cycles and the more controversial macro cycles, incorporating cutting-edge work on grammaticalization, and drawing on examples from many languages and language families, this book accessibly guides readers through the state of the art in the field. With practical methodological guidance on how to identify and investigate linguistic cycles, and an array of useful pedagogical features, the book provides a coherent framework for approaching, understanding, and furthering research in linguistic cycles. This text will be an indispensable resource for advanced students and researchers in historical and diachronic linguistics, language typology, and linguistic and grammatical theory. Elly van Gelderen is Regents’ Professor of English and Linguistics at Arizona State University, USA. Her most recent books include The Linguistic Cycle: Language Change and the Language Faculty (2011), The Diachrony of Meaning (2018), and Third Factors in Language Variation and Change (2022).

THE LINGUISTIC CYCLE Economy and Renewal in Historical Linguistics

Elly van Gelderen

Cover image: Photo of M31 courtesy of Randy Pfeiffer First published 2024 by Routledge 605 Third Avenue, New York, NY 10158 and by Routledge 4 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN Routledge is an imprint of the Taylor & Francis Group, an informa business © 2024 Elly van Gelderen The right of Elly van Gelderen to be identified as author of this work has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data Names: Gelderen, Elly van, author. Title: The linguistic cycle : economy and renewal in historical linguistics / Elly van Gelderen. Description: New York, NY : Routledge, 2023. | Includes bibliographical references and index. Identifiers: LCCN 2023003819 (print) | LCCN 2023003820 (ebook) | ISBN 9781032224336 (hardback) | ISBN 9781032224329 (paperback) | ISBN 9781003272564 (ebook) Subjects: LCSH: Linguistic change. | Grammar, Comparative and general—Grammaticalization. | Cycles. Classification: LCC P142 .G449 2023 (print) | LCC P142 (ebook) | DDC 417/.7—dc23/eng/20230501 LC record available at https://lccn.loc.gov/2023003819 LC ebook record available at https://lccn.loc.gov/2023003820 ISBN: 978-1-032-22433-6 (hbk) ISBN: 978-1-032-22432-9 (pbk) ISBN: 978-1-003-27256-4 (ebk) DOI: 10.4324/9781003272564 Typeset in Times New Roman by Apex CoVantage, LLC

CONTENTS

Preface List of tables List of figures List of Abbreviations and primary sources 1

Introduction

viii x xii xiii 1

1 What is the linguistic cycle? 3 2 What kinds of cycles exist? 8 3 How and why to study cyclical change 11 3.1 The practical side 11 3.2 The theoretical side 14 4 Major questions in the study of cycles 17 5 Terminology, resources, and glossing 20 6 Conclusions and outline 23 Suggestions for further reading 24 Review questions and exercises 24 2

A history of cyclical change 1 2 3 4 5

The eighteenth and early nineteenth centuries 27 The late nineteenth and early twentieth century 30 The mid and late twentieth century 34 Recent generative work 36 Recent non-generative work 38

26

vi

Contents

6 Conclusions 40 Suggestions for further reading 40 Review questions 40 3

Micro cycles: determiner and verbal cycles

41

1 2 3 4

Definition of a micro cycle 42 The determiner cycle 43 Copula cycles 53 Tense and aspect cycles 58 4.1 The imperfective cycle 59 4.2 The perfective cycle 68 4.3 Imperfective and perfective renewal in Basque 72 5 Mood and modal cycles 73 6 Voice cycles 82 7 Conclusions 90 Suggestions for further reading 91 Review questions and exercises 91

4

Micro cycles: polarity and discourse cycles

93

1 Negative cycles 94 1.1 Jespersen’s negative cycle 94 1.2 Givón’s negative cycle 97 1.3 Croft’s negative cycle, also known as the negative existential cycle 100 2 The interrogative cycle 106 3 Complementizer cycles 111 4 Pragmatic cycles 116 4.1 A definition 117 4.2 Temporal sources 119 4.3 Emphatic pronoun cycles 121 5 Interactions between micro cycles 122 6 Conclusions 124 Suggestions for further reading 125 Review questions and exercises 125 5

Macro cycles 1 Definition and controversies 129 2 Analytic and synthetic cycles 131 3 Agreement cycles or head marking cycles 140

128

Contents

vii

3.1 Subject agreement cycles 140 3.2 Object cycles 148 3.3 Morpheme order 154 4 Case cycles or dependent marking cycles 156 5 Interactions involving macro cycles 162 6 Conclusions 167 Suggestions for further reading 167 Review questions and exercises 168 Appendix 170 6

Explanations and mechanisms

172

1 2 3 4

Comfort and clarity 173 External influence 177 Construction grammar 183 Early Minimalism: structural and featural economy 187 5 Later Minimalism: minimal search 197 6 Attractor states and linguistic cycles 200 7 Conclusions 203 Suggestions for further reading 203 Review questions and exercises 203 7

Conclusions and future directions

204

1 Insights from cycles 204 2 Criticisms of the cycle 206 3 Future directions 208 4 Conclusions 210 Review questions and exercises 210 Final advanced exercise 210 Possible answers to the review questions and exercises References Index

211 217 247

PREFACE

Linguistic cycles describe phenomena that recur in unidirectional, systematic patterns. They provide insight into the faculty of language by (a) illustrating the economizing principles in a syntactic derivation and (b) displaying the array of universal semantic features. Cycles also give typological insight on the possible shapes of languages. Linguistic cycles involve loss and renewal. Analogies have been made between linguistic cycles and geological processes, for example, as a language grinding itself down and eroding (Naumann 1992); or chemical ones, for example, as bleaching (Sweetser 1988); or biological, for example, as natural selection (Benítez-Burraco 2017) and hermit crabs (Heath 1998). Those characterizations suggest that language is a system on its own, independent of the role of the speaker. In this book, even though I may use terms like language loss and renewal, I assume that it is the speaker of a language who erodes and renews. The book is written for an audience similar to that of other specialized textbooks, such as, Hopper and Traugott’s (2003) Grammaticalization, Siewierska’s (2004) Person, and Narrog and Heine’s (2021) Grammaticalization. I have taken these as guides. Readers are assumed to have a basic familiarity with linguistic concepts, such as morpheme, analytic, synthetic, tense, mood, aspect, definiteness, argument, and have a basic knowledge of morphology and syntax. An introduction to linguistics (using e.g. Fromkin et al. 2019 or O’Grady et al. 2017 as textbook) or to typology (using e.g. Whaley 1997 or Payne 1997 [2010]) should be sufficient preparation. Good glossaries are available at https://glossary. sil.org/term, www.uni-due.de/ELE/LinguisticGlossary.html, and http://cjpountain.sllf.qmul.ac.uk/linggloss.htm. I have tried to keep Chapters 1 through 5 free of technical explanations. Sections 3, 4, and 5 of Chapter 6 are an exception, and in a more descriptively oriented course, these can be skipped.

Preface

ix

Reviewers have suggested adding phonological cycles, such as the first consonant shift, since these are perfect examples of loss and renewal. I hope someone will work on these cycles, but I don’t have the expertise. Bermúdez-Otero and Trousdale (2012) may be a place to start. Inevitably, many of the ideas and examples are based on earlier work on the linguistic cycle authored by me, for example, van Gelderen 2007a, b, 2008a, b, 2009a, b, 2011, 2013, 2016. Sources are grammars, corpora, and linguistic articles. Where the source is not listed, they are from personal knowledge. Translations of quotes are generally my own as well. Examples from the history of English are used more than any other language; as one reviewer wrote, this is “didactically acceptable.” As for the stylistic choices, textbooks are used in classrooms where the book, instructor, and students all participate. Therefore, I use ‘we’ when that’s the case. When ‘I’ is used, it is in contexts where I argue something specific. With apologies to a reviewer who doesn’t really like my “tendency to write almost everything in present tense,” I continue to write in the present tense when talking about historical stages to emphasize their current relevance and to reliven the prose. I have benefitted very much from the comments of 22 reviewers whose help Ze’ev Sudry requested and also thank him for his initial query to write this book before he left Routledge. Thanks very much to the members of the ASU Syntax Reading Group, in particular Johanna Wood, Derek McCarthy, Ming Chen, Annette Hornung, Kaya Fallenstein, Jacob Willson, and Servo Kocu. I also valued being a guest in Maj-Britt Mosegaard Hansen’s reading group on Pragmatic Cycles and am very appreciative to Maj-Britt and Bernd Heine for comments. Special thanks to John Powell, who has enthusiastically helped clarify the prose and has thought about the audience and the exercises. Recent PhD students asked great questions: Angela, Annette, Cristian, Haroon, Jaber, Mastourah, Mekhlid, Mansour, Muneer, Robert, Umar, and William, among many others.

TABLES

1.1 1.2 1.3 3.1 3.2 3.3 3.4 4.1 4.2 4.3 4.4 5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9 5.10 5.11 5.12 5.13

Cyclical developments in Egyptian Examples of cyclical change Pronouns in four Athabaskan languages Definite articles cross-linguistically in WALS Varied positions for renewal of deictic particles in Germanic Forms in -ende in the Peterborough Chronicle Changes in the have perfect in Old and Middle English The stages of the NEC from Sanskrit to Hindi/Urdu Numbers and percentages of demonstrative objects (Dem) with after and fronting Reanalyses involving after Changes in the history of Dutch Greenberg’s index of syntheticity Locus of marking Whole-language typology Distribution of pronominal subjects Frequency of verbal agreement Pronouns in Old French Loss of verbal agreement from Old to Modern French Pronouns and OMs in Tukang Besi Pronouns in Marshallese Pronouns in Proto Micronesian The order of V and subject agreement affix The order of V and object agreement affix Dependent marking

4 9 12 44 51 65 70 101 113 116 126 133 133 133 141 141 143 143 151 151 152 154 155 156

Tables

5.14 Cases for Sanskrit -a stems for deva ‘god’ and the demonstrative pronoun 5.15 Portmanteau morphemes with first and second person subjects and other objects 5.16 Some pre- and postverbal objects in the Orléans corpus 5.17 Pronouns and subject agreement in Arabic 5.18 Suffixing and prefixing 6.1 Changes in the Peterborough Chronicle, especially after 1130 6.2 Features of house and build 6.3 What sets the cycles in motion, using generative grammar 7.1 The linguistic cycles in this book 7.2 Other cycles, not discussed or not in detail

xi

160 164 167 169 169 180 193 199 205 209

FIGURES

1.1 1.2 1.3 2.1 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 4.1 4.2 5.1 5.2 5.3 6.1 6.2

Start followed by an -ing verb The use of never in COHA Begin followed by an -ing verb Model of language acquisition The article ‘the’ followed by a noun The demonstrative ‘that’ followed by a noun Tense, aspect, and mood grammaticalization Classification of aspectual oppositions The imperfective cycle Verb auxiliary continuum Sources of passives That there + noun That + noun + there Failed to and a verb in COCA The negative existential cycle (NEC) Synthetic analytic continuum Attachment type The first person singular pronoun separated from and adjacent to the auxiliary/verb was Changing complements to dislike The negative cycle

25 212 212 37 49 49 58 59 63 76 83 213 214 99 100 132 135 215 184 195

ABBREVIATIONS AND PRIMARY SOURCES

ABL ABS ACC AFOC ANTIP ART ASP AUX C c Chron CL CM COND CONNEG COP CP DAT DEC DEF DEM DET EMPH ERG EX

ablative absolutive accusative assertive focus antipassive article aspect auxiliary complementizer circa Chronicle classifier common gender conditional negative nonfinite copula Complementizer Phrase dative declarative definite demonstrative determiner emphatic ergative existential

xiv

Abbreviations and primary sources

EXCL F FUT GEN HPP INC INF INST IPF IRR LA LMP LOC M MID N NP NEC NEG NOM OE OB OM OTA P PASS PF Pol PP PR PROG PRT PST Q QUA REFL REL S SAE SAI SU SUBJ

exclusive number feminine future genitive Head Preference Principle inclusive number infinitive instrumental imperfective irrealis Labeling Algorithm Late Merge Principle locative masculine middle voice noun noun phrase Negative Existential Cycle negative nominative on a noun; nominalizer on a verb Old English object object marker Oxford Text Archive preposition, or plural in glosses passive voice perfect polarity preposition phrase present progressive particle or participle past question qualifier reflexive relative singular Standard Average European Subject Auxiliary Inversion subject subjunctive mood

Abbreviations and primary sources

T TMA TNS TP V v VOC WALS * = – 1 2 3

xv

tense (and familiar in the case of pronouns) tense, mood, and aspect tense tense phrase verb or polite (with pronouns) light verb vocative World Atlas of Linguistic Structures reconstructed or ungrammatical; actual use will be clarified clitic affix first person second person third person

Primary sources

BNC CdES COCA COHA DOE Gutenberg MED OED Orléans Corpus

British National Corpus (www.english-corpora.org/bnc) Corpus d’entretiens spontanés (www.uwe.ac.uk/hlss/llas/iclru/ corpus.pdf) Corpus of Contemporary American English (www.englishcorpora.org/coca) Corpus of Historical American English (www.english-corpora. org/coha) Dictionary of Old English (https://tapor.library.utoronto.ca/ doecorpus/) See www.gutenberg.org for references to Jane Austin, Mark Twain, etc. Middle English Dictionary (https://quod.lib.umich.edu/m/ middle-english-dictionary/dictionary) Oxford English Dictionary Part of the ELICOP Corpus (http://bach.arts.kuleuven.be/ pmertens/corpus/search/t.html

1 INTRODUCTION

One of the meanings of ‘cycle’ in the Oxford English Dictionary (OED) is “a regular order or succession in which things recur.” The term was either borrowed from French cycle or from Latin cyclus – the exact origin of technical terms is sometimes not clear – and is first encountered in 1387 with a chronological meaning. It has been used in astronomy, physics, geology, medicine, botany, zoology, and mathematics, to name a few. The OED doesn’t mention a linguistic use, but that use has been available since the nineteenth century. In linguistics, we take a cycle to describe a round of linguistic changes taking place in a systematic manner and direction. The term typically describes a sequence of linguistic changes in which a word grammaticalizes, that is, loses lexical meaning, gains more grammatical use, shortens in form, and is renewed by another lexical item. A word may “grind itself down,” in von der Gabelentz’ words of 1901, in order to be replaced by a new one to counteract “this erosion and the destruction of the language.” As far as I have been able to determine, the term ‘linguistic cycle’ was first used by Lefèvre (1877, 1893) to refer to the change from analytic to synthetic languages. The terms analytic and synthetic will be used frequently. These notions refer to a spectrum of languages, with the former expression characterizing those that possess relatively fewer affixes and the latter term describing those with relatively more affixes. The most well-known of cycles is the Negative Cycle, and it plays out in the history of English and many other languages. For instance, Old English has a negative ne that is initially renewed through not and then replaced by it. This development is shown in (1): ne is the original negative in Old English, weakens to a prefix, and gets reinforced by another negative originating from a phrase na wiht ‘no creature’ and changes its shape into not. Subsequently, the original DOI: 10.4324/9781003272564-1

2

Introduction

negative ne drops out, and its replacement, the negative not, loses semantic and phonological impact again when it forms a single word with auxiliaries, such as don’t, won’t, and can’t. (1)

The Negative Cycle in the history of English Old English Middle English later Middle English Modern English ne > n(e)- . . . na wiht/not > not > -n’t

Because lexical elements that renew older forms in cyclical change are not (always) identical to the old ones, one way of characterizing a cycle is as a spiral, as in von der Gabelentz (1901: 256) and in Meillet’s (1912: 140) description of it as “une sorte de développement en spirale” [‘a kind of development as a spiral’]. Jespersen (1917) discusses the tendency for negatives to weaken and then be subsequently strengthened, but the term ‘Jespersen Cycle’ isn’t used until Dahl (1979). In fact, aware of the breadth of the definition of other cycles, Jespersen very specifically rejects the concept of cyclical change in the broader sense because he sees languages as moving towards flexionless stages in a unidirectional manner and not developing new morphology (Jespersen 1922; Chapter 21.9). Incidentally, as van der Auwera (2009) points out, Gardiner (1904) had already given a description of a Negative Cycle, as did Meillet (1912), so the idea must have been known before 1917. However, because there are many Negative Cycles, I will on occasion use Jespersen Cycle to characterize the cycle in (1). After a hiatus of work on the linguistic cycle, Tauli (1958, 1966), Hodge (1970), and Givón (1971, 1976) return to this topic. Their work is diverse: Tauli and Hodge still focus on the major changes between analytic and synthetic stages, and whereas Hodge reintroduces the term ‘cycle,’ Tauli avoids this term, instead preferring ‘analytic and synthetic tendencies.’ Givón’s work emphasizes reconstructing earlier stages of Bantu languages and understanding change in terms of cyclical change. The 2000s experience an interest in cyclical change, especially by generative linguists. That framework regards unidirectional change as a window on the faculty of language. There is also an interest in Pragmatic Cycles by non-generativists, for example, Hansen and Visconti (2009). Although a majority of the effort in that time period focuses on negative cycles (Kiparsky & Condoravdi 2006; Larrivée & Ingham 2011; Vossen & van der Auwera 2014; Willis et al. 2013; Breitbarth et al. 2020), cycles involving determiners (Egedi 2014; Wood 2016), prepositions (Hegedüs 2014), complementizers (Jäger 2010; Bácskai-Atkári 2014a; Bácskai-Atkári & Dékány 2014), modals (Jędrzejowski 2016), and pronouns (van Gelderen 2011; Mithun 2013; Bahtchevanova & van Gelderen 2016; Givón 2016a) are also examined. In keeping with Hodge (1970), Givón (2018: 198), Bouzouita et al. (2020), and others, I will be using the term ‘cycle’ in this book rather than ‘spiral’ because the latter is “much less usual” (Hagège 1993: 149), and renewal doesn’t always occur through a different lexical item but also the reuse of the earlier form.

Introduction

3

The remainder of this introductory chapter outlines in more detail what a cycle is, what kinds of cycles are generally accepted, how and why to study cycles, and big picture questions regarding the linguistic cycle. 1

What is the linguistic cycle?

In this section, I’ll define the linguistic cycle, provide brief examples of cycles, and discuss the two mechanisms underlying cycles, grammaticalization and renewal. The linguistic cycle is a designation used to describe language change taking place in a systematic manner and direction. One definition I will employ in this book appears in (2). (2)

The Linguistic Cycle involves [initial formulation] a set of changes where lexical words or expressions weaken to grammatical morphemes or where grammatical items become more grammatical. The original words and expressions are renewed by lexical expressions that may be different or the same as the original.

A typical cycle thus comprises of grammaticalization followed by renewal, which is in turn followed by grammaticalization and so on. This definition is based on the oft-quoted characterization by von der Gabelentz, as provided in (3). (3)

The history of language moves in the diagonal of two forces: the impulse toward comfort, which leads to the wearing down of sounds, and that toward clarity, which disallows this erosion and the destruction of the language. The affixes grind themselves down, disappear without a trace; their functions or similar ones, however, require new expression. They acquire this expression, by the method of isolating languages, through word order or clarifying words. The latter, in the course of time, undergo agglutination, erosion, and in the mean time renewal is prepared: periphrastic expressions are preferred . . . always the same: the development curves back towards isolation, not in the old way, but in a parallel fashion. That’s why I compare them to spirals. (von der Gabelentz 1901: 256)

An isolating language is highly analytic and, as line 5 in the quote mentions, these languages undergo changes where independent words become affixes through agglutination, which itself wears off again. Chapter 5 will discuss the change of languages from isolating to agglutinative and back to isolating in more detail. This change can also be characterized as going from analytic to synthetic to analytic and so on. Hodge (1970) has done more than anyone to feed recent ideas on the cycle with his short article titled ‘The Linguistic Cycle’ in which he examines the

4

Introduction TABLE 1.1 Cyclical developments in Egyptian

Proto-Afroasiatic Old Egyptian Late Egyptian Coptic

analytic synthetic analytic synthetic

*Sm sM Sm sM

(from Hodge 1970: 5, * indicates a reconstructed stage)

overall changes in morphological structure throughout the history of Egyptian. He uses lower and uppercase to give a visual representation of full cycles from synthetic ‘sM,’ that is, a language with lots of inflectional morphology as indicated by the capital M and lower case s for less syntax, to analytic ‘Sm,’ that is, a language with a lot of syntax, indicated by the capital S, and less morphology, indicated by lowercase m. By more or less syntax Hodge means the degree of reliance on function words and word order; I return to the history of this language in Chapter 5. Hodge’s representation is provided in Table 1.1 Jespersen (1917) is among the first to discuss changes in negatives as a cycle, but without using that term. He cites examples from many languages and provides a discussion about weakening and strengthening tendencies. There is a wealth of data on Negative Cycles in a myriad of languages from other, more recent sources, for example, van der Auwera and Vossen (2016), Vossen and van der Auwera (2014), Mithun (2016), Larrivée and Ingham (2011), Willis et al. (2013: 7, 21, 169), Chatzopoulou (2019), Breitbarth et al. (2020), to mention but a few. A typical chain of changes from the history of English was given in (1) in the previous section; it is repeated here as (4). (4)

The Negative Cycle, starting on the left Old English Middle English later Middle English Modern English ne > n(e)- . . . na wiht/not > not > not/-n’t . . . nothing NEG NEG NP/ADVERB NEG NEG NP

Examples of the stages of the Negative Cycle are given in (5) for English. They include the following: (5a) shows the use of a negative ne by itself; (5b) shows a negatively marked argument nowiht together with ne; (5c) has a negative contracted with the verb into nes and a negative adverbial nawhit, a contracted form of the negative indefinite na wiht ‘no thing/creature’; (5d) shows the adverbial not by itself; and (5e) shows the cliticization of not onto the verb; and (5f) the reinforcement in colloquial English by another negative. (5)

a.

ne hyrde ic cymlicor ceol gegyrwan not heard I more.beautiful ship prepare ‘I never heard before of a ship so well prepared’ (c1000, Beowulf 38–39).

Introduction

5

b. And þa æfter micelre tide sæt ic ana in þam and then after much time sat I alone in that westenne and ic ne geseah nowiht buton eorðan. Desert and I NEG saw nothing except earth ‘And after a long time, I was alone in the desert and saw nothing except earth.’ (DOE, c1000, Lives of Saints, 310–311) c. Nes þis meiden nawhit heruore imenget in hire mod inwið. Not.was this maiden not therefore troubled in her mind within ‘This maiden was not troubled in her mind because of this.’ (c1250, Katerine, 28, 21–22) d. Yit it semeth that He wolde not leve thee thus lightly ‘Yet it seems that he wanted not leave you thus lightly.’ (c1400, Cloud of Unknowing, 241–242) e. And to þis I cannot answere þee bot þus: ‘I wote neuer.’ ‘And to this I can’t answer you except thusly: I knew never.’ (c1400, Cloud of Unknowing, 450–451, www.lib.rochester.edu/ camelot/teams/cloud.htm) f. I don’t want nothing better’n this. (1876, Mark Twain, Tom Sawyer, Chapter 13) Cycles involve grammaticalization, which is a process whereby new grammatical categories are created from lexical categories. This process may go hand in hand with a loss of phonological weight and semantic specificity. It is often put as the cline in (6a), which includes both semantic loss (content > grammatical) and morpho/phonological reduction (word > clitic > affix); the cline in (6b) emphasizes the degree of association. (6)

Grammaticalization involves the following changes: a. content item > grammatical word > clitic > inflectional affix > zero. (Hopper & Traugott 2003: 7; Narrog & Heine 2021: 130) b. free > periphrastic > affixal > fusional (Dahl 2004: 106)

Heine and Reh (1984: 15) define grammaticalization as a change where “linguistic units lose in semantic complexity, pragmatic significance, syntactic freedom, and phonetic substance.” I will assume these additions to the definition in (6). Definitions of grammaticalization typically not only include the primary change, as in (6), from content item to grammatical word, but also secondary grammaticalization, from grammatical word to more grammatical. Kuryłowicz (1964: 52) makes this distinction in his definition of grammaticalization as “advancing from a lexical to a grammatical or from a grammatical to a more grammatical status.” Andersen

6

Introduction

(2008) labels these two processes differently: lexical elements becoming grammatical are called grammation, and grammatical elements becoming other grammatical element are called regrammation, but I won’t make a strict distinction. The decrease in phonological content in (6) is, however, not a necessary consequence of the loss of semantic content (see Kiparsky 2011; Kiparsky & Condoravdi 2006; Hansen & Visconti 2009; Hoeksema 2009). For instance, Kiparsky (2011: 19) writes “in the development of case, bleaching is not necessarily tied to morphological downgrading from postposition to clitic to suffix.” Thus, Kiparsky and Condoravdi (2006) find no evidence for phonetic weakening in the Greek Negative Cycle. Ansaldo and Lim (2004), Post (2007), and Bisang (2008) point out that grammaticalization in East Asian languages may also show marked differences from the cline in (6), particularly regarding phonological reduction. Schiering (2010: 85) links the presence of cliticization and affixation to language-type: “[t]he highest potential for erosion is predicted for stress-based languages, whereas reduction is expected to be limited to the junctural context in syllable- and mora-based languages.” So cluster simplification is possible in all types, but vowel reduction/weakening are typical for languages where stress is realized by pitch, intensity, and duration. His case studies are stress-based German and syllable-based Turkish. In addition, Kiparsky and Condoravdi suggest pragmatic and semantic reasons for the start of that change. Hansen and Visconti (2009), likewise, argue that renewal takes place in pragmatically marked contexts, for instance, in negating discourse-old propositions. In this book, I show that both beginnings are possible: phonological weakening can constitute the first change, later triggering renewal, or pragmatic strengthening can be first, later triggering phonological loss. This means we’ll have to revise the definition in (2) to that in (7), where (a) is the original definition and (b) is added. (7)

The Linguistic Cycle involves [revised formulation] (a) a set of changes where lexical expressions weaken to grammatical morphemes and grammatical items become more grammatical. The original words and expressions are renewed by lexical expressions that may be different or the same as the original. or (b) a set of changes where the semantic and pragmatic features are strengthened by a new expression which triggers a weakening of the older form.

These two forces have been referred to as push and pull changes, respectively (see e.g. Breitbarth et al. 2020: 74). Hopper and Traugott (2003: 2–3) stress that grammaticalization has a diachronic as well as synchronic perspective. In outlining the typical characteristics of grammaticalization of be going to, they identify a change from a very specific purposive directional construction, as in (8), to an inferred future, as in (9).

Introduction

(8) (9)

7

I am going to marry Bill (i.e. I am travelling to marry Bill). I am going to like Bill.

In the change from (8) to (9), the lexical verb go changes to an auxiliary, because in (9) go needs another verb in the sentence to function as a future marker. Once this change has taken place, the new form can undergo processes that auxiliaries typically undergo, such as phonological reduction in (10a). The main verb does not undergo such weakening, as the ungrammaticality of the reduced lexical verb gonna in (10b) shows. (10) a. I’m gonna leave for the summer. b. *I’m gonna the store later. Even if gonna in (10a) is now a future auxiliary, it retains some of its earlier flavor of intent to do something and of motion, but this may also disappear, and be going to/gonna might become a general future, replacing will/’ll in what could be called a Future Cycle. Nesselhauf (2012) provides a very precise account of the changes in the various future markers (shall, will, ’ll, be going to, be to, and the progressive) over the last 250 years. She identifies three crucial features, intention, prediction, and arrangement, and argues that as the sense of intention is lost and is replaced by the sense of prediction, new markers of intention will appear, that is, renewal takes place. One such candidate for renewal is want where intention is expressed in (11a), and it is starting to gain the sense of prediction, as in (11b). (11) a. The final injury I want to talk about is brain damage. (Nesselhauf 2012: 114) b. We have an overcast day today that looks like it wants to rain. (Nesselhauf 2012: 115) The second aspect of the linguistic cycle is renewal, which is discussed less often than grammaticalization. For von der Gabelentz, as we can see in (3), renewal is due to a need for clarity “which disallows . . . erosion and the destruction of the language.” Hagège (1993: 150) similarly states that speakers “are unconsciously striving to make their words clearer.” Hopper and Traugott’s definition is that “[r]enewal results primarily in alternate ways of saying approximately the same thing, or alternative ways of organizing linguistic material” (2003: 122). Smith (2006: 141–142) defines renewal as “the refilling of a semantic domain with new phonetic material, often a periphrastic construction, thus renewing the expressivity of that domain after the natural forces of language change have weakened its semantic content.” As Givón (2018: 199) puts it colorfully, “[a]s more attractive alternatives become available, decrepit old constructions are retired, replaced by shining new ones.” This is expressed as (12a), which mirrors (7a).

8

Introduction

However, it is also possible that the renewal occurs first for pragmatic reasons, as stated in (12b), which is congruent with (7b). (12) Renewal occurs (a) when a grammatical category has become less salient and needs clearer expression and (b) when emphatic needs result in a twofold expression of a grammatical category. In (5), we saw the weakening of a negative and a renewal through another negative. The renewal or clarifier may also be a minimizer (pas ‘step’ in French), a positive adverb (wehe ‘indeed’ in Egyptian; see Gardiner 1904), or a maximizer (all the tea in China or for a million, as in English (13)). These are all reinterpreted from small or definite into negative. (13) a. You wouldn’t want to be in Sydney for all the tea in China. (COCA news 2000) b. She wouldn’t stay at home for a million bucks. (COCA website 2012) As we’ll see in Chapter 4, the Negative Cycle can be argued to start as either due to (12a) or (12b). Concluding, a cycle describes a change from an independent lexical word to a more dependent grammatical morpheme accompanied by a renewal of the erstwhile lexical word. Renewal can also set a cycle in motion. 2

What kinds of cycles exist?

In this section, I distinguish between cycles that are generally accepted and those around which some skepticism exists. Examples of uncontroversial cycles comprise the negative, future, modal, aspect, interrogative, copula, and determiner. Following van Gelderen (2013), these will be referred to as micro cycles. More controversial are macro cycles, that is, cycles that shift a language from analytic to synthetic and from synthetic to analytic. I provide a list of cycles in Table 1.2 and will discuss the difference between these types of cycles later in this section. Each of these cycles is complemented by a renewal of the element, which I have put in bold and which will become clearer when we discuss the cycles in detail in Chapters 3 to 5. I sometimes speak of the Case Cycle or the Copula Cycle, but there are always more than just one cycle! Many of the early works on cyclical change analyze the phenomenon as a change in the typological character of a language, e.g. from analytic to synthetic, as in von der Gabelentz (1901) and Hodge (1970) mentioned previously. This view, however, is not always accepted, and there are also challenges in defining

Introduction

9

TABLE 1.2 Examples of cyclical change

Micro Cycles Determiner Cycles demonstrative > article and original demonstrative optionally reinforced TMA Cycles motion verb > future volition verb > modal and original lexical element renewed telic adverb > perfective aspect > past tense Negative Cycles negative argument > negative adverb/particle > zero and negative renewed negative verb > auxiliary > negative > zero Interrogative Cycles negative > polarity Q > zero and original optionally renewed wh-element > polarity > zero Copula Cycles demonstrative > copula > zero and copula optionally renewed verb/adposition > copula > zero Complementizer Cycles PP/adverb > complementizer and original lexical form renewed Macro Cycles = Analytic to synthetic to analytic Subject and Object Agreement Cycles demonstrative/emphatic > pronoun > agreement > zero and optional new original Case Cycles verb/noun > preposition > case > zero and new V/N

}

}

}

}

the terms analytic and synthetic. Heine et al. (1991: 245), building on the work of Givón (e.g. 1971, 1976), distinguish three kinds of cyclical change. The first only refers to “isolated instances of grammaticalization,” when a lexical item grammaticalizes and is then replaced by a new lexeme. An example would be the Old English lexical verb willan (‘want’) being used as a future marker and then being replaced by a new volition verb want, as well as the change from (8) to (9). Other examples of this type of cyclical change are discussed regarding thus by Wood (2016) and regarding rather by Gergel (2016). One could argue that these isolated changes have wider implications and should therefore be counted as examples of the second type, which refers to “subparts of language, for example, when the tense-aspect-mood system of a given language develops from a periphrastic into an inflexional pattern and back to a new periphrastic one” (Heine et al. 1991: 245). Since negatives often interact with tense, mood, and aspect, they should be counted here and not as isolated changes. Examples of the second type of change are provided in Mithun (2011, 2016), Veselinova (2016), van der Auwera and Vossen (2016). Because they are hard to separate and involve subparts of the grammar, I collapse the first two types into one and regard them as micro cycles, defined as in (14).

10

Introduction

(14) A micro cycle involves (a) individual changes where a lexical word or phrase grammaticalizes as a grammatical morpheme or a grammatical word as a more grammatical expression and is renewed by another lexical expression or (b) cyclical changes in a subpart of the grammar involving grammaticalization and renewal. The third type of cyclical change that Heine et al. (1991) identify applies to entire languages and especially to language types, and I therefore refer to these as macro cycles, as defined in (14). (14) Macro cycles involve grammaticalization and renewal in ways that affect entire languages and that change language typologies. The descriptions of languages changing from analytic to synthetic to analytic by von der Gabelentz and Hodge, given previously, fit this kind. More recent discussions of this change appear in McWhorter (2016), Szmrecsanyi (2016), and Haspelmath (2018). When pronouns are reanalyzed as agreement, this is an increase in its synthetic character. Evidence for the change from pronoun to agreement can be seen in Navajo (15) where the agreement is a shortened pronoun. (15) a. (Shi) diné bizaad yíní-sh -ta’ I Navajo language 3–1S-study ‘As for me, I am studying Navajo.’ b. (Ni) na-ni -l-nish? You ASP-2S-CL-work ‘Are you working?’

Navajo

As will be explained further in Chapter 5, I will consider the Pronoun to Agreement Cycle as one involving a change in language type (revising my arguments in van Gelderen 2013). The reason is that it signals a change towards a language with a very synthetically marked verb, also known as head marking language (as the term is used in Nichols 1986). Similarly, the Case Cycle will be considered a macro cycle since it changes the language towards synthetically marked nominals and is also known as dependent marking. Heine et al. (1991: 246) argue that there is “more justification to apply the notion of a linguistic cycle to individual linguistic developments,” for example, the development of future markers, of negatives, and of tense, rather than to changes in typological character, as in the development from analytic to synthetic and back to analytic. Their reasons for caution about the third type of change, that is, a cyclical change in a language’s typology, is that we don’t know enough about older stages of languages. This cautionary sentiment is reflected in

Introduction

11

the work of other linguists, and whereas most researchers are comfortable with cycles of the first and second kind, many are not with cycles of the third kind. As mentioned before, Jespersen (1922; Chapter 21.9) criticizes the concept of cyclical change. His criticism is based on his views that languages move towards flexionless stages in a unidirectional manner and that they do not develop new morphology. Jespersen’s views cannot be correct because languages and families such as Finnish, Altaic, and Athabascan increase in morphological complexity through a cyclical process (see van Gelderen 2011). Later views are reviewed in Chapter 5. This book discusses both micro and macro cycles and will show the processes at work are very similar. Lastly, Hansen (2014, 2020: 168) and others propose the term ‘Cycles of Pragmaticalization’ for the change where a lexical item gains pragmatic functions, as in when a temporal adverb is used as discourse marker. These are also micro cycles and will be discussed in Chapter 4. Concluding, the literature distinguishes between cycles that involve isolated parts of the grammar or subparts of the grammar and those that involve entire grammars. I refer to those as micro cycles and macro cycles, respectively, and will discuss them in separate chapters. 3

How and why to study cyclical change

Narrog and Heine (2021: 25–40) discuss how to study grammaticalization, and I will follow their discussion in this section as it applies to cycles. I will also provide some practical suggestions on how to get started in this line of research and reinforce that in the exercises to the chapters. I finish by looking at motivations for cycles. 3.1 The practical side

Narrog and Heine mention three reasons to study grammaticalization, diachronic, synchronic, and for reconstruction, and provide examples of specific studies. For instance, a diachronic approach would study the use of modals or the progressive over a period of time and come up with a scenario. A synchronic approach would discuss the use of particular constructions as more or less grammaticalized. The latter method could also compare the stage of grammaticalization of a lexical item across different languages. Narrog and Heine provide a comparison of the verb ‘threaten’ in eight languages to see in which it has grammaticalized with a meaning of ‘be about to’ most. They mention Portuguese, German, and English as having the verb in the most grammaticalized stage. In (16a), the English threaten is a lexical verb, and in (16b), it has grammaticalized to a marker of an imminent event. This is also possible in Portuguese, as (17) shows.

12

Introduction

(16) a. The bouncer threatened us. b. It threatened to rain. (17) Um gordo e rubicundo merceeiro . . . ameaçava Portuguese a fat and reddish merchant threatened estalar tôdas as costuras da farda. tear all the seams of.the costume ‘A fat, reddish trader was about to burst out of his attire’ (de Lima 2006, cited in Narrog & Heine 2021: 33). Finally, in order to apply internal reconstruction and the comparative method to reconstruct earlier stages, reverse grammaticalization can be applied. Givón has successfully applied this to the reconstruction of Bantu and Athabaskan (e.g. Givón 2000). The same three-way division can be made for research on cycles. People study cycles diachronically; for example, the Determiner Cycle has been scrutinized from Latin to modern Romance (Harris 1977; Giusti 2001) and from Old Norse to modern Scandinavian (Faarlund 2004). Linguists study it to find where the cycle starts and how it proceeds. Synchronically, particularly in relation to the Negative Cycle, it has been argued that specific languages are in stage 1, 2, or 3 of Jespersen’s Cycle (see e.g. van der Auwera & Vossen 2016). In terms of reconstruction, cycles have proved useful in reconstructing negation in Athabaskan (cf. van Gelderen 2008b), on reconstructing pronouns and demonstratives in the Yuman family (cf. Langdon 1968, 1977; Gordon 1986), and pronouns and demonstratives in Pashto (cf. Powell 2021). In this book, my methodology is mainly diachronic and synchronic. On the one hand, I examine well-documented cyclical changes that have been described in the literature, and on the other hand, I look at grammars of contemporary languages to see evidence of past cycles. If you, the reader, are someone who has never researched grammaticalization or cycles before, it may help to first read the chapters on micro cycles in this book. This will familiarize you with the types of change that occur. Then, you could select pronouns or negation or demonstratives and compare these in languages you know something about or can find grammars of. The languages can be in the same family or can be in varieties of the same language. Take, for instance, the pronouns in four Athabaskan languages, shown in Table 1.3, of

TABLE 1.3 Pronouns in four Athabaskan languages

Navajo

Apache

Koyukon

1 sg

shi

shíí

see

Dëne Sųłiné si/sį

2 sg

ni

nih(í)

nen(h)

nën

(Bray 1998 for (Western) Apache; Jetté & Jones 2000 for Koyukon; Cook 2004 for Dëne Sųłiné)

Introduction

13

which we’ve seen the Navajo pronouns in (15). From this comparison, you’ll notice that the pronouns are shared in the family. To see a possible cycle in action, you’d look at the verbal inflection and observe, for example, in (15), that the first (and second) person markings on verbs in these languages are phonologically weakened forms of the full pronouns of Table 1.3. All things being equal, the pronoun could have lost phonology and become the agreement, or the agreement could have been enhanced phonologically to become the pronoun. The Pronoun Cycle will tell you the former is the right conclusion, as we’ll discuss more in Chapter 5. Another example may be to consider a word that could in principle grammaticalize, such as the demonstrative that in English, and then imagining where the renewal might come from, such as a new adverb there. Then you could check a contemporary corpus (British National Corpus/BNC or the Corpus of Contemporary American English/COCA) to see if the use of there is increasing with a determiner. Since that is also a complementizer, you could restrict the search by searching for that as a d(eterminer): simply input the following syntax into the search bar (use * to skip words) [that_d there] or [that_d * there]. The first search returns sentences, such as in (18), and the second, as in (19). (18) (19)

This here’s my place, and that there car belongs to me. (COCA 2005 movies) That man there must be over 15 feet tall! (COCA 2017 movies)

In Chapter 3, we’ll take a look at this cycle more and see that this involves a change occurring in English. This may be a good point to explain how I have selected the studies included in this book. Many of the cycles I discuss are known from the literature. I haven’t made a balanced typological survey of a set of the world’s languages but have used known descriptions of cycles and, in addition, have looked at descriptions of, for instance, negation and determiners in various grammars. These grammars are of languages from different families and areas, even if the sample of languages isn’t balanced typologically. If enough is known about the history of a language, cyclical change will become obvious, for example, the current work on specific languages (e.g. Hungarian by Bacskai-Atkari 2014b; Bácskai-Atkári & Dékány 2014; Kiss & Mus 2022) and on specific cycles (e.g. the Comparative Cycle by Jäger 2021). One of the reviewers asked if I am cherry-picking the examples that work well. Where possible, I list counterexamples, such as with the Voice Cycle and with languages that resist the Pronoun Cycle. So far, I have reviewed three ways that grammaticalization and cycles can be studied: by looking at one stage of a language, that is, synchronically; by examining it in stages, that is, diachronically; and by using it to reconstruct earlier stages. I have also suggested a few ways to research cycles. Next, I will mention some of the reasons that linguists study cycles.

14

Introduction

3.2 The theoretical side

Since linguists explore language for different reasons – some are interested in its (communicative) use and others in the cognitive structures that produce it – this different focus will be evident in the reasons why cycles are studied. I will first consider the original explanation, given by von der Gabelentz, in terms of clarity versus comfort, and used by Meillet and Jespersen. I then provide some specific reasons for various current approaches and come back to these in Chapter 6. Cycles have often received functional explanations, as in (3) in the previous section by von der Gabelentz (1901: 256), namely, in terms of a weakening of the endings due to comfort and a strengthening of the original idea due to clarity: “the impulse toward comfort, which leads to the wearing down of sounds, and that toward clarity, which disallows this erosion and the destruction of the language” (von der Gabelentz 1901: 256). In the original German, von der Gabelentz uses “Deutlichkeit” (‘clarity’) and “Bequemlichkeit” (‘comfort’) as the competing factors. I’ll provide a bit more background about what the two terms mean for him. Von der Gabelentz gives specific examples of comfort, and they include the unclear pronunciation of everyday expressions, the use of a few words instead of a full sentence, that is, ellipsis (pp. 182–184), “syntaktische Nachlässigkeiten aller Art” (‘syntactic carelessness of all kinds,’ p. 184), and loss of gender (p. 254). He also gives various examples of clarity, namely, special exertion of the speech organs (p. 183), “Wiederholung” (‘repetition,’ p. 239), periphrastic expressions (p. 239), replacing words like sehr ‘very’ by more powerful and specific words such as riesig ‘gigantic’ and schrecklich ‘frightful’ (p. 243), using a rhetorical question instead of a regular proposition, and replacing case with prepositions (p. 183). For von der Gabelentz, the course of the spiral is connected with the agglutination theory, as evident in (3), that says that all affixes were once independent words (p. 255). The examples he gives of comfort and clarity do not specifically relate to the micro cycles, however, and even the cases of words turning into affixes are all pretty general – not grounded in very specific concrete examples. In (3), von der Gabelentz states that languages may have affixes that then require a new expression after they are ground down. The new expression may be constructed “through word order or clarifying words.” He argues that languages change from fusional and agglutinative systems to isolating systems and then again develop into agglutinating ones. Meillet’s (1912) work on language change as grammaticalization is an obvious source for ideas on cyclical change. His examples of grammaticalization are many: the French verb être ‘to be’ going from lexical verb to auxiliary, aller ‘to go’ changing from a verb of motion to a future marker, and the Greek thelô ina ‘I wish that’ changing to a future marker that is much reduced in phonology,

Introduction

15

namely, tha. As for cyclical change, Meillet argues that languages add words to obtain a more intense expression (“pour obtenir une expression intense”), then weaken to become simple grammatical categories (“de simples outils grammaticaux”), and then new words are added. After this stage, weakening starts and so on without ending (“sans fin”) (Meillet 1912: 140). Jespersen has reservations about cycles, especially macro ones, although he discusses linguistic change as a tension, a ‘tug-of-war,’ between the speaker’s needs and those of the community. Jespersen does not see this ‘tug-of-war’ resulting in cyclical change and in that way it is different from von der Gabelentz’ two tendencies. He writes that “the correct inference can only be that the tendency towards ease may be at work in some cases, though not in all, because there are other forces which may at times neutralize it or prove stronger than it” (1922: 262), although he mainly gives instances of weakening, that is, phonetic ease. It isn’t always easy to decide which sounds are easier and in which contexts. Most examples of ease that Jespersen gives are phonetic, for example, [h] being an easier fricative than [s] or [f]. His example of the Negative Cycle is not put in terms of a cycle but in terms of tendencies to weaken and strengthen. In Jespersen (1941: 15), he continues the use of tendencies: “In linguistic changes we see the constant interplay of two opposite tendencies, one of an individual, and the other of a social character, one towards ease and the other towards distinctness,” clearly echoing von der Gabelentz’ formulation in (3). Having provided more background on the traditional explanation of the cycle as a balance between comfort and clarity – but more on that ‘balance’ in Chapter 6 – I now turn to explanations given in more recent times. The most enthusiastic proponents of linguistic cycles seem to be generative linguists, witness the number of recent publications, for example, van Gelderen (2007a, 2008a, b, 2009a, b, 2011), Willis et al. (2013), and Bouzouita et al. (2020). The reason behind this interest is that generativists see uniform linguistic change, grammaticalization as well as renewal, as a window on the faculty of language. Grammaticalization is seen as driven by the child acquiring the language. Explicitly, a child may reanalyze a phrase as a head and a head as a more grammatical head from the language they hear, that is, as having a simpler structure and fewer semantic features. Grammaticalization is economical from an acquisition point of view. Renewal is worthy of study as well in that it provides insight into which features are needed (and therefore need to be renewed after the final stage of their grammaticalization) and which are not. A negative will never disappear, so the cline in (6) never reaches zero before a renewal is present. In the case of a dual number, such as Old English wit ‘we two,’ it can reach zero without being renewed. This provides insight into core grammatical features. What is the set of concepts universal to our species and others? Within generative grammar, McCawley (1971) and Katz and Fodor (1963) emphasize the universal character of semantic representations. They use semantic markers such as [human],

16

Introduction

[young], and [male], to decompose the meaning of a word “into its atomic concepts” (Katz & Fodor 1963: 186). Chomsky (1965: 142) writes that “semantic features . . . are presumably drawn from a universal ‘alphabet’ but little is known about this today and nothing has been said about it here.” The ability to categorize is not unique to humans, however. Certain animals are excellent at categorization, for instance, prairie dogs have sounds for specific colors, shapes, and sizes (Slobodchikoff 2010), and Bickerton (1990) suggests that pre-linguistic primate conceptual structure may already use symbols for basic semantic relations. Other theoretical frameworks are interested in grammaticalization. For instance, Bybee’s (2010) usage‐based theory relies on grammaticalization. For her, grammaticalization is crucial in providing insight into the mechanisms of grammar. High(er) frequency of usage encourages phonetic reduction and loss of compositionality and structure. An instance given in Bybee (2002) is the grammaticalization of be going to, which we also saw discussed previously. The high frequency of be going to gives rise to a neuromotor routine whereby it is processed as one unit. Cycles do not feature in this approach but are consistent with it. Sociolinguistic and stylistic approaches have also been taken. For instance, shorter, more grammaticalized forms are used more often by younger speakers and males (e.g. Lorenz 2020; Levshina & Lorenz 2022). And some innovations are introduced in more colloquial contexts (e.g. Finegan & Biber 2001). Construction Grammar is interested in constructionalization and constructional change, as we’ll see in Chapter 6. Goldberg (1995) and Croft (2001), in particular, have developed this framework with a focus on form-meaning pairs. Traugott (2008) points out why this framework works well to explain grammaticalization as it pays attention to all linguistic parts, phonology, morphology, syntax, and pragmatics. However, she does not engage in how Construction Grammar could explain renewal, that is, the other part relevant to the cycle. Traugott turns out to be an exception, and as Narrog and Heine (2021: 263) point out, Construction Grammar is not particularly concerned with grammaticalization for a number of reasons: for example, some stress change can go in every direction (Hilpert 2021). Probably as a result, I have also seen no work focused on the linguistic cycle in Construction Grammar. Like Traugott, Sommerer (2015) and Culicover (2021) discuss Construction Grammar and grammaticalization but avoid the cycle as well. Finally, Haspelmath (2018), using a broad functional approach, examines the Analytic to Synthetic Cycle in a new light. He writes that “older tightly bound constructions often get competition from new constructions based on content items, which may eventually replace the older patterns.” He calls this an anasynthesic spiral, something we’ll come back to in Chapter 5.

Introduction

4

17

Major questions in the study of cycles

We have some sense for what keeps some cycles going. As was mentioned previously, because negatives are so important in language, pragmatic strengthening is frequent, and this stimulates new forms next to pragmatically weaker ones (see also Meisner & Pomino 2014). Modal cycles are also omnipresent, as Nesselhauf (2012) and van Gelderen (2020) have shown. Using Traugott and König’s (1991) influential inferencing model, one can say that modals originate in verbs of motion, ability, volition, and intention that ‘invite’ certain inferences of futurity and possibility. A new modal is the result of the reanalysis of a pragmatic function into a grammatical category or feature. Looking at changes in a cyclical manner has brought up new questions. Some of the crucial descriptive questions in relation to the mechanisms of the linguistic cycle are summarized in (20). These hold for both micro and macro cycles. (20) Questions in relation to the linguistic cycle are as follows: a. Which cycles exist and why? This boils down to: which semantic and grammatical features participate in cycles? b. What are the sources of renewal once a cycle has desemanticized a lexical item? c. Which process starts the cycle? d. Are there typical steps in a cycle? e. Do cycles have preferred stages? f. What structural factors favor/disfavor cycles? Do cycles conflict or conspire with each other? g. What’s the role of language contact? Turning to the specific questions in (20): (a) and (b) are best answered by studying Kuteva et al. (2019), who provide an encyclopedia of grammaticalization. In it, they identify categories that are typically grammatical in languages of the world: time, space, cause, completion/telicity, intention, existence, negation, number, and person. The list of changes also shows the typical sources of renewals. The prevalence of certain cycles shows which of these categories are important for all languages and shed insight on the (human) cognitive faculty. Lehmann (1982/2015) discusses channels of various grammaticalizations, and this too is helpful. The grammaticalization literature lists rarer cycles too. One example is Mithun’s (2016) distributive cycle, where verbal affixes cause the event/state to be spread in space, time, and participants. An answer to (a) and (b) is found in Chapters 3, 4, and 5, where I focus on the main cycles, for example, negatives, determiners, tense, mood, aspect, agreement, and case, and these involve the same grammatical features that typologists identify (see Bickel & Nichols 2005: 220). Some cycles not discussed in this book are listed in Chapter 7.

18

Introduction

Answers to question (c) remain controversial: will the old form first weaken phonologically or not? As we have seen before, although Meillet (1912: 139) acknowledges the role of the weakening of pronunciation (“un affaiblissement de la prononciation”), he believes that what provokes the start of the (negative) cycle is the need to speak forcefully (“le besoin de parler avec force”). Therefore, the loss in phonological content is not a necessary consequence of the loss of semantic content (see also Hoeksema 2009). Veselinova (2016) provides an interesting perspective on the question of when the renewing element appears and how long it stays around: stages with variation are quite stable. Kiparsky (2011: 19) argues “in the development of case, bleaching is not necessarily tied to morphological downgrading from postposition to clitic to suffix.” As mentioned before, the answer to this question is that either scenario occurs, and this point will be discussed in more detail in Chapter 6 after we’ve examined specific cycles. As for question (d), van der Auwera (2009) carefully lays out the typical steps of the Negative Cycle but also some variation in the stages. Van der Auwera and Vossen (2016) provide further evidence for solid Jespersen Cycles in two languages of the Americas, with certain peculiarities, such as the role of the irrealis as renewing stage. Gergel (2016) compares English and German comparative temporal adverbs that acquire modal meanings and shows why they end up undergoing slightly different changes. In this book, I provide some structural steps that all cycles go through and provide a reason behind them in Chapter 6. Question (e) will be addressed in Chapter 6 as well. We know that some languages prefer verbal agreement, especially in the case of agreement with the subject, over marking the argument through an independent pronoun. Languages also use negative particles more frequently than clitics. The ultimate answer isn’t forthcoming, but I provide an inventory of the ‘preferences.’ Answering (f) contributes to the ‘definition’ of the linguistic cycle. From a structural point of view, two main types of cycle occur: those where a full phrase strengthens the original lexical word, that is, where doubling occurs, such as in the traditional Negative Cycle (Jespersen 1917; van der Auwera 2009), and those where a new word is used, without co-occurrence of the old and new, such as in Croft’s Existential to Negative Cycle (Croft 1991; van Gelderen 2008a). Veselinova (2016) discusses this cycle, and many researchers comment on how ‘their cycle’ is a typical negative cycle or not, for example, Mithun (2016) and Pye (2016). The circumstances surrounding some of the other cycles are still mysterious, such as why (Subject) Pronoun Cycles do not occur in East Asian languages. Chinese, Thai, Korean, and Indonesian (to name but a few) do not show tendencies to phonologically reduce the subject pronoun or to restrict its occurrence to a preverbal position. Subject pronouns themselves have been renewed through

Introduction

19

nouns in some of these languages (Thai and Indonesian), and in these, there is a variety of pronouns to choose from, but there has been no instance where the pronoun is reanalyzed as agreement marker. Structural change factors, such as word order and the particular position of the verb ‘to be,’ may play a role; for example, negatives are focused and therefore may be highlighted by ‘be’ and a combination of ‘be,’ and the negative can become the new negative. Cycles may conflict, and this structural conflict may constitute a slowing (or accelerating) factor; for example, Croft’s and Jespersen’s Cycles seem to work together (in Athabaskan and Ugric, see van Gelderen 2011), and Subject and Object Pronoun Cycles do interact (see Bahtchevanova & van Gelderen 2016). Copula verbs have many sources, demonstratives, adpositions, existence and motion verbs, and the choice of which of these emerges as the copula must be due to structural factors. English has a strong tendency to recruit verbs, and that may be because it has a prominent T(ense) position; many Semitic languages and most creoles reanalyze demonstratives as copulas because these languages use topicalizations. To answer question (e), I will also examine interactions of cycles in Chapter 4. Regarding (g), external reasons for accelerating or impeding the rate of change need to be studied. Heine and Kuteva (2005) have argued for accelerated grammaticalization due to language contact. Van der Auwera and Vossen’s (2016) data show areal clusters in the stages of the negative cycle and Mithun suggests language contact as a potential contributor to cyclical change in the languages of Northern California because certain cycles appear in certain linguistic areas. She shows for Central Pomo, a Pomoan language, that directional prefixes on a verb are renewed through adverbs that are themselves being cliticized on a verb, which is a feature present in the other languages as well. Szmrecsanyi’s (2021, 2016) data show that something major happened to English around 1200: the introduction of articles and the loss of case marking. A contact explanation with Celtic or Old Norse would of course make sense (see e.g. Filppula et al. 2002, 2008; Emonds & Faarlund 2014, respectively). McWhorter (2007) argues that languages with a lot of adult language learning remain in one stage. His examples include English, Mandarin Chinese, Persian, Colloquial Arabic, and Malay. Most of the phenomena these languages have in common involve loss of complexity, for example, gender marking, affixes, case, obligatory TMA marking, and so on. Related is the question of complexity and simplicity when the former is seen as due to social isolation and the latter as due to language contact (see Trudgill 2011). The linguistic cycle predicts that any language will accrue complexity when it changes from analytic to synthetic and lose this when it goes from synthetic to analytic again, and the same is true with the other (micro) cycles. Modern English, with more non-native than native speakers, is still accruing complexity, as shown in Szmrecsanyi (2016). An answer to question (g) will be investigated in Chapter 6.

20

Introduction

5

Terminology, resources, and glossing

In this section, I discuss some of the relevant terminology, in particular the difficulty distinguishing a word, from a clitic and an affix. The distinction between a word and a phrase is also considered and becomes relevant in Chapter 6. Finally, I provide some details on the origin of my data and how I have glossed them. Some terminology concerning cycles is controversial. The cline in (6), repeated here as (21), presupposes a clear semantic distinction between content (or lexical) word and grammatical word and a morphosyntactic distinction between word and clitic and between clitic and affix, but these distinctions are not always so clear-cut. (21)

content item > grammatical word > clitic > inflectional affix > zero.

Lexical words are content items, have detailed semantic content, belong to an open class, can be stressed, and can support arguments (assign an Agent or a Goal argument); grammatical words lack these properties. For instance, when willan was a main verb, it could appear by itself and assign Agent and Theme to its arguments, as in (22), where God is the Agent and it the Theme. (22)

Se ealdorman gewat þa ða hit wolde God. The nobleman departed then then it wanted God ‘The nobleman died when God wanted it’ (OED, c1000, Ælfric Lives of Saints 432).

There are, however, words that do not completely belong to either group. Some lexical words may never completely be reanalyzed as grammatical words, something Corver and van Riemsdijk (2001) have called semi-lexical categories. Adpositions are among these because they are not an open class (evidenced by the fact that they are rarely borrowed) but can be stressed and contribute specific meaning. Examples in (23) show the gradual nature of this class. (23)

Lexical/content – despite, above

semi-lexical – out, off (particles)

grammatical of, to (case markers)

On the morphosyntactic side of (21), it is likewise notoriously hard to distinguish words, clitics, and affixes. Haspelmath (2011a: 32) writes that there is no “good answer to the question of how to define the notion of word in a clear and consistent way.” He reviews several word-defining properties, e.g. semantic, orthographic, phonological, and morphosyntactic independence. For the latter, he examines such criteria of morphosyntactic wordhood as potential pauses, a free occurrence, an external mobility and internal fixedness, and concludes that these are not “of much help for identifying morphosyntactic

Introduction

21

words” (p. 39). Likewise, Schiering et al. (2010: 657) conclude that “the ‘word’ has no privileged or universal status in phonology, but only emerges through frequent reference of sound patterns to a given construction type in a given language.” Easterday et al. (2021) allow the orthography of the language to decide. The distinction between affix and clitic is also hard to make, and certain clitics are treated as bound or as free by different authors (e.g. Szmrecsanyi 2012 and Haspelmath 2011a, respectively). Haspelmath (2011b) puts it as (24). (24)

Since I do not think that words and affixes can be identified consistently across languages, I use the terms “word,” “clitic,” and “affix” in a loose sense here, referring to the kinds of things that are generally separated in linguists’ notations by spaces, equals signs and hyphens, respectively. This is not satisfactory, but I cannot think of a better alternative for the purposes of this chapter. (Haspelmath 2011b, footnote 3)

Haspelmath (2022) also questions the inclusion of clitic in the cline in (21) because there are very few cases of this change. I will continue to use the cline in (21) as it fits determiner and pronominal cycles. Siewierska (2004: Chapter 2) provides a very thorough discussion of independent and dependent markers; in her case, the focus is on person markers, but the same holds for any other marker. By independent, she considers words that are separate and able to bear stress and, by dependent, words that cannot be stressed and are phonologically and morphologically dependent on other words. The latter group is divided in weak, clitic, bound, and zero. Weak forms are not attached but may appear in syntactically restricted positions and clitics seem bound to specific positions (p. 27). Bound forms are what are called affixes in other work, that is, those markers that can only be attached to one word type. As Siewierska notes frequently, what are “considered to be clitics by one author may be treated as bound or independent by another” (p. 26). In this book, I consider an independent word one that can occur in isolation, me in (25) but not I. (25) Q: Who is it? A: Me/*I. A clitic needs to attach to another word, but that word can be anything, as (26) shows for the first person nominative pronoun that attaches to an auxiliary or a conjunction. An affix can only attach to a particular category, for example, the third person -s in (27) attaches only to verbs.

22

Introduction

(26) a. I=m happy doing that ‘I’m happy doing that.’ b. as=I know. ‘As I know.’ (27) He walk-s in the morning. When possible, I’ll mark the clitic boundary as =, as in (26), and the affix boundary as –, as in (27). If it is not known if the bound form is a clitic or affix, I’ll use –. In addition to word, clitic, and affix, I will also use the syntactic characterizations of phrase and head of a phrase. In most cases, this distinction is clear because typically a phrase consists of more than one word. Take a noun phrase (NP) such as the one in (28). It is an NP because it is headed by the noun discussion. (28)

The interesting discussion about the lunar eclipse.

In (generative) structures, heads and phrases occupy different positions, and many of the changes that we will see in this book are not only from word to affix, as in (21), but also from phrase to head (word), as in the case of Prepositional Phrases changing to complementizers, as shown in (29), and exemplified in (30). This change will be elaborated on in Chapter 4. (29) Phrase > head After that after (30) After þat Crist had ordeynid his apostlis > after I met wyth hym in þe strett ‘After that Christ had ordeigned his apostles.’ ‘After I met him in the street.’ As for sources for the data used in this book, I have read handbooks and grammars (and am grateful to the careful work of the authors) in order to find evidence of cycles and cyclical change. I haven’t surveyed a typologically and geographically systematic sample. Where relevant, I have consulted the online World Atlas of Language Structures (www.wals.info) to get an idea of where certain phenomena might be found. Since I use a variety of languages and stages of languages, the examples I found constitute only the tip of the proverbial iceberg and are heavy on Indo-European and languages of North America. The corpora used include the Corpus of Contemporary American English (COCA), the Corpus of Historcal American English (COHA), the British National Corpus (BNC), the Corpus d’entretiens spontanés (CdES), the Orléans Corpus, and the Corpus de Español. I have also made use of individual electronic texts, made available by the Oxford Text Archive and the Dictionary of Old English project (DOE) at the University of Toronto. The latter allows one to search all of Old English. With many of these texts, I have used the concordance program MonoConc. For languages such as Norwegian, Swedish, Dutch, Urdu/Hindi, and

Introduction

23

French, I have used Google (advanced) searches. The choice of which to use depends on the size of the query. For instance, when looking at all instances of French moi ‘me,’ it is more convenient to select a smaller corpus. For recent innovations, such as the change from must have to musta, I have used Twitter data. As for the dates of my historical English sources, if the precise date is known, a(nno) is used; if the date is an approximation, c, for circa is used. Most of the bibliographical references are as given in the electronic resources, such as the DOE corpus or Oxford English Dictionary (OED). If the reference can be found easily, via the OED, DOE, COHA, or COCA, it will not be listed in the bibliography. I did not want to clutter up the bibliography with works I use only once or twice as example sentences. References to Beowulf are from Fulk et al. (2008), to the Old English Gospels from Skeat (1881–7), to the Anglo-Saxon Chronicles (Peterborough and Parker) from Thorpe (1861), to Layamon’s Brut from Brook and Leslie (1963), to Chaucer from Benson (1987), and to Shakespeare from the OTA electronic edition of the First Folio. For references to the Paston Letters, Jane Austin, Mark Twain, etc., see the editions available at www.gutenberg.org. For references to Orosius, Cura Pastoralis, Cursor Mundi, Katerine, Handlyng Synne, and the Cely Letters, see Bately (1980), Sweet (1871), Morris (1874–1893), d’Ardenne (1977), Sullens (1983), Hanham (1975), respectively. I simplify glosses where appropriate but have kept the glosses as much as possible as they were in the original source. When discussing negation and the agreement on the verb or demonstrative is not relevant, I leave that information out. Once in a while, I have left special characters out, for instance, tones in Chinese sentences, since native speakers say they can get the meaning without them in a full sentence. Data from corpora and Google searches often incorporate unusual spellings or punctuation. I have left these as they were in the original. It is often very hard to decide which name of a language to use. I have been pragmatic in this. For instance, Chipewyan is an older name that is well-known to linguists, but the speakers prefer Dëne Sųłiné (even though they are not in agreement about the spelling of it). I have therefore used Chipewyan/Dëne Sųłiné. I have tried to use language names that were explicitly chosen by its speakers, for example, Tohono O’odham instead of Papago, and Athabascan (with b and c) rather than Athapaskan. Other problems arise using umbrella terms like Persian, French, Mandarin, Urdu/Hindi, and of course, English. 6

Conclusions and outline

This chapter has provided an introduction to the linguistic cycle. It gave a definition, examples, and distinguished between micro and macro cycles. I also offered some background on how and why to study cycles, the major questions surrounding the cycle, and some terminological conundrums.

24

Introduction

The outline of the book is as follows. In Chapter 2, a history of the linguistic cycle is provided, from before its first mention in 1877 to the present. Chapters 3 and 4 discuss eight micro cycles and Chapter 5 four macro cycles. The approach in these chapters is data-driven and typological. The book compares different cycles to see how the grammaticalization (e.g. with first person verbs) and renewal (as (9a) or (b)) start. Chapter 6 reviews explanations of the cycle, functional ones that emphasize communicative needs and generative ones that focus on internal reanalysis. Chapter 7 is a conclusion. Suggestions for further reading On grammaticalization: Hopper and Traugott (2003) Grammaticalization and Narrog and Heine (2021) Grammaticalization. On the Linguistic Cycle: Hagège (1993) The Language Builder: An Essay on the Human Signature in Linguistic Morphogenesis and Bouzouita, et al. (2020) Cycles in Language Change.

Review questions and exercises

1. Define the following concepts in your own words: cycle, macro cycle, micro cycle, analytic, synthetic, and spiral. 2. Look up a grammatical word in the OED and see what it says about the origin and development of this word, for example, the modal auxiliary may. How could this information help you understand cycles? 3. How would you go about selecting a cycle to study and actually studying it? 4. Assume the English negative not is weakening phonologically and semantically/pragmatically. Where might a renewing element be found: verbs, adverbs, or nouns? Check in COCA or COHA if the use of this particular lexical element has increased or not. Consult their guidelines on the syntax of search words in order to find different possible renewing elements. 5. Section 4 mentions that the dual doesn’t renew frequently. English is an example of that since Old English has dual first and second person pronouns, wit ‘we two’ in (31) and git ‘you two,’ respectively. The -t ending in these is probably an affixed version of the word twa ‘two.’ (31) Wit þæt gecwædon cnihtwesende we.two that said being.boys ‘The two of us agreed, being boys’ (Beowulf 535–538). Modern English hasn’t renewed these lost dual pronouns. Is there a way in Modern English to express the dual, and could that become part of a cycle? Does Old English (32) give you a clue?

Introduction

FIGURE 1.1

Start followed by an -ing verb



25

2 A HISTORY OF CYCLICAL CHANGE

Work on grammaticalization and the linguistic cycle relies on insights from comparative and historical linguistics. The latter fields of study have deep roots in the Greek, Latin, Medieval, and Renaissance periods, and beyond. The eighteenth century, in particular, provides important breakthroughs in the understanding of the relationships between human languages, for example, in works by Étienne Bonnot de Condillac, Adam Smith, Gottfried Wilhelm Leibniz, William Jones, and John Horne Tooke. The insights include (a) languages are related; (b) content words reanalyze as grammatical words; (c) synthetic languages change to analytic ones and vice versa; (d) there is loss and renewal due to ease and clarity, respectively; and (e) change can be linear (with an end point) or cyclical (with no end point). The nineteenth and twentieth centuries see more work on these issues. Some of the relevant works are Bopp (1816, 1833), von Schlegel (1818), von Humboldt (1822, 1827–9), Curtius (1870, 1885), Whitney (1875), Lefèvre (1877, 1893), Meillet (1912), Bally (1913), Sapir (1911, 1921), Havers (1931), Tauli (1958, 1966), Kuryłowicz (1964), Benveniste (1968), Hodge (1970), Givón (1971, 1979), and Lehmann (1982/2015). The first uses of the terms ‘cycle’ (in 1877) and ‘grammaticalization’ (in 1912) appear in this period. The twenty-first century witnesses an interest in cycles for theoretical reasons, for instance, in the works of van Gelderen (2011), Willis et al. (2013: 7, 21, 169), Breitbarth (2014), Chatzopoulou (2019), Breitbarth et al. (2020), and Bouzouita et al. (2020). The insights here include (a) the presence of back-and-forth tendencies between comfort and clarity, that is, grammaticalization and renewal, in cyclical change; (b) the unidirectionality of the changes; and (c) explanations for these based in typology and the principles of the faculty of language. DOI: 10.4324/9781003272564-2

A history of cyclical change

27

Eighteenth and early nineteenth century precursors to grammaticalization and the cycle are discussed in Section 1. This work on the relationship between languages and the regular changes that they undergo is instigated by questions such as which language is the original language. Nevertheless, the work forms the beginning of the systematic study of language change. In Section 2, I proceed to the late nineteenth and early twentieth centuries when more data on ancient and lesser-known languages become available. Sections 3 to 5 examine modern approaches starting in the mid twentieth century when Tauli, Hodge, and Givón add data and cyclical change acquires theoretical significance. I mention the modern approaches for completeness in this historical chapter but will cover more of these in Chapter 6. This chapter can, of course, only provide a cursory overview of the history of grammaticalization and the cycle. The sketch here is based on some of my own readings of primary sources, helped by the works of Pedersen (1931), Chomsky (1966), Robins (1967), Winfred Lehmann (1967), Coseriu (1972), Said (1978), Schwegler (1990), Christian Lehmann (1982/1995), Nerlich (1990), Hopper and Traugott (2003: Chapter 2), and McElvenny (2016). Chapters 3 to 5 focus on describing certain cycles, and Chapter 6 provides more overview in terms of explanations for the cycle without providing historical depth. 1

The eighteenth and early nineteenth centuries

The seventeenth century sees an interest in the pre-Babel language or the perfect one; for example, Stiernhielm’s (1671) De Linguarum Origine Præfatio [Preface on the Origin of Language] examines the origins of languages and argues, as others had done with other languages, that Swedish was the language of logic, spoken before the fall of the Tower of Babel (see Eco 1995: 97). Though that consideration no longer seems relevant, Stiernhielm’s discussion of genetic relationships helps set the stage for the fascination with language change and the processes behind it. However, as Pedersen (1931: 9) puts it, it is not until “the eighteenth century [that] we meet with a new spirit. . . . For the first time, real inquiry starts.” Relevant work in the eighteenth century includes that of de Condillac (1746), Smith (1759), Leibniz (1768), Jones (1807), and Horne Tooke (1786–1805) (see Robins 1967: 150–159). This time period introduces the basic ingredients necessary to formulate views on a linguistic cycle, in particular the notion of grammaticalization – although that term is not used yet – and an acknowledgment of the change from analytic to synthetic and vice versa. I will provide some quotes from these works to show that (a) they recognize grammaticalization, if not by name, and (b) often see an improvement or decay in the changes, for example, that they see change as linear, not cyclical. Many of the motivations behind their work are currently less relevant, such as the concern for the original or perfect language.

28

A history of cyclical change

De Condillac (1746) is captivated by the origin of language, which, according to him, arises out of gestures and cries. Relevant to later debates on grammaticalization is his notion that abstract vocabulary results from concrete words and that the noun is the basis, as in (1). (1) La langue fut longtemps sans avoir d’autres mots que les noms qu’on avait donnés aux objets sensibles, tels que ceux d’arbre, fruit, eau, feu, et autres dont on avait plus souvent occasion de parler. (De Condillac 1746, Part 2, Chapter 9) [‘Language went a long time without having anything else than the names which had been given to sensible objects, such as those of tree, fruit, water, fire, and others which they had more occasion to talk about.’] According to Lehmann (2015: 1), de Condillac explains the “personal endings of the verb through agglutination of personal pronouns,” something we will study as the Pronoun Cycle in Chapter 5. However, I have only been able to find the quotes in (2); the first one of which says that verbs are originally not inflected and thereby suggests that endings were later additions, and the second quote, which suggests endings were added. There is no talk about the source of the endings, however. (2)

a. Les verbes, dans leur origine, n’exprimaient l’état des choses que d’une manière indéterminée. Tels sont les infinitifs aller, agir. (De Condillac 1746, Part 2, Chapter 9, §85) [‘Verbs, in their origin, only expressed the state of things in an indeterminate way. Such are the infinitives, go, act.’] b. Quand on se fut fait des verbes, on remarqua facilement que le mot qu’on leur avait ajouté pour en distinguer la personne, le nombre, le temps et le mode, avait encore la propriété de les lier avec le nom qui les régissait. [‘When verbs were made, we noticed quickly that the word which had been added to them to distinguish the person, the number, the time and mode still had the property of linking them with the noun which ruled them’]. (De Condillac 1746, Part 2, Chapter 9, §94)

Smith (1759 [1869]: 312) acknowledges a change from synthetic to analytic when he writes that “[i]t is, perhaps, worth while to observe that those prepositions, which in modern languages hold the place of the ancient cases, are, of all others, the most general, and abstract, and metaphysical; and of consequence, would probably be the last invented.” Like many in that time period, he also makes value judgments on the direction of the change, as when he states that the “simplification of the rudiments of languages renders them more and more imperfective” (1759 [1869]: 323). This means an argument for a uni-directional approach.

A history of cyclical change

29

Leibniz had an excellent knowledge of Sanskrit, Hebrew, Chinese, and spoke various other languages. He is concerned with relationships between languages and families (1768) and, like Stiernhielm, with finding an original language or the impossibility of it. He is a proponent of etymology in the modern sense, that is, looking for cognates on the basis of sound and meaning. He also has an early interest in river names, which shows, among other matters, how wide the geographical area of the Basques was. His philosophical outlook is Platonic: we know the principles of language subconsciously. Likewise, Jones (1807) notices similarities among ancient languages. He is, of course, very well-known for seeing the connection between Sanskrit, Greek, and Latin, which gives a real impetus to work on comparative and historical linguistics. Jones in his famous characterization of Sanskrit as having “a wonderful structure; more perfect than the Greek, more copious than the Latin” (1807: 34) also suggests a unidirectional movement towards reduced ‘perfection.’ Horne Tooke (1786–1805) takes the noun and the verb as basic and derives all other categories from it, for example, conjunctions, adverbs, and prepositions derive from nouns and verbs (Horne Tooke 1786: 24, 33; see also Robins 1967: 156–157). He regards inflectional elements as deriving from full verbs, but as Robins notes, only some of his identifications are correct. Work in the early nineteenth century includes that of scholars such as Franz Bopp, Friedrich von Schlegel, and Wilhelm von Humboldt. Bopp (1816, 1833) provides a useful overview of how abstract, grammatical vocabulary develops from earlier concrete vocabulary. Bopp studies the grammatical structure of the related languages and the origin of their grammatical forms. In (3), talking about the origin of case affixes, he argues that they arise from earlier independent words and provides many examples of a phenomenon that may be described as a cycle, as expressed in (4), taken from the English version. (3) Ihrem Ursprunge nach sind sie, wenigstens gröfstentheils, Pronomina. [In origin, they are for the most part pronouns.] (Bopp 1833: 248) (4) As also in verbs the personal terminations, i. e. the pronominal suffixes – although, in the course of time, they are no longer recognised and felt to be that which, by their demonstrable origin, they imply and are – are replaced, or, if we may use the expression, commented on by the isolated pronouns prefixed to the verb; so, in the more sunken, insensible state of the language, the spiritually dead case-terminations are, in their signification of space, replaced, supported, or explained by prepositions, and in their personal signification by the article. (Bopp 1885: 128, bolding mine) I read (4) to mean (a) that verbal endings originate in pronouns to be renewed by isolated pronouns and (b) that (synthetic) case is replaced by (analytic) prepositions and articles.

30

A history of cyclical change

von Schlegel’s (1808) Über die Sprache und Weisheit der Indier compares Sanskrit, Latin, and Greek, noting many similarities that will later be reformulated as the First Consonant Shift (e.g. [p] > [f]). He describes changes from ancient synthetic languages to modern analytic ones and uses value judgments to describe languages, with synthetic being the highest stage to be reached. Von Humboldt (1822, 1827–9) rejects the value judgments connected with the stages as well as von Schlegel’s definitions of analytic and synthetic. He does think the stages reach an endpoint, that is, are noncyclical. Von Humboldt (1836) emphasizes the creative ability of human language and the human mind (see Robins 1967: 174). Humboldt was also essential in adopting the three-way distinction among isolating, agglutinative, and flexional languages. In short, in examining relationships between languages and regular change, this period prepares the way for later work on grammaticalization and cyclical change. Although some concern is with the original and most perfect language, systematic changes and macro cycles are implicit. 2

The late nineteenth and early twentieth century

The late nineteenth and early twentieth century period continues to study grammaticalization and cyclical change and in fact uses the terms ‘cycle’and ‘grammaticalization’ for the first time. I’ll discuss some of the main ideas chronologically. Curtius (1870, 1885) may be the first one to use the term Bequemlichkeit (‘comfort’) as a reason behind phonological change (see Melchert 1975) and phonetic Verwitterung (‘weathering’). In his work, there is a tension “brought about by the tendency to greater comfort in articulation” and “the need to preserve the distinctness of meaning-bearing elements in words,” according to McElvenny (2016). The latter notes that Whitney (1875: 53) similarly uses “tendency of economy” and continues [w]hat Whitney has in mind here has as much affinity to modern grammaticalization theory as anything Gabelentz had to say. Whitney’s illustrations of this process include such classic examples of present-day grammaticalization as the English suffix -ly and its Romance counterpart -ment(e), and the Romance future as represented by French donner-ai. (Whitney 1875: 122–124) Whitney’s own words appear in (5), and they are indeed prescient to those of von der Gabelentz, as quoted in Chapter 1. Talking about changes in English, he says that it (5)

has so nearly stripped its root-syllables of all the apparatus of suffixes with which they were formerly clothed, and left them monosyllabic. All this has come about mainly through the instrumentality of the tendency to ease

A history of cyclical change

31

and abbreviation. . . . [However] it does not lose what it once possessed in the way of inflectional apparatus without providing some other and on the whole equivalent means of expression. (Whitney 1875: 105–107) The words ‘cycle’ or ‘spiral’ are not mentioned by Whitney, however. As I noted in the introductory chapter, the first mention of a ‘linguistic cycle’ in English, German (‘Sprachzyklus’), or French (‘cycle linguistique’) – the most likely languages at that time – is from Lefèvre (1877) in quotes provided in a bit of detail in (6) and (7). Chapter 8 of his book is titled “Le cycle du langage.” (6)

Il y a un cycle du langage, une évolution naturelle, où le développement de la pensée a sa part d’influence, mais qui se déroule manifestement selon des lois à peu près fatales. Le langage s’élève du rudimentaire au complexe, et du complexe au simple; au premier éveil de la curiosité humaine ont répondu les raciness démonstratives, sorte de geste vocal qui désignait le lieu plus que le caractère des objects; puis les raciness verbales (ou nominales), en très petit nombre, sont venues définir les choses par quelqu’une de leur qualités ou actions apparentes; de la juxtaposition ou agglomération de ces deux sortes de signes sont nées les parties du discours, d’abord vaguement accusées par l’indépendance mobile des racines accolées. (Lefèvre 1877: 143, my bolding) [There is a cycle of language, a natural evolution, where the development of thought has its share of influence, but which obviously takes place according to almost fatal laws. Language rises from rudimentary to complex, and from complex to simple; to the first awakening of human curiosity the demonstrative roots responded, a sort of vocal gesture which designated the place more than the character of the objects; then the verbal (or nominal) roots, in very small number, came to define things by some of their apparent qualities or actions; from the juxtaposition or agglomeration of these two kinds of signs are born the parts of speech, at first vaguely marked by the mobile independence of the attached roots.]

(7)

Le cycle linguistique est dominé et dirigé par l’instinct et le besoin de la clarté. La penseé, après avoir été dégrossie, allégeé et enrichie par le langage, veut trouver en lui un instrument docile et précis; il n’est plus utile que le mot, en répondant à son appel, vienne miroiter devant elle et se vanter de ses métamorphoses; il faut que, sans laisser un doute à l’interprétation, il serre de près la substance des objects, don’t il n’exprimait dans le principe que les qualités vagues. L’état analytique, où les mots sont à la fois démarqués et spécialisés, est donc pour nous le terme dernier et le but d’évolution linguistique. (Lefèvre 1877: 151, my bolding)

32

A history of cyclical change

[The language cycle is dominated and directed by instinct and the need for clarity. Thought, after having been roughened, lightened and enriched by language, wants to find in it a docile and precise instrument; it is no longer useful that the word, in responding to her call, shimmers before her and boasts of her metamorphoses; it is necessary that, without leaving any doubt in interpretation, it closely squeezes the substance of the objects, in principle it only expresses the vague qualities. The analytical state, where words are both demarcated and specialized, is therefore for us the final term and the goal of linguistic evolution.] Interesting trends (marked in bold) in these quotes are the following: (a) there is a linguistic cycle, (b) there is a natural development in language that is lawbound in (6), (c) languages change from analytic to synthetic and back (“simple to complex”) in (6), (d) there is a balance between comfort and clarity (“instinct and clarity”) in (7), and (e) change is goal-oriented in (7) although that seemingly conflicts with (c) and goes against the notion of cyclicity. In 1893, Lefèvre uses the term twice, first in a general sense, as in (8), and then in a comparative way in (9). Here too, there is a contradiction in that unidirectionality from synthetic to analytic seems implied, one stage being deemed better as well. (8)

(9)

le linguist remonte, de la phase analytique moderne à l’étage flexionnel, puis à l’agglunination, et au monosyllabisme; et, selon toute vraisemblance, il y a bien là un cycle de langage. (Lefèvre 1893: 15, my bolding) [the linguist follows back the language from the modern analytical stage to the inflected stage and then to the agglutinative and monosyllabic ones; and in all likelihood, there appears a linguistic cycle.] De ces deux langues, l’anglais et le français, l’une a l’autre dégageés, dans une mesure suffisante, de tout embarras grammatical, l’une a l’autre arriveés au terme du cycle linguistique, laquelle emportera le prix? (Lefèvre 1893: 301, my bolding) [Of these two languages, English or French, both sufficiently disengaged from the embarrassments of grammar and both having arrived at the term of the linguistic cycle, who shall get the prize?]

Lefèvre’s book is translated into English a year later, and the terms used there are ‘complete cycle’ and ‘linguistic cycle,’ respectively (Lefèvre 1894: 20; 424). Another early use of the term is by Lefèvre’s contemporary, Henry (1883: 56), who discusses a “cycle d’évolution” (as quoted in Klippi 2004: 184) in the context of where analogy takes place. As is well-known, the early twentieth century witnesses the first use of the term ‘grammaticalization’ in Meillet (1912: 148). However, as we’ve seen, the ideas were in the air, as McElvenny (2016) also observes in the quote in (10).

A history of cyclical change

33

(10) Expressivité was a central concept in the contemporary stylistic work of such figures as Charles Bally (1865–1947), which drives speakers to shun old and hackneyed linguistic expressions and promotes the creation of new, more vivid forms. Bally (1965[1913]: 41–43) also illustrated one of his discussions of expressivité with those now familiar examples of grammaticalization in the Romance languages, the cyclical development of the periphrastic future and the emergence of the -ment(e) suffix. While Meillet was not uncritical of Bally’s work in stylistics and other related efforts of the time (see e.g. Meillet 1910, 1913, 1926), he cites, in a footnote later added to his classic grammaticalization paper (Meillet 1912: 148), Bally and Leo Spitzer (1887–1960) for the light they cast on the ‘rôle du sentiment dans la création des formes grammaticales’ (‘role of sentiment in the creation of grammatical forms’). Among other major achievements of this time period, Edward Sapir is essential for suggesting a language typology and for a way to measure it. He writes that “[a]ll languages differ from one another but certain ones differ far more than others” and that this “is tantamount to saying that it is possible to group them into morphological types” (Sapir 1921: 121). Methods for calculating the analyticity or syntheticity of a language is something we come back to in Chapter 5 (though not Sapir’s classification). Relevant to our discussion is his formulation of language change as drift: “The drift of a language is constituted by the unconscious selection on the part of its speakers of those individual variations that are cumulative in some special direction” (1921: 155). The example of drift he gives is of the loss of case in current English whom. It is a drift because it conforms with the loss of other cases in its earlier history. He ends the chapter on drift with the tendency in English to incorporate loans as compensating “for something that was weakening within” (1921: 170). Thus, although Sapir doesn’t recognize cyclical change per se, drift is an instance of unidirectional change, and compensation is a type of renewal. However, as part of his early structuralist viewpoint, he is not interested in diachrony, and that remains the case for a number of years in the US. Later American structuralists, in particular Bloomfield and his followers, were not interested in diachrony either. That includes later practitioners, such as the founder of the generativist framework, Noam Chomsky. A last person to mention in this period is Havers (1931), who references a cycle in (11) that is unique in many ways. (11) Wenn man von einem Kreis- oder Spirallauf in der Sprachgeschichte spricht, so meint man damit die Tatsache, daß jüngere Generationen oft wider auf den Sprachgebrauch längst vergangener Sprachperioden verfallen. (Havers 1931: 6, my bolding) [When one speaks of a cyclical or spiral course in the history of language, one means the fact that younger generations often fall back on the language used in long past language periods.]

34

A history of cyclical change

His examples show a change from synthetic to analytic, so I interpret his mention of younger generations in (11) as their renewing within the synthetic/analytic macro cycle. Apart from this reference, there is not much on ‘Kreis/Spiral’ in his book. Later on, he mentions an analytic “Tendenz” (p. 157), as part of the driving forces (“Triebkräfte”) that keep language changing. Among the tendencies he describes are clarity (“Anschaulichkeit”), emotional discharge (“emotionale Entladung), economy (“Kraftersparnis”), and striving towards beauty (Streben nach Schönheit”). All of these terms are used without reference to other literature, as if the ideas were so much in the air so as not to need citation. In short, this section has shown that comfort, erosion, ease, and economy all appear around the 1870s, just before the first use of the linguistic cycle and grammaticalization. What is odd is that the early use, such as in Lefèvre, is not always consistently cyclical. 3

The mid and late twentieth century

Partly as a result of the synchronic bend of structuralist and generative frameworks, there is a period of “amnesia” about grammaticalization, as Lehmann (1982/2015) puts it, in the early to mid twentieth century. Thanks to scholars like Valter Tauli, Jerzy Kuryłowicz, Emile Benveniste, Carlton Hodge, and Tom Givón, the research increases again and is picked up by Christian Lehmann, Bernd Heine, Elizabeth Traugott, and others. Grammaticalization research in turn leads to work on cyclical change. Tauli (1958: 50–51) avoids the term the ‘linguistic cycle,’ using ‘tendencies’ instead. He distinguishes five principles, as shown in (12), the first two being very similar to those in von der Gabelentz (and Jespersen). The last three involve conscious choices by the speaker, so they are subsumed under ‘clarity’ in other definitions. (12) If we presuppose that the human mind underlies all linguistic change, then we may assume that the following major impulsive forces collaborate in structural changes: (1) tendency towards clarity, (2) tendency towards ease or economy of effort, (3) emotional impulses, (4) aesthetic tendencies, (5) social impulses. (Tauli 1958: 50) The first tendency in (12) involves expressiveness and avoidance of homonyms by creating grammatical categories and is behind the rise of periphrastic expressions and “phonologically stronger morphemes.” The second tendency toward ease entails a preference for shorter expressions as well as a preference for the reduction of grammatical categories and variability. The third involves emotional expressivity, as in to impress someone or to be expressive. The fourth

A history of cyclical change

35

favors irregularity and variability of expression, as in the use of a variety of syntactic constructions. The fifth is concerned with rules of politeness and in particular with the use of personal pronouns, something I come back to in Chapter 5. The value of Tauli’s work is in the many examples of change that he includes in his work, in particular, when he discusses analytic and synthetic tendencies in the loss of independent pronouns and the rise of case markers. Kuryłowicz (1964) and Benveniste (1968) are both important for reviving grammaticalization after Meillet’s earlier coining of the phrase, and they are mentioned here because their work is relevant to the cycle. Kuryłowicz, as mentioned in Chapter 1, clarifies what grammaticalization does: it not only describes the change from lexical to grammatical but from grammatical to more grammatical. He also provides many examples of the grammaticalization of adpositions. Benveniste adds many examples, such as the well-known auxiliary renewal of the possessive habere ‘have’ in Latin to future in French. Hodge (1970) presents a short article on the history of Egyptian, which I have mentioned in the previous chapter. His article marks the first major use of the linguistic cycle in 50 years or so. The lasting value of the work is the description of a complete cycle, a macro cycle, from analytic to synthetic to analytic to synthetic. He also mentions the possibility of such a cycle in Chinese, something I come back to when discussing macro cycles. In wide-ranging research on many languages, Givón has shown grammaticalization and renewal processes. For instance, Givón (1971) famously argues that yesterday’s syntax is today’s morphology, and Givón (1976) formulates an Agreement Cycle, without giving it this title, and returns to that subject in Givón (2016a). In Givón (1979 [2018: 198], he plots the “lifecycle of grammatical constructions” as a curve with three main phases. However, he spends less than a page on describing these stages. The first phase is one where the new construction is very similar to the old. The example he gives is of the Hittite relative that uses a conjunction similar to that used in other clause-chaining constructions. It loses this conjunction in stage two. Stage three is one where the construction becomes irregular and ceases to be transparent through phonological erosion, which makes the construction ripe for renewal. This formulation is clearly one of a micro cycle and is indeterminate as to where the change starts: what made the relative use the conjunction in the first place? This period sees the revitalization of work on grammaticalization with the work by Lehmann (1982 [2015]), Heine and Reh (1984), Traugott and Heine (1991), and Hopper and Traugott (1993 [2003]). Perhaps the most data-rich in this group is Traugott and Heine’s two-volume edited work, which provides both a theoretical underpinning and case studies on a variety of languages and topics. They are relatively well-known and do not focus on renewal, so I won’t dwell on them much.

36

A history of cyclical change

The grammaticalization studies are instrumental in doing the legwork for insights into the cycle. Lehmann (1982 [2015]: 22) makes a clear distinction between the two diachronic relations “y continues x” and “y replaces x.” Within a grammaticalization scale, the relation “y continues x” is equivalent to the relation “x is grammaticalized to y.” However, the relation “y replaces x” is neither a relation of grammaticalization nor of degrammaticalization. We shall call it, with Meillet’s (1915/1916) “renouvellement” in mind, the relation of renovation, also called renewal in the literature. Although Lehmann characterizes a cycle by describing the two processes typical of a cycle, he doesn’t use the actual term, except in reference to people like Meillet or Hodge. In conclusion, the mid and late twentieth century see a reemergence of grammaticalization, which is a necessary step in looking into cycles. However, there is minimal mention of actual cycles. 4

Recent generative work

In this section, I sketch why generative grammar was initially reluctant to work on grammaticalization and on (cyclical) change. I then outline how, ironically, this framework currently embraces the study of grammaticalization and that of cycles more than other frameworks, see for example Willis et al. (2013), Breitbarth et al. (2020), and Bouzouita et al. (2020). Generative grammar has its beginnings in the late 1950s with the work of Noam Chomsky emphasizing the innate linguistic knowledge. It focuses on the internalized system that allows native speakers to form and understand grammatical sentences. As a result of the emphasis on native speaker intuitions and the lack of those speakers for earlier periods, mainstream generative grammar has not generally used insights from historical linguistics except in work by Elizabeth Traugott, Paul Kiparsky, and David Lightfoot, to name some of the early generative historical linguists. In this tradition, language change is seen as a setting of choices by the child acquiring a language based on the language it is exposed to. A major argument against using grammaticalization is the unidirectionality assumed by most linguists working on grammaticalization. If a learner just reacts to a certain input, there should not be unidirectional changes, according to Newmeyer (1998), Lightfoot (2006), and others. Newmeyer (1998: 226) exclaims that “there is no such thing as grammaticalization” (italics left out) and that grammaticalization is an epiphenomenon. Klima (1965: 83) formulates a model of generative language change emphasizing the discontinuous nature of change and reanalysis by the learner. (Since generative grammar takes a purely synchronic approach, there is no reanalysis in

A history of cyclical change

37

the strict sense. A child acquires a language based on the available data and does not reanalyze. It is just a convenient term.) I reproduce it as Figure 2.1 based on Andersen (1973). It includes changes that adults can make (minor innovations) and that are then the input for the new generation. Figure 2.1 expresses that the exposure to a language triggers a grammar in an individual that is referred to as I(nternal)-Language, which, in turn, produces an E(xternal)-Language. What this model shows is that language acquisition is discontinuous: the I-language (or grammar) of one generation may differ from that of the next. Working in a generative model, Roberts and Roussou (2003) and van Gelderen (2004a, 2011) look at the systematic change presented by grammaticalization and formulate principles of Universal Grammar that are assumed to be part of the innate starting point of the child. They argue that grammaticalization (and cycles for van Gelderen 2011) provide a window on the faculty of language. Some of the economy principles formulated this way are provided in (13). (13) Economy Principles a. Head Preference Principle (HPP): Be a head, rather than a phrase. b. Late Merge Principle (LMP)/Upwards Reanalysis: Merge as late as possible. c. Feature Economy: Minimize the semantic and interpretable features in the derivation. As will be explained in more detail in Chapter 6, these principles account for micro cycles like the Negative and Determiner Cycles and for macro cycles such as the Pronoun and Case Cycle.

FIGURE 2.1

Generation n

Generation n+1

Universal Grammar (UG) + Exposure to language (input) = I-Language

Universal Grammar (UG) + Exposure to language (input) = I-Language

E-Language + minor innovations

E-Language

Model of language acquisition

(based on Andersen 1973)

38

A history of cyclical change

Evidence from language acquisition in, for instance, the work of Cournane (2014, 2017) shows that “children make structural mapping errors that, if left unchecked, are compatible with the innovations we see in the historical record” (2014: 103), that is, that (13b) is indeed at work. Since the mid 2000s, generative grammar has undergone a paradigm shift from its early emphasis on Universal Grammar to a focus on factors not specific to the faculty of language, for example, Chomsky (2013, 2015). If the faculty of language developed in humans only 100,000 to 200,000 years ago – as is speculated – it makes sense to attribute less to language-specific principles. “UG must meet a criterion of evolvability” (Chomsky et al. 2019: 230). If there were principles specific to language, they wouldn’t have had enough time to evolve, and if there had been time, the changes would possibly have made languages very different from each other. The more general factors are known as third factor principles and include principles of efficient computation, as in (14). (14) Principles not specific to the faculty of language. (a) Principles of data analysis that might be used in language acquisition and other domains; (b) principles of structural architecture and developmental constraints that enter into canalization, organic form, and action over a wide range, including principles of efficient computation, which would be expected to be of particular significance for computational systems such as language (Chomsky 2005: 6). Although third factor in nature, structural and featural economy, as in (13), still depend on linguistic information, and if less emphasis is placed on these, they should be reformulated as nonlinguistic, that is, genuine third factor principles. In Chapter 6, I provide such an account for micro and macro cycles. In particular, I will argue that the third factors of minimal search and determinacy also provide insight into the systematic change presented by cycles. 5

Recent non-generative work

In this section, I discuss functionalist and Construction Grammar approaches to the cycle. I come back to these in Chapters 4 and 6, respectively. The current chapter provides an overview of who works on the cycle, but later chapters provide more in-depth accounts. Functionalism in linguistics foregrounds the role of language in communication and takes language and linguistic change as shaped by communicative needs of the hearer and the speaker. There are many varieties of functional linguistics. Many take as their intellectual origin the Prague School, which looks

A history of cyclical change

39

at how various levels (syntactic, semantic, and pragmatic) interact. Well-known functional principles are given in (15). (15) a. Old information goes before new information. b. Structurally light information goes before heavy. André Martinet, Simon Dik, and Michael Halliday are well-known theorists of this framework. According to the former, “la communication reste la function primaire et centrale du langue” [Communication remains the primary and central function of language] (Martinet 1955: 41). As seen previously, functionalist approaches, such as Heine et al. (1984) and Hopper and Traugott (2003), look into grammaticalization. Recently, functionalist approaches have turned to cycles as well. For instance, Hansen (2020: 167) presents an edited special issue on Pragmatic Cycles, to which I return in Chapter 4. She and others focus on cycles where the driving force is functional, that is, pragmatic. Assuming the definition of the cycle, as in von der Gabelentz, pragmatics is indeed relevant but part and parcel of the cycle, not a special type. Construction Grammar has, from its start, been concerned with language change, in particular with syntactic change. Construction Grammar is a network of form-meaning pairs that are called a ‘construction,’ and this knowledge is usage-based, not innate, as in the generative framework. That pairing of form and meaning is idiosyncratic and unpredictable. Hilpert (2021: 38–41) explains how its diachronic approach differs from other cognitivefunctionalist frameworks. Construction Grammar capitalizes on the notion of the unruliness and unpredictability of change, and that renders it very different from, for instance, grammaticalization theory, which makes testable predictions. What are the advantages of Construction Grammar? Construction Grammar deals “with many changes that are in fact bidirectional” (Hilpert 2021: 43) so that a noun can be used as a verb and a verb as a noun. It is therefore comfortable with degrammaticalization, such as the use of ish as an independent word, derived from an affix. Construction Grammar is interested in how both the form and function of a construction change. Since it is usage-based, Construction Grammar makes copious use of corpora and examines the competition between constructions, for example, between an -ing clause and a to- infinitive. Barðdal and Gildea (2015: 37) confirm that “constructions high in type frequency attract members from constructions lower in type frequency,” which, of course, begs the question what made one construction (e.g. an -ing complement) more frequent than another (e.g. a to infinitive). In Chapter 6, I discuss three Construction Grammar approaches that consider grammaticalization, Traugott (2008), Sommerer (2015), and Culicover (2021), but neither of these is really concerned with cyclical change.

40

A history of cyclical change

6

Conclusions

This chapter has reviewed the history of the linguistic cycle. The first mention of the phrase is in 1877, as far as I know. Foundational precursors before 1877 and relevant work after 1877 up to the present were also discussed. I divided this history into four stages, one that occurs before the linguistic cycle is used, a second stage in which the cycle and grammaticalization are introduced, a third stage that features the reintroduction of the cycle and more work on grammaticalization, and the current period with more theoretical interest in the cycle. Although in the earliest period the cycles are sometimes formulated in contradictory ways and involve few actual examples, it is clear that cyclical change has been on the minds of linguists for hundreds of years. Suggestions for further reading On the history of linguistics before the twentieth century: Robins (1967) and Petersen (1931 [1962]). During and after the twentieth century: Newmeyer (1998) and Harris (2021).

Review questions

1. When are the first discussions of the linguistic cycle? Do they involve the macro or micro cycle or both? 2. Are you familiar with a specific theoretical framework? How does a framework influence the types of data a linguist considers and the questions they ask? 3. I used Google Books to search for the earliest instance of the ‘linguistic cycle.’ Are there other ways you could search? Advanced exercise

Find information where the language you selected in Chapter 1 is spoken, who speaks it, and what the status of the language (safe, endangered, dormant) is. Are there grammars of one sister language or even an older stage of the language?

3 MICRO CYCLES Determiner and verbal cycles

In this chapter and the next, I describe the Determiner, Copula, Tense, Aspect, Mood (TAM), Voice, Negative, Interrogative, Complementizer, and Pragmatic Cycles. These are micro cycles in that they affect subparts of a grammar and don’t shift the major typology of a language. For instance, in analytic languages, negatives and determiners can be synthetic, and in turn, in synthetic languages, negatives and determiners can be analytic. In fact, a grammatical category can be in different stages of a micro cycle regardless of the prevailing language type. The choice of micro cycles in this chapter includes the Determiner, Copula, TAM, and Voice Cycles. I start with the Determiner Cycle because it takes place inside the nominal group, whereas the other cycles in this chapter and the next involve the clause. The Copula, TAM, and Voice Cycles are grouped together because they consist of verbal categories, and copulas are sources for TAM and voice. Some of the data on the Determiner and Copula Cycles can also be found in van Gelderen (2011) and that of the Perfective, Imperfective, and Voice Cycles in van Gelderen (2018). Micro cyclical changes are interesting from the point of view of the morphosyntactic features that are found universally. Nouns need definiteness or case, or aspect marked on the verb (see Leiss 2000), and verbs need tense, mood, aspect, voice, and pronouns or agreement (or all of these). Copulas are not generally needed, unless they are used to express a certain tense or aspect, but can easily be lost, and their renewal is frequent. Cycles give us insight in the features that are needed and how they are expressed. The outline of this chapter is as follows. In Section 1, I restate the definition of micro cycles from Chapter 1 and provide a bit more background. Sections 2 DOI: 10.4324/9781003272564-3

42

Micro cycles

through 6 discuss examples of micro cycles, namely, cycles involving determiners, copulas, TAM, and voice markers. Section 7 is a conclusion. 1

Defnition of a micro cycle

As mentioned in Chapter 1 and repeated in (1), micro cycles involve individual changes in a lexical word, for example, in a negative or a modal, or changes in the subparts of the grammar, such as when a synthetic TMA system is replaced by a periphrastic one. These changes do not affect the basic typology of a language. (1)

A micro cycle involves (a) (b)

individual changes where a lexical word or phrase grammaticalizes as a grammatical morpheme or a grammatical word as a more grammatical expression and is renewed by another lexical expression or cyclical changes in a subpart of the grammar involving grammaticalization and renewal.

The changes in (a) and (b) are sometimes hard to keep separate. Is it possible to think of the change from the motion verb going to to the future gonna as an individual change or as one connected to the appearance of other semi-modals, such as wanna and hafta? Fischer (2003: 20) doesn’t think it is an individual change and coins the term ‘Modality Cycle’ to describe the changes from a main verb to a deontic and epistemic modal and a later renewal of the deontic modals, such as must, by semi-modals like have to. It is hard to think of micro cycles as ever involving changes in just one word. Gergel’s (2009) description of the change from the adverb rather in Old English, as in (2a), to a modal auxiliary, as in (2b), is possibly an example of an individual change. (2)

a.

b.

Quirinus Þa eode to ðam cwarterne hraðe Quirinius then went to the prison quickly ‘Quirinius then quickly went to the prison’ (c1000, Ælfric Homilies 24.78; from Gergel 2009: 247). & saide þat þai wolde neuer faile Kyng Arture, and raþere to bene dede: ‘and said that they would never fail King Arthur and (would) rather be dead’ (c1250, Brut 3, 82.2486; from Gergel 2009: 258).

Although micro cycles involve subparts of a grammar, as was mentioned in the introduction to this chapter, they do not affect the basic typology of a language. Scandinavian languages are analytic in their lack of inflection on verbs (e.g. har

Micro cycles

43

in (3) remains the same regardless of the person and number of the subject), and their lack of case marking on nouns (e.g. potatis in (3) is the same in all functions). These languages show evidence of a Determiner Cycle, however, because determiners are cliticized, as in (3) from Swedish. (3)

Jag/du/hon har ätit potatis=en I/you/she have eaten potato=DEF.CM ‘I/you/she have/has eaten the potato.’

Swedish

Scandinavian has gender (e.g. common gender in (3)) and number on determiners, nouns, and adjectives so it can be said that the DP internal syntax is synthetic where the clausal syntax is analytic. Another example involves negation. A language family that is (poly)synthetic can show synthetic or analytic stages of a Negative Cycle in its daughter languages. Athabascan is such a family; in most of its daughter languages, the negatives are quite different even though the other morphemes are similar. For instance, in Lower Tanana (4a), the negative is formed with a -dh prefix and an -ęę suffix, whereas, in the closely related Upper Tanana (4b), the negative is an independent word. Note how the basic word for ‘ice/freeze’ is very similar: ten and tän, respectively. (4) a. tendh ghaaghetltenęę t -n -dh -gh -gh -es -ł FUT QUA NEG QUA QUA 1S CAUSE ‘I won’t freeze it solid’ (from Kari 1993: 55). b. k’aa tinak-tän NEG 1S.FUT-freeze ‘I won’t freeze it solid’ (from Kari 1993: 55).

Lower Tanana -ten -ęę ice NEG Upper Tanana

Thus, even though the languages in (4a) and (4b) are both synthetic, their negatives can be synthetic or analytic. This section has (re)introduced the micro cycle before going to actual examples of micro cycles. I start with the Determiner Cycle. 2

The determiner cycle

All languages are said to have demonstratives (Diessel 1999a), but not all have articles. In the World Atlas of Language Structures (WALS), 50% of the languages examined (308/620) have either a definite word or affix, and 11% (69/620) use demonstratives as articles, as shown in Table 3.1. That leaves 39% of the languages in this set (243/620) without an article.

44

Micro cycles

TABLE 3.1 Definite articles cross-linguistically in WALS

Definite word distinct from demonstrative Definite affix on noun Demonstrative word used as marker of definiteness No definite article but indefinite article Neither definite nor indefinite article Total:

216 92 69 45 198 620

(Dryer 2013a)

That articles derive from demonstratives has been known for a long time for Romance. For instance, Raynouard (1816: 41) remarks “[d]es savants français et étrangers ont souvent observé que l’article . . . dériva du pronom ILLE” [‘French and foreign scholars have often observed that the article . . . derives from the pronoun ILLE’]. Early on too, von Schlegel (1818: 28) writes “un pronom démonstratif quelconque se transforme en article” [‘any demonstrative pronoun transforms itself into an article’]. Greenberg (1978: 61 ff.) formulates the “cycle of the definite article” with three stages, as definite article, non-generic article, and noun marker. According to him, Amharic has gone through all stages (Greenberg 1978: 59). In this section, I mainly focus on a narrower cycle, namely, the change from a demonstrative to an article with the renewal of the deictic features of the original demonstrative. I will use the term ‘Determiner Cycle’ for this set of changes since demonstratives and articles are both determiners. A representation of this scenario is given in (5) for the typical changes, and actual examples appear later in this section. (5)

The Determiner Cycle involves a. demonstrative > article > affix > zero = grammaticalization b. demonstrative + locative > demonstrative ‘see’ + demonstrative > demonstrative = renewal

As demonstratives lose deictic force, they “are being constantly replaced by new demonstratives usually formed by older ones by the addition of new deictic elements, etc.” (Greenberg 1978: 77). In (5b), the first type of renewal is shown through a locative adverb that modifies the head noun. The deictic meaning is replaced by a non-deictic one, also referred to as psychological deixis in Johannessen (2006) or recognitional function in Diessel (1999a: 106). Another frequent renewal is through an imperative verb, with the meaning of ‘look’/‘see.’ This has occurred in the history of Romance and Germanic. For instance, the demonstrative this originated “by adding se, si (probably = Gothic sai ‘see, behold’) to the simple demonstrative” (OED s.v. this). The loss of the article is rare, but it

Micro cycles

45

occurred in the history of Aramaic (Gzella 2015: 15) as well as in some varieties of English, as I show later. Apart from Greenberg, Demonstrative Cycles have been discussed by a number of linguists, for instance, Harris (1977), Epstein (1995), Bernstein (1997), Lyons (1999), Millar (2000), Putnam (2006), Rupp (2007), Roehrs (2010), Rupp and Tagliamonte (2017), Vindenes (2018), Bernstein et al. (2021), and Rauth and Speyer (2021). I will rely on their insights and data in what follows. Three different stages of grammaticalization, as in (5a), are visible in the history of Romance. In (6), Latin ille is a demonstrative indicating location away from the speaker and addressee, which becomes the definite marker le in French (7), and a definite marking suffix in Romanian (8). (6) (7) (8)

ille liber ‘that book’ le soleil ‘the sun’ om-ul bun man-DEF good ‘the good man.’

Latin French Romanian

The renewal in (5b) is also exemplified in Romance. Since the demonstratives lose some locative/deictic function, they are reinforced in (later) Colloquial Latin as ecce iste ‘see this’ and ecce ille ‘see that’ (Greenberg 1978: 76; Giusti 2001: 169), which in turn grammaticalize to cest/cist and cel/cil, respectively, in Old French and become ce(tte)/ce(s) in Modern French, as in (9). Harris (1977: 256) claims the latter forms are articles since they do not mark distance, as indicated in the gloss in (11), and that “only forms with . . . -ci ‘here’ and -là ‘there’ [, as in (10),] can be regarded as demonstratives” (260). (9)

ce moment French DET moment ‘this/that moment.’ (10) mais ma femme elle vivait à ce moment-là encore French but my wife she lived at DET moment-there still ‘but my wife was still alive at that moment’ (CdES corpus). Schematically, the stages of grammaticalization and renewal are shown in (11). (11) a. Latin ille > b. Colloquial Latin ecce ille >

French le > Old French cell >

Romanian -ul Modern French ce/s

46

Micro cycles

Greenberg’s cycle has other stages, as mentioned. In Stage I, demonstratives become articles, as in (5a); in Stage II, these in turn reanalyze as non-generic markers; and then, Stage III, they become noun class markers before finally disappearing. The latter two shifts are part of a macro cycle that will be discussed in Chapter 5 as the Case Cycle. Here, I look at Stage I, that is (5a), and the renewal of the demonstrative, that is (5b). Lehmann (2015: 40ff) argues that demonstratives mark three components: deixis, definiteness, and a category, and these components are relevant in describing the cycle in (5). The first component to be lost in language change is typically the deictic element, the second one is the definite marking, and the third is whether it is an independent category (an NP) or a dependent one (a D). Lehmann provides numerous examples of a demonstrative losing deictic marking and becoming a definite marker with anaphoric function, which then loses definiteness. He doesn’t examine what replaces the erstwhile deictic and definiteness functions. In what follows, I look at both the typical grammaticalization of the demonstrative, as in (5a), and its renewal, as in (5b). This is what will be called the Determiner Cycle. In previous work (van Gelderen 2007a), I have called it the Definiteness Cycle, but this cycle is more about loss and renewal of deixis, in addition to definiteness, and therefore the term Determiner Cycle is more appropriate. I’ll first provide some examples and then focus on what is renewed and how this takes place. Apart from the changes in Romance, the reanalysis of a demonstrative to an article has occurred in Mayan, Salish, Egyptian, Niger-Congo (Greenberg 1978), Nilo-Saharan (Greenberg 1963), Korean (Sohn 1999), Germanic (Raidt 1993), and Uto-Aztecan (12), from Tohono O’odham (Shaul 1986). The clitic noun marker g= in Tohono O’odham goes back to an older demonstrative (Shaul 1986: 51) that still occurs in Tohono O’odham as a demonstrative, as in (12b), and as a third person pronoun. (12) a. b.

‘Ab ‘o hihim g=cewagi Tohono O’odham there AUX walk.P ART=clouds ‘Here come the clouds’ (Zepeda D’ac ‘O’odham 1982). Cipkan ‘o hegai ‘uwĭ work.PRT AUX that woman ‘That woman is/was working’ (Zepeda 1983: 8).

So Tohono O’odham has undergone the change in (5a), and the old demonstrative continues to be used. That means there has been a split with the old determiner keeping its deixis, which makes the renewal of the demonstrative not necessary. Another such example can be seen in the history of Basque. Martínez-Areta (2013: 294), among others, writes that “[t]he Basque definite article comes from

Micro cycles

47

the cliticization of the third grade demonstrative” (i.e. the most distal) but the conversion “did not cause this demonstrative to disappear.” The change is given in (13a), with the reconstructed form marked by an asterisk. Reconstructions can be done by comparing different varieties of Basque and on the basis of what makes most sense linguistically. Examples of the Modern Basque article and distal are provided in (13bc). (13) a. Modern Basque Proto- Basque etxe-a < *extxe har house-DEF house that yonder ‘the house’ ‘that house there’ b. gizon-ari Basque man-DAT.DEF ‘to the man.’ c. gizon hari man that.yonder.DAT ‘to that man yonder’ (Martínez-Areta 2013: 294). Examples of the shift in (5a) in English can be shown using the Peterborough Chronicle, written during the eleventh and twelfth centuries in eastern England (however, see Millar 2000 for more details). In (14), from the entry of the year 1130, demonstratives are used regularly before nouns (e.g. se is masculine singular nominative, seo feminine singular nominative, and Þæt neuter nominative). However, in (15), from the year 1137, reduced forms suddenly appear before nouns. Unlike earlier demonstratives, these cannot be used by themselves (cf. Wood 2003) and are therefore seen as articles. The change to (15) is due to a different scribe, who has internalized a new system. (14) Đes feorðe dæges þæræfter wæs seking Heanri on Roueceastre. & seburch forbernde ælmæst. & seærcebiscop Willelm halgede Sancti Andreas mynstre & ðaforsprecon biscop mid him. & sekyng Heanri ferde ouer sæ into Normandi on heruest. ‘On that fourth day thereafter, (that) King Henry was in Rochester, when that town was almost consumed by fire; and (that) Archbishop William consecrated St. Andrew’s monastery, and those aforesaid bishops with him. And (that) King Henry went over sea into Normandy in autumn.’ (Peterborough Chronicle, anno 1130) (15) ðis gære for þe king Stephne ofer sæ to Normandi & ther wes underfangen forþi ðæt hi uuenden ðæt he sculde ben alsuic alse the eom wes. ‘This year, (the) King Stephen crossed the sea to go to Normandy and was received there because they thought he was like the uncle’ (i.e. his uncle). (Peterborough Chronicle, anno 1137)

48

Micro cycles

The Ormulum is from the same geographic area as the Peterborough Chronicle but from a little later in the twelfth century. It too shows early articles, and it illustrates that, as the demonstratives in (14) are replaced by phonologically weaker articles in (15), the demonstrative itself continues. The Ormulum has more varied nouns, so in (16), the nouns are listed with the gender they have in Old English. In Old English, þe and þatt would only be expected with masculine and neuter nouns, respectively, and even though, in (16a), þe is used with a masculine noun and þatt with a neuter one, that is not the case in (16b). Here, the situation is the reverse, and þe is used with a neuter and þatt with a masculine noun. The use of þatt in (16) is therefore compatible with it continuing to be a demonstrative for all nouns regardless of gender. (16) a. Þe laferrd haffde litell rum. Inn all þatt The lord.M had little room in all that miccle riche. big kingdom.N ‘The lord had little room in that big kingdom’ (OED, c1175, Ormulum 8490). b. He toc þe recless & te blod He took the Incense.N and the blood.N & ȝede upp to þatt allterr. And went up to that altar.M ‘He took the incense and the blood and went to the altar’ (OED, c1175, Ormulum 1082). Once the demonstrative weakens, we expect a renewal and that occurs several centuries later. The first clear proximal and distal reinforcements are from the sixteenth century, as in (17). (17) a. b. c. d.

The best wyse that we may Hast vs outt of this here. ‘The best manner that we may hasten ourselves out of this here.’ (OED, c1500, Towneley Plays I. xv. 181) To Samaria and them partes. (OED, 1596 H. Clapham, Bible Hist. 92) The warres and weapons are now altered from them dayes. (OED, 1598, Barret Theor. Warres I. i. 4) On leaving yours and Mr. B.’s hospitable House, because of that there Affair. (OED, 1741, Richardson Pamela III. xxxviii. 404).

As Rupp and Tagliamonte (2017) argue, the early renewals are very deictic in nature, and it isn’t until the eighteenth century that non-deictic complex demonstratives are found. The stage where the article is reduced to zero in (5a) has not been reached in standard English but has happened in certain varieties,

Micro cycles

49

for example, York English (18a) and for certain speakers of American English in (18b). (18) a. b.

You know, Ø night life’s brilliant. (Rupp & Tagliamonte 2017: 82) That’s artificial intelligence for you – camera focusing on your dog while burglar gets in – let’s hope dog’s less stupid. (Twitter @sonjameier2, August 2022)

The corpus data don’t show much change in the frequency of ‘the’ and ‘that’ over the period from 1990 to 2019, as shown in Figures 3.1 and 3.2, respectively. COCA gives the total numbers in the top line; take a look at the bottom line for the normalized frequency, that is, per million words. Although the total frequencies don’t change much, we see that ‘the’ is less and ‘that’ more frequent in the spoken register. That suggests a change in progress towards the use of ‘that.’ Having given some examples of grammaticalization and renewal, I now turn to two other questions. As in the case of some other cycles, the question arises what starts this cycle, a weakened deictic D(eterminer Phrase), as in (5a), or a pragmatic need for extra emphasis, as in (5b). The second question is how the renewal takes place, that is, which is the ambiguous context such that reanalysis is possible. Such a context is also known as the bridging context. As just shown, in English, (5a) seems to lead the change because the renewal of (5b) is attested after the change from (14) to (15), as in (17ab). Likewise, Rupp and Tagliamonte (2017) find that the complex deictic demonstrative is

FIGURE 3.1

The article ‘the’ followed by a noun (downloaded 18 April 2022)

FIGURE 3.2

The demonstrative ‘that’ followed by a noun (downloaded 18 April 2022)

50

Micro cycles

not found in the Linguistic Atlas of Early Middle English (ranging from 1150 to 1325) as well as in similar corpora and this confirms a later origin. Vindenes (2018) finds that renewal drives two changes in various stages of Scandinavian, namely, the reinforcement of the demonstrative through a form of the verb sai ‘see,’ as in (19), and that by the locative adverbs. We’ve seen both scenarios already in other languages. stæinn sā-si, es vas nura goði. Old Norse (19) Æft Rōulf stændr after Hróðulfr stands stone this.S.M.NOM-see he was nura chief ‘In memory of Hróðulfr stands this stone; he was (nura)-chief.’ (Flemløse-sten, Vindenes 2018: 646–647)

Vindenes doesn’t provide empirical evidence as to whether (5a) or (5b) leads the change in (19). Instead, he argues tentatively that [s]ince it seems difficult to use such explanations [weakening first] without assuming some type of therapeutic change . . . a better approach might be to look for other factors, for instance in the discourse context, that could motivate the reinforcement. (647) Vindenes’ second case study, the reinforcement of demonstratives by deictic elements, again provides no empirical data as to which reason drives the change but merely asserts that “[n]ew demonstrative variants must arise before old ones disappear” (659). I find it quite plausible that emphatic reinforcement can drive the cycle, but empirical evidence is hard to come by. Another question is how the determiner is replaced. I will answer this by looking at a number of Germanic languages and varieties in which the demonstrative is renewed by an erstwhile sentence adverb. Here, it is easy to see ambiguous or bridging contexts. For instance, in (17a), the adverb there is ambiguous between being a sentence-final sentence adverb or being a reinforcer of the proximal this. In (17b), that there is already reanalyzed as a complex determiner because it is no longer ambiguous. This reinforcement of the demonstrative has not become part of the standard language because, in present-day standard English (but see Rupp and Tagliamonte 2017 for exceptional varieties), both the proximal adverb here and distal there are still in an ambiguous position, as (20) and (21) show, and could easily be input to reanalysis. (20) Don’t bring all this stuff here to the city. (COCA 2015 Magazine) (21) You see that guy there, he’s the biggest liar in Congress. (COCA 2019 Spoken)

Micro cycles

51

Adverbs can also precede the determiner, as in (22) to (24), and that could lead to another form of renewal, one we’ll see happened in Afrikaans. In these sentences, here and there are sentential locatives, but at some point, they could be reanalyzed as part of the nominal phrase. These are other bridging contexts, enabling reanalysis. (22) Well, there’s a chance here this thing ain’t running again. (COCA 2017 TV) (23) By the time Beck gets back here this thing will be over and done with. (COCA 2008 TV) (24) Yeah. throw it out there that way. (COCA 2015) The three possible positions for renewal are listed in Table 3.2 and will be elaborated on later (but note that Roehrs 2010 recognizes another position for Romance). All of these arise through ambiguity with a locative adverbial; the post-D has to grammaticalize as N-less construction first, as in (17a). The same renewal through a locative adverb occurs in Dutch (25), where the cooccurrence of the distal adverb daar and the demonstrative die is shown. However, the adverb can also come before the noun, as in (26) and (27), and result in ambiguity. (25) die man daar is mijn vader Dutch that man there is my father ‘That man there is my father’ (book title). (26) En moet je daar die man zien met dat fluitje. And must you there that man see with that whistle ‘And look at that man there with the whistle.’ (www.letras.com/don-quishocking/zonnig-mexico/) wie is hier die man met zoveel ervaring. (27) Ik dacht al I thought PRT who is here that man with so much experience ‘I was thinking who is that man with so much experience here.’ (www.onemorething.nl › . . . › Afreageerdraadje) In Afrikaans, the demonstrative from the source language, die, is used as the definite article and the former demonstrative is reinforced with a locative adverb dié; hierdie and daardie (or daai) are the demonstratives.

TABLE 3.2 Varied positions for renewal of deictic particles in Germanic

Pre-D (‘there that N’) Afrikaans (29), Dutch (26)

post-N (‘that N there’) French (10), English (21)

post-D (‘that there N’) English (17d)

52

Micro cycles

(28) Hierdie plaatjie laat jou ‘n gedetaileer boom van hierdie taal familie sien. Afrikaans This picture lets you a detailed tree of this language family see ‘This picture lets you see a detailed family tree of this language family.’ (http://home.unilang.org/main/families.php?l=af) The renewer occurs before the demonstrative in Afrikaans, and although we don’t have historical evidence, there could be ambiguous constructions where the adverb could be a sentential locative, as in Dutch (26) and (27). These ambiguous structures are not infrequent in Dutch, for instance, and could lead the way to renewal. (Standard) Dutch and English show incipient stages of change, but as mentioned, Afrikaans (28) and (29) show that the deictic adverb, both proximal and distal, modifying the entire DP is now part of the DP and the deictic nature has weakened. (29)

Daardie teenstrydighede was egter nie soseer in die man Bram Fisher nie

Afrikaans

There.those contradictions were however NEG so.much in the man Bram Fisher NEG

‘Those contradictions were however not so much in the man Bram Fisher not (but in . . .)’ (Mandela speech, 1997, www.anc.org.za/ancdocs/history/mandela/1997/ sp971128.html)

Other varieties of Germanic also show this renewing element, as (30) attests from German. Dutch and Belgian varieties show the same (cf. De Bont 1962; Pauwels 1959; Raith 1993), as in (31). (30) Hier die /hi:di:/

man leiden, German variety (Saarbrücken) Here that people can one guide aber da denen traut man nicht über den Weg but there those trusts one not across the road ‘One can guide the people over here but over there one doesn’t trust them to cross the road.’ (Raith 1993: 282)

(31) Hier die

vrouw

Leute

zal

kann

ons wel

de weg wijzen

Dutch variety (Aarschots)

here that woman will us PRT the direction show ‘This woman will show us the right direction’ (Pauwels 1959: 1).

What is the trigger for the innovation, for example, in Afrikaans? Raidt (1993: 285) writes that the Afrikaans forms in (28) and (29) “emerge suddenly around

Micro cycles

53

1880 without any noticeable period of transition.” So suddenly, the more Dutchlike dese and die are replaced by hierdie and daardie, respectively, and and die becomes restricted to the article. What sets the change in motion is unclear. In Dutch and English, there is no evidence of the change to complex demonstrative yet. Dutch and Afrikaans show pre-D and post-N positions for renewal and, for English (17d), we have seen the post-D possibility. The latter is also possible in a number of other Germanic varieties, as Putnam (2006), Roehrs (2010), and Rauth and Speyer (2021) have shown. In Section 2, we have examined the change of the demonstrative to an article and the renewal of the demonstrative through deictic features. A complete cycle has occurred in the history of Romance and of English where the demonstrative has grammaticalized to an article, and the former has been renewed. The evidence from English (and Tohono O’odham) suggests that the change in (5a) precedes that in (5b). For Romance, Scandinavian, and Germanic languages other than English, the evidence is not so clear. The section has also addressed different bridging contexts, one where the sentential adverb precedes and one where it follows the determiner. For instance, in Afrikaans, the former was taken and, in French, the latter. In languages such as English and Dutch, adverbs in either position could be come reanalyzed. 3

Copula cycles

Li and Thompson (1977) are among the first to examine the set of changes in copulas, and Katz (1996) is one of the first to note the systematic nature of these changes and to discuss the phenomenon as a cycle. Copula cycles occur in many typologically and genetically different languages: Turkish, Uto-Aztecan, Chinese, Hebrew, Arabic, Maltese, Kenya Luo, Lango, Logbara, Nuer, Wappo, West Greenlandic, Sranan, and Saramaccan. The complements to copulas typically describe identity, location, or quality. The sources for copulas depend on this meaning, that is, a third person pronoun will often become a copula of identity, a locative adverb or posture verb one of location. In turn, copulas themselves are a source for passive, focus, and progressive markers, to name just a few of the ones mentioned in Kuteva et al. (2019: 123–132), and languages may lack copulas. Like the Determiner Cycle, the Copula Cycle consists of two stages, as shown in (32). The grammaticalization stage typically has the copula attach to something else, and then the copula will get lost. It can be renewed in a number of ways, as shown in (32b).

54

Micro cycles

(32) The Copula Cycle involves a. copula > copula + X > X b. demonstrative/3rd person pronoun > locative adverb > posture verb > unaccusative >

= grammaticalization copula copula = renewal copula copula

}

I’ll provide some examples of loss and renewal in what follows. One of the earliest known instances of a Copula Cycle involves the Indo European reconstructed pronoun *es. Benveniste (1971) and Shields (1978, 2004) are among the proponents of the hypothesis that the third person pronoun is reanalyzed as part of the ‘be’ paradigm. According to the latter, traces of its pronominal use can be seen in, for instance, Umbrian, Old Irish, and Old High German before it undergoes “morphological resegmentation” (Shields 1978: 73). A “new demonstrative stem in *i- soon” develops, attested in, for example, Latin is [NOM S M] and id [NOM/ACC S N]. Shields suggests the change in (33). (33)

es (bheu-) mene pater > es mene pater > is es mene pater He is my father 3S.be my father 3S 3S.be my father

Shields doesn’t speculate on the loss of the copula in Indo European, but the most likely cause for the copula’s loss is that the copula merges with another grammatical category (X in (23a)), for example, a negative, and that the new form functions as the renewal for the negative. This scenario has indeed played out in varieties of Arabic. In Standard Arabic, the present tense copula agrees in number and gender with the subject, as in (34) to (36). (34) Ahmad huwa ad-duktoor Arabic Ahmad COP ART-doctor ‘Ahmad is the doctor.’ (35) Huda hiya ad-duktoor-ah Arabic Huda COP.FS ART-doctor-FS ‘Huda is the (female) doctor.’ (36) Ana huwa ad-duktoor Arabic I COP ART-doctor ‘I am the doctor’ (Alsaeedi 2015: 38–41). The copula is homophonous with the third person pronoun, as (37) shows, and originates from that form. In current spoken Arabic, the use of huwa is a bit ‘awkward.’ (37) Huwa yaqra Arabic 3SM read.IMPF ‘He is reading.’

Micro cycles

55

The erstwhile pronoun can be negated in the present tense in the same way as a verb, as in (38), and is therefore considered a copula verb. Once the number and gender are lost on the demonstrative, this form can turn into a negative particle, as in (39) and (40), from Cairene Arabic. (38) faTma ma-hiyya :-sh Egyptian Arabic) il-mas’u:la Fatima NEG-3SF-NEG the-responsible ‘Fatima is not the one responsible’ (Edwards 2006: 53). miš (39) Mohammed hina NEG here Mohammed ‘M. isn’t here’ (Diem 2014: 2).

Cairo Arabic

(40) hiyya miš Cairo Arabic iggawwizzit? she NEG married ‘Hasn’t she married?’ (Woidich 2006: 341) Sentences such as (38) may therefore be seen as participating in a Negative Existential Cycle (NEC), which occurs in a number of varieties of Arabic. In (39), there is a negative copula miš that derives from a form like (38), in particular from ma-hu-šay [NEG-COP-NEG], a copula inside a negative brace. The copula itself originates from a (minimally inflected) demonstrative hu. This negative copula miš, no longer inflected, is now being generalized for emphasis, as in (40). These changes are typical for the NEC, also known as Croft’s Cycle, which will be discussed in the next chapter. Typically, languages that undergo this negative cycle also seem to participate in Copula Cycles (cf. Alsaeedi 2019). The renewal of a copula can come from many sources. A demonstrative/third person pronoun frequently reanalyzes as a copula, as we’ve seen in Arabic. Similarly, in West Greenlandic (41), tassa is a copula but can also (still) be used as a demonstrative pronoun. (41) Hansi tassa pisurtaq West Greenlandic Hansi that leader ‘Hansi is the leader’ (Fortescue 1984: 72, from Hengeveld 1992: 192). In (41), the copula indicates identity, as does the erstwhile demonstrative shi in Chinese (42). (42) Shi wo Mandarin Chinese be 1S ‘It’s me’ (Hui-Ling Yang, p.c.).

56

Micro cycles

Creole languages frequently grammaticalize demonstratives into copulas, as in (43), where the distal demonstrative becomes the copula of identity and continues to be used as demonstrative, as (43) also shows. (43) da somma da wan boen somma Sranan DEM person COP a good person ‘that’s a good person’ (Grammatik 1854: 6; Arends 1986: 107). There are different sources for locative copulas, mainly particles, prepositions, and (positional) verbs (see Hengeveld 1992; Nicholas 1996; Stassen 1997; Pustet 2003). Note the locative adverb ‘there’ as source in (44) and the positional verb zitten ‘to sit’ in (45) and (46). (44) Etnel ɓi/o ɗɛ wan malenge-ma. Saramaccan Etnel PST/IRR COP one lazy-MA ‘Etnel was/will be a lazy cat’ (Veenstra 2014). (45) Jan zit in Frankrijk Dutch Jan be in France ‘Jan is in France’ (Hengeveld 1992: 238). (46) so my hat sat in my office for a year collecting dust (COCA Spoken 1993) Viberg (2013) compares posture verbs used as copula verbs across English, Finnish, French, and German and finds that they have different ratios. Finnish and French use very few, Swedish very many, and English is in between. English derives most of its copula verbs from intransitive (unaccusative) verbs, such as go, turn, remain, stay, and appear, as in (47). (47) He went/turned/remained/stayed/appeared quiet. The reanalysis of the intransitive as copula also requires the reanalysis of the adverb as adjective. This was possible because there is a stage of the language where they have a similar morphological shape. Visser (1946: 65) notes that, even in the sixteenth century, certain adjectives could be used as adjective or adverb. Sentence (48) is such an ambiguous case; (49) and (50) show the transition from an earlier to later use. (48) Since which she was removed to Kimbolton, Where she remains now sick. ‘Since when she was removed to Kimbolton, where she now remains, sick.’ (Visser 1963–1973: 195, c1600, Shakespeare, Henry VIII, 4.1)

Micro cycles

57

(49) the hole body of Christes holy church remaine pure. (Thomas More Works 183 F8, Visser 1963: 195). (50) Our greatness will appear Then most conspicuous. ‘Our greatness will appear most conspicuously then.’ (OED, 1667 Milton Paradise Lost ii. 257) This change has also occurred in Yuman, as Munro (1977) shows. On the basis of what looks like nominative marking on the predicate of a copula, Munro argues that (51) is a complex sentence where an existential intransitive verb has a sentential subject whose last noun gets marked nominative by -č, as shown in (52). The ‘real’ subject is responsible for agreement on the verb (here zero). Mojave (51) John kwaθ?ide:-č ido-pč John doctor-NOM be-TNS ‘John is a doctor’ (Munro 1977: 445). (52) [John kwaθ?ide:-č] ido-pč (Munro 1977: 450) However, this is not a transparent construction because the subject that agrees with the verb in (53) is not the one marked with nominative case. As a result, many (younger) speakers see the predicate noun as the subject and innovate a person marking on the verb that agrees with this new nominative, as (54) shows. (53) ?inyep kwaθ?ide:-č ?-ido-pč Mojave 1S doctor-NOM 1-be-TNS ‘I’m a doctor.’ (54) kwaθ?ide:-ny ?inye-č ?-ido-pč doctor-DEM 1S-NOM 1-be-TNS ‘The doctor is me’ (both Munro 1977: 452). She notes that speakers of other Yuman languages are marking the first noun as well as the second, as in (55). (55) nyu-č pa?ichwa:-v-č-yu-m Yavapai 3S-NOM enemy-DEM-NOM-be-TNS ‘He is an enemy’ (Kendall 1972, from Munro 1977: 470). Concluding this section, copulas arise frequently from demonstrative pronouns and in turn can become negatives, as in Arabic, and then the copula could be renewed again. They also originate from posture verbs and (unaccusative) intransitive verbs. As to where the cycle starts, we have seen that in Arabic (39), the copula has merged with a negative before a new copula has arisen, so this

58

Micro cycles

cycle starts with (a) and then is optionally renewed. The Indo European reconstruction in (33) also assumes the copula disappears first. 4

Tense and aspect cycles

Tense is defined by Comrie (1985: 1; 6) as “the grammaticalisation of location in time” and aspect as the “grammaticalisation of expression of internal temporal constituency” (of events, processes etc.). Tense and aspect are complex notions. Tense is about points in time, the present, past, and future, but the present can frequently be used to mark the future. Tense and aspect also interact; for example, perfective aspect often marks past tense and imperfective present tense. For intricacies about tense and aspect, see Comrie (1976, 1985); Dahl (1985); Bybee et al. (1994), among others. Tense and, especially, aspect go through frequent cycles. Well-known is the Chinese verb liao ‘finish,’ which ends up as perfective le marker. Typically, these markers evolve from full verbs, as in Figure 3.3, to become an auxiliary marking aspect and subsequently tense. Tense also originates from adverbs. Figure 3.3 adds the development of modals, which I will discuss in Section 5. The auxiliary (like the copula we’ve seen in the previous section) may attach to another word, e.g. a verb, pronoun, or negative – the X in (56a) – and lose semantic content and remain as an agreement or negative marker. The form may then disappear altogether. After changing from a verb to an auxiliary, the original verb may be renewed, as in (56b). (56) The TMA Cycle involves a. V > TMA + X b. new verb

>

X

= grammaticalization = renewal

As is clear from Figure 3.3, tense markers arise from adverbs or from aspect markers; they don’t directly evolve from verbs themselves. Although Kuteva et al. (2019: 484) give some examples of a verb changing to past, their examples

full verb

modal verb

serial verb

auxiliary verb

adverb FIGURE 3.3

mood marker

aspect marker tense marker

Tense, aspect, and mood grammaticalization

(see Lehmann 2015: 39)

Micro cycles

59

are cases where the verb is an aspect marker first. Kuteva et al.’s (179) example of a verb meaning ‘finish’ changing to a past tense is one whose “pathway appears to be part of a more extended chain” in which perfect aspect precedes the past. Present tense (Kuteva et al. 2019: 347–348; 485) originates as progressive aspect. An instance where an aspect function precedes a tense function is seen in the Germanic past tense morpheme (-ed in English, -te/-de in Dutch, etc.), which originates as an auxiliary *do and is then reanalyzed as a perfective aspect marker before it becomes a past tense marker (see e.g. Tops 1974). I will therefore discuss the past in connection with the Perfective Cycle. Future tense markers express irrealis mood and will therefore be dealt with in the next section. In the remainder of this section, two cyclical changes in aspect are discussed: (a) the Imperfective Cycle in which a progressive appears and is lost and renewed and (b) the Perfective Cycle in which perfective prefixes are lost and are renewed by telic particles and by the auxiliaries have and be. I finish with a renewal of both imperfectives and perfectives in Basque. 4.1

The imperfective cycle

As for the Imperfective Cycle, languages develop and lose progressives, which Bybee and Dahl (1989: 55) define as a situation that “is in progress at reference time.” To distinguish between different imperfective aspects, Figure 3.4 may be helpful. Starting from the top, the imperfective can be divided into habitual and non-habitual. Habitual aspect describes an activity that occurs habitually, while continuous aspect is a term that covers the (non-habitual) imperfective of all verbs, both activities and states. Actions that are ongoing are marked differently than states in many languages, namely, as progressive as opposed to nonprogressive.

FIGURE 3.4

Classification of aspectual oppositions

(from Comrie 1976: 25)

60

Micro cycles

The progressive aspect (PROG) may, however, generalize to include stative aspect as well, and then it becomes a marker of continuous or even imperfective (IMPF) aspect; see Bybee and Dahl (1989) for examples. As the marking changes in function, it moves upwards in Figure 3.4. This is known as the Imperfective Cycle and schematized in (57) (cf. Bybee et al. 1994: 139). (57) The Imperfective Cycle V/P > Progressive

>

continuous

> Imperfective

As Bybee et al. (1994: 127; 139) note, there is, however, no language that marks continuous (or non-progressive). They also note that it “is difficult to view the so-called present tense as a ‘tense’ (126) since the present covers imperfective.” I agree and will not discuss the present although an imperfective could in principle change to a present tense. Another representation of the Imperfective Cycle is given in (61), which is adapted from Deo (2015). She doesn’t distinguish imperfective from continuous, but I will give examples that show the progression, as in (57). What (58) shows is the change from no progressive to optional to obligatory progressive and back to no special progressive. (58)

(a) (b) (c) (d)

IMPF > PROG zero-PROG = IMPF emergent-PROG categorical-PROG generalized-PROG

languages Russian, Arabic Arabic varieties, Dutch, French English, Swahili Turkish, Tigre

In (58), stage (a) means that there is one general imperfective form for states and activities, as (59) shows, and no special progressive for activities. (59) a. ana I b. ana I

a-erif Standard Arabic IMPF-know, ‘I know.’ a-ktub IMPF-write, ‘I write habitually/am writing now.’

However, note that Standard Arabic has an auxiliary with the past imperfective, as in (60), so that progressive could start to spread to the present as well. (60) kunt a-ktub Standard Arabic be.PST.1S IMPF-write, ‘I was writing.’ Stage (58b) has an optional, emerging progressive, which is frequent in many varieties of Arabic, as shown by Holes (1990) and Bakir (2010) for the Gulf varieties, Saddour (2009) for Kuwaiti Arabic, Jarad (2015) for Emirati Arabic,

Micro cycles

61

and by Altamimi (2021) for Najdi Arabic. The verbs jaalis/qaaʕaɪd ‘sitting’ and gaayem ‘standing’ can both mark duration with activity verbs, as in (61a) from Najdi, but not with stative ones, as in (61b). (61) a. hom jaalis-iin ya-ktub-un el-waʤib Saudi Najdi Arabic they PROG-3P.PL.M 3P.M writeIMPF-PL the-assignment ‘They are writing the assignment’ (Altamimi 2021: 76). b. *antum jals-iin ta-krah-un e-safar li ʤidah you PROG-2P.M.PL PL.M 2P.hateIMPF-PL the-travel to Jeddah Intended: ‘You hate traveling to Jeddah’ (Altamimi 2021: 79). However, Al Aloula (2021), studying the difference in acceptability between durative and stative verbs finds that some speakers accept (61b), so there may be a change to stage (58c). Other varieties of Arabic use a preposition that precedes the imperfective, for example, Egyptian, as described by Saddour (2009), as optional progressive markers. Dutch and French are similarly in stage (b), with the optional progressive in (62a) and (63a) and a present tense with a similar meaning in (62b) and (63b). Sentences (62c) and (63c) show the impossibility of a progressive with stative verbs, as expected. (62) a. Ik ben een boek aan het lezen. Dutch I am a book on the reading b. Ik lees nu een boek. I read now a book ‘I am reading a book now.’ c. *Ik ben dat antwoord aan het weten. I am that answer on the know ‘I know that answer.’ (63) a. Elle est en train d’écrire une lettre she is in process PRT.write a letter b. Elle écrit une lettre. she write.PRES a letter ‘She is writing a letter.’ c. *Elle est en train de savoir. she is in process PRT know ‘She is knowing.’

French

Stage (58c) is relevant to Modern English, where all ongoing activities are marked with -ing. English makes a clear distinction between nonprogressive and progressive aspect. These distinguish between stative and activity verbs, respectively. Modern English marks nonprogressive as simple present or past,

62

Micro cycles

as in (64ab), and -ing is not possible, as (64c) shows. Present and past progressives are marked through -ing, as is shown in (65ab). (65c) shows that the simple present is not possible for activity verbs unlike states in (5). (64) a. b. c. (65) a. b. c.

I know French. I knew French five years ago. *I am knowing French. She is walking right now. She was walking yesterday. *She walks right now.

stative (none in COCA; 3 in BNC) activity (not attested)

As we’ll see later, the -ing marking may be in the process of being extended to stative verbs. This means that the -ing is being used for continuous aspect and has become an imperfective, as represented in stage (58d). Stage (58d) is the last stage, where the progressive has generalized to other imperfectives. Deo mentions this last stage as one relevant to currentday Hindi. Speakers from the nineteenth century use (66) for the present progressive, but present-day speakers see this as a habitual. Instead, the younger speakers use (67) for the progressive, which is optional for older speakers. (66) Mẽ kitab lykhti hũ Urdu/Hindi I book writing.F be.1S ‘I (a female) am writing a book’ (older use). ‘I (habitually) write books’ (present-day use). (67) Mẽ kitab lykh rehi hũ I book write sit.F be.1S ‘I (a female) am writing a book.’ Note that the original ‘be’ and present participle in (66) is already a form that went through the cycle. As Slade (2013: 566–567) argues, the auxiliary was originally a renewing light verb indicating a progressive. The original present in Classical Sanskrit is given in (68) and is similar to (66). (68) calan bhavati/ go.PR.PRT be(come).PR3S ‘keeps on going.’

asti be(come).PR3S

Classical Sanskrit

As mentioned, the cycle in (58) doesn’t distinguish habitual and continuous. Adding that gives us Figure 3.5. This shows that the original progressive is lost and is renewed by another, similar to the one in varieties of Arabic in (60).

Micro cycles

FIGURE 3.5

63

The imperfective cycle

Urdu/Hindi does use the present participle and auxiliary strategy with stative verbs, as in (69a), which I have included in Figure 3.5. In fact, as (69b) shows, even the innovative progressive is already being used with statives as well. (69) a. me is ko pasand kerti hu᷉. Urdu/Hindi I this OM pleasing do AUX ‘I like this.’ b. Ham ap ka show ko bhot pasand kar raha ha we your POSS show OM very pleasing do remain AUX ‘We are really liking your show’ (Facebook). I’ll now show how the history of English shows a cycle of optional progressive to obligatory progressive to imperfective marking. As de Wit et al. (2013) and de Wit (2017: 54) have shown, “under pressure of the grammaticalizing imperfective,” the present tense also shifted to a perfective meaning. I will not discuss that interaction, however. Durative present tense verbs in older stages of English do not need an -ing. This is shown for Old English in (70), for Middle English in (71), and for Early Modern English in (72). (70) nu ic arisu cwið drihten Now I rise said lord ‘Now I rise up said the lord’ (c800, Vespasian Psalter 11.6, Visser 663). (71) What do ye, maister Nicholay? What do you, master Nicholay ‘What are you doing, master Nicholay’ (c1400, Chaucer, Miller’s Tale 71.3437). (72) What say you, Scarlet and John? ‘What are you saying, Scarlet and John’ (c1600, Shakespeare, Merry Wives I, i, 155).

64

Micro cycles

Throughout the history of English, there have been optional ways to mark progressive action by means of a preposition and a participle, as in (73) from Old English, (74) from Middle English, and (75) from seventeenth century English. As in Modern English, none of these involve states. Hence, the auxiliary and participle optionally mark the progressive, not continuous aspect. (73) Exorciste beoð on getacnunge Cristes gespelian Exorciste is on teaching Christ’s story ‘Exorciste is teaching the story of Christ’ (DOE, c1000, De Ecclesiasticis Gradibus). (74) þær he wes an slæting there he was on hunting ‘he was hunting there’ (c1250, Layamon, Caligula 6139). (75) I am upon writing a little treatise (Pepys’s Diary 31 Dec 1666, from Visser 1963–1973: 1998). In addition to the forms with a preposition, there is, in Old English, a form in -ende (or -ande or -inde depending on the regional variety) that resembles the modern construction in (65a), namely, (76) and (77), and in Middle English, one in -yng/-ing, as in (78). It is often argued that in Old English these participles are adjectives (cf. Mossé 1938, I: 3; Denison 1993: 373–377; Ziegeler 1999) and that the constructions involve copulas followed by adjectival forms. (76) ac se æglæca ehtende wæs but that monster pursuing was ‘but that monster was chasing’ (c1000, Beowulf 159).ð (77) þe þer were wuniende REL there were living ‘who were living there’ (c1200, Lambeth Homilies 41, from Mossé 1938, I: 81). (78) We han ben waitynge al this fourtenyght ‘We have been waiting these two weeks’ (c1400, Chaucer, Knight’s Tale 38.929). If these are adjectives, they are based on activity verbs. Table 3.3 provides all forms in -ende in the Peterborough Chronicle, which gives a good sense of Early Middle English use, and that shows that these are activity verbs which possibly makes -ende into an (early) optional progressive. The form in -yng (or -ing) in (78) is a Middle English innovation, but whether it is a direct continuation of the -ende form continues to be contested (see Mossé 1938: II, 36, Jespersen 1940: 415, and Ziegeler 1999). In Chaucer, it is not that frequent after a form of to be (it is frequent as a verbal noun and adjective).

Micro cycles

65

TABLE 3.3 Forms in -ende in the Peterborough Chronicle

delnimende ‘taking part,’ wuniende ‘living,’ (on)feohtende ‘fighting,’ fleogende ‘fleeing,’ winnende ‘winning,’ hergende ‘plundering,’ tiligende ‘striving,’ wircende ‘working,’ biddende ‘praying,’ (land)sittende ‘sitting,’ forðbringende ‘bringing forth,’ scynende/scinende ‘shining,’ byrnende ‘burning,’ tyrwigende ‘annoying,’ wexende ‘waxing,’ dreogende ‘threatening,’ waniende ‘waning,’ ridende ‘riding,’ gangende ‘going,’ and sprecende ‘speaking.’

Some evidence that the two constructions are connected is the fact that, in many texts, both forms occur seemingly ambiguous, as (79ab) show. (79) a. Where þe dragun was wonande where the dragon was living ‘where the dragon lived’ (c1300, Handlyng Synne 1760). b. Wher þe old man wonyng was where the old man living was ‘where the old man lived’ (c1300, Handlyng Synne 8504). In the fifteenth century Paston Letters, a variety of letters spanning four generations, the special progressive of (80) and (81) is relatively rare (around 20 perhaps in a corpus of over 250,000 words) but possibly already expresses continuous aspect, as (82) shows. (80) þer ys comyng . . . more than a thowsand ‘More than a thousand are coming’ (Paston Letters #187, anno 1465). (81) where the seid felechep is abydung ‘Where the above-mentioned fellowship abides’ (Paston Letters #40, anno 1452). (82) syche mony that he is owyng ‘such money that he owes’ (Paston Letters #336, anno 1469). However, the progressive is definitely optional because the present is typically expressed as in (83). (83) a. I send you ‘I am sending you’ (Paston Letters #3, anno 1425). b. I make þis day a new apelle ‘I am making a new appeal today’ (Paston Letters #4, anno 1426). By the time of Thomas More, that is, the early part of the sixteenth century, the progressive is still “employed rather sparingly” (Visser 1946: 248), as in (84).

66

Micro cycles

Visser (p. 662) says, about both More and Shakespeare that “at the time the choice between the two possibilities did not yet depend on any fixed principle.” The simple present is used frequently, as in (85). (84) Now she’s going to my wife (c1600, Shakespeare, Merry Wives III, ii, 36). (85) a. Whether go you. . . . To see your wife ‘Where are you going’ (c1600, Shakespeare, Merry Wives II, ii, 10). b. But what saies shee to mee? (c1600, Shakespeare, Merry Wives II, ii, 75) In Shakespeare, there are a number of cases where the -ing progressive is used with verbs such as live, but not with know and see. So I conclude that -ing is definitely used by Shakespeare as a progressive. The use of the progressive increases especially in the nineteenth and twentieth centuries, see Scheffer (1975: 110ff.). Checking certain simple present and progressive forms (eat/is eating; run/is running) in Jane Austen’s early nineteenth century Emma, Persuasion, and Pride and Prejudice shows that by this time the use is modern, with the simple present being used for habitual aspect, as in (86), and the progressive for activity verbs, as in (87). (86) a. b. (87) a. b. c.

I dare not let my mother know how little she eats (Emma II, ch 9). whenever she comes to Highbury (Emma I, ch 10). he is writing about it now (Persuasion ch 23). At this moment . . . Mr Elton is shewing your picture . . . (Emma I, ch 7). If you are looking for my master . . . he is walking towards the little copse (Pride & Prejudice II, ch 7).

Considering this, the evidence points to the unmarked aspect changes in the nineteenth century. The sharp increase in the progressive around 1800 has been the subject of studies from Scheffer (1975), Strang (1982), Kranich (2010), Ziegeler (1999), Smitterberg (2005), and Hundt (2004). The latter three focus on changes in the nature of the subject. I’ll now turn to changes currently affecting the progressive and possibly changing it into an imperfective marker. Evidence for this change is that the progressive -ing is being used with statives, as in (88) and (89). (88) a. the skiers are loving this – about a foot of snow for you. (COCA Spoken 2009) b. You know, I really am loving producing films, television, and stage. (COCA Spoken 2013) (89) a. I’ve been going around, and I am liking what I see in the classrooms. (COCA Spoken 2002)

Micro cycles

67

b. So how’s he liking his new single status? (COCA Magazine 2012) c. I don’t know that an anti-Gingrich or Romney or Obama group on Facebook that has a lot of people liking it is going to have a big impact on the campaign, but this is how people organize themselves these days online. (COCA Spoken 2012) The change to -ing is affecting a large set of stative verbs, for example, (90) (90) a. b. c. d.

How is it tasting? (waiter) Pizza is tasting an awful lot like something other (COCA news 1992) How his throat is feeling today (COCA spoken 2014) This is sounding bad (COCA spoken 2015)

Even verbs like know in (64) occur – although infrequently – with the -ing, as (91) shows. (91) a. We’ve been knowing this stuff has been going on in the NFL. (COCA Spoken 2005) b. And so everybody in town was knowing that this was happening. (COCA Spoken 2009) c. Well, I’ve been knowing Ross ever since he was about 11. (COCA Spoken 2014) Rautionaho and Fuchs (2021) put this change as follows: “The spread of the progressive from dynamic to stative verbs started in the seventeenth century, and slowed down in the late twentieth century.” They continue, “Overall, stative progressives have not become more frequent in the last twenty years, and that the group of stative verbs is highly heterogeneous.” So far, I have shown that -ing becomes obligatory as a progressive in the nineteenth century and starts on its way to become a general imperfective marker. Mossé (1938, II, 2ff.), based on Streitberg (1891), attributes the popularity of the progressive to the demise of the (perfective) aspectual system occurring from the thirteenth century on. Considering (58), that connection may be correct in that, for instance, Arabic and Russian, mark the perfective but not the progressive. However, the history of English shows quite a lag between obligatory perfective marking (no longer strict in Old English) and obligatory progressive marking (starting in the nineteenth century), and they may be independent of each other. The Imperfective Cycle provides a possible scenario for what’s happening to verbs like love and like in (88) and (89): the -ing construction is expanding its scope to imperfective.

68

Micro cycles

4.2 The perfective cycle

As for the Perfective Cycle, Bybee et al. (1994: 105) propose a cline, which I have adapted as (92). A resultative refers to a state brought about by a past event with the resultant state as its focus, an anterior refers to an event in the indefinite past that has present relevance (also known as perfect), a perfective signals a completed action, and the past places an event before the speech time. (92) The semantics of the Perfective Cycle involve a. Resultative > Anterior > Perfective b. New resultative

> Past

To make this cycle more understandable, I provide a hypothetical example. If a resultative particle, e.g. up in eat up, would be reanalyzed as an anterior and perfective, as in (93), a new resultative might appear. Anterior > Perfective > Past (93) a. Resultative > I ate up the apples I up ate the apples I upate NP I upate NP (= they are still eaten) (and can’t eat more) (= finished) (I ate before now) b. new particle ate through

The resultative stage in (93) would emphasize the end point of the event and the lasting effect on the apples, the anterior would mark that the eating event has current relevance, the perfective shows that eating is seen as bounded, and the past that the eating happened before speech time. The renewal of the resultative could be another particle or auxiliary. One example of a Perfective Cycle is the loss of ge- (and other resultative affixes) in the history of Germanic. The resultative nature of the ge- prefix exists in Gothic, as Streitberg (1891), Lloyd (1979), and EyÞórrson (1995), among others, have argued and is shown in (94). (94) jah sums qam qiþands: frauja, sai, Gothic and another came saying: Lord, look, sa skatts þeins þanei habaida galagidana in fanin DEM money your, REL.MS.ACC have.PST.1S lay.PP.MS.ACC in cloth (Gothic Bible, Luke 19.20) Gothic’s descendant Old English shows the use of ge- as resultative (cf. e.g. McFadden 2015), as well as its use in the last stages of (93), namely, used as perfective (or past) in (95) (cf. Wischer & Habermann 2004; Broz 2013).

Micro cycles

69

(95) Her Hengest & Æsc gefuhton wið Walas Here Hengest and Ash fought againt & genamon unarimenlicu and took countless ‘In this year Hengest and Ash fought against the Welsh and took countless things’ (c1100, Peterborough Chronicle, 473.1). In Middle English, in the later part of the Chronicle, the resultative particles in (96) and (97) take over the resultative function that ge- held previously. (96) Sum he iaf up Some he gave up ‘Some castles, he gave up’ (Peterborough Chronicle, anno 1140). (97) he uuolde iiuen heom up Wincestre he wanted give them up Winchester ‘He wanted to give Winchester up to them’ (Peterborough Chronicle, anno 1140). The particles would be expected to continue on their path to anterior and perfective, but that hasn’t happened yet, and the Modern English particles still indicate result, as in (98). The particle is still in the resultative stage of (93). The full stages are shown in (99). (98) They issued it out. (99) Perfective Cycle Resultative > Anterior > Perfective > ge- (Gothic and Old English ge- (Old English) up, out, etc (Modern English)

Past

Detges (2000) examines the change in (92) in English and Spanish and concludes that the change is brought about by discourse strategies of the speaker. A resultative denotes a present result that is relevant to the subject of the sentence. Two inferences can be drawn from the resultative: the event was in the past, and the subject is the Agent of the event. Either of these inferences can be conventionalized, which in the former case results in a past tense reading with current relevance and in the latter in the highlighting of the Agent, as in (100). (100) þu þe self hafast dædum gefremed þæt . . . you yourself have by.deeds achieved that ‘You have achieved through your deeds that . . .’ (c1000, Beowulf 953– 954, Detges 2000: 350).

70

Micro cycles

A second example of the Perfective Cycle, possibly resulting from the loss of ge-, involves changes in have. The resultative use of the have auxiliaries (and be) with a participle has been argued by many to occur in Germanic (Traugott 1972). It arises from the possessive have that has an object and where the participle indicates the result. Visser (1963–1973: 2189) mentions (101) as resultative. (101) Gyt ge habbað eowre heortan geblende? If you have your hearts blinded ‘If your hearts are blinded’ (DOE, c1000, West Saxon Gospels, Mark 8.17). Carey (1995) argues for a change from resultative to anterior in Old English and her results are given in Table 3.4. Present state adverbials, such as nu ‘now’ in (102), are compatible with resultatives and are gradually replaced by anterior adverbials, such as ær ‘before’ in (103) and (104). (102) ða cwæð se Wisdom: Nu hæbbe ongitenþine ormodnesse then said that Wisdom: Now I have understood your unhappiness ‘Then Wisdom said: I now understand your unhappiness.’ (c900, Boethius 13.15) (103) and hæfde ær his ðing þearfum gedælede and had before his things needy divided ‘He had earlier divided his things among the needy.’ (DOE, c1000, Ælfric Homilies, Godden 289.29) (104) Headda abbot heafde ær gewriton hu Wulfhere . . . Headda abbot had before written how Wulfhere ‘Headda the abbot had before written how Wulfhere . . . ‘ (c1100, Peterborough Chronicle, 350). This shows a change from resultative to anterior, in accordance with (99). Once the simple perfective arises, it becomes a competitor for the simple past, as Mustanoja (1960: 504) explains. TABLE 3.4 Changes in the have perfect in Old and Middle English

Present state Anterior

Alfred 900 CE 50 (38.8%) 0

Aelfric 1050 CE 37 (35.6%) 0

Layamon 1250 CE 20 (12.4%) 8 (5%)

Gawain 1375 CE 0 7 (14.9%)

Total

129

104

161

47

(from Carey 1995: 86)

Micro cycles

71

The auxiliaries have and be are not generally used for anterior aspect (yet) in Old English. If Old English doesn’t have a special anterior, what expresses the present and past perfect? MacLeod (2012: 226–227) writes that the preterite [simple past] and the periphrastic perfect and pluperfect were available in Old English as a means of expressing the semantic and pragmatic content belonging to the perfect and pluperfect domains, but that these two formal categories differed in their perceived stylistic value. So in Old English, have and a past participle sometimes express anterior aspect and obligatorily do so in late Middle English. The cline in (99) also predicts a stage where the anterior becomes perfective and then past. This hasn’t happened in English but it has in German and Dutch, where the phenomenon is known as the ‘Präteritumschwund’ [past-disappearance] (cf. Abraham & Conradie 2001) because the simple past is replaced by have and participle, as shown in (105). (105) Ik heb dat gisteren gedaan I have that yesterday done ‘I did that yesterday.’

Dutch

There are stages of English where (106) to (108) appear, from different centuries and varieties, but this is not part of standard English. (106) a. he hath acordyng to youre desyre spoken Yeluerton yesterday. ‘He spoke to Yelverton yesterday, as per your wish’ (Paston Letters, #184, anno 1465, Margaret). b. I haue send to Ser Thomas Howys yesterday . . . ‘I sent yesterday to Sir Thomas Howys’ (Paston Letters, # 190, anno 1465, Margaret). c. I haue yesterday sent to Herry . . . ‘I sent to Harry yesterday . . . ’(Paston Letters, #352, anno 1472, John III). (107) The river has fallen yesterday and to-day nearly eighteen inches. (COHA 1820) (108) Anyway (pause) three people have phoned yesterday. (BNC spoken) These might suggest that the English present perfect is a simple past in those varieties. Various scholars have discussed the loss of the simple past in English, for example, Zieglschmid (1930). However, there are different changes happening. On the one hand, Elsness (1997), investigating the relationship between the simple past and the present perfect, shows that, after an initial increase of the perfect, there is currently a move back to a use of the simple past or at least not

72

Micro cycles

a reanalysis as past, as in other Germanic languages (Elsness 1997: 347). On the other hand, Fryd (1998) and Walker (2011, 2017) have shown that certain varieties are in fact starting to use the present perfect as emphatic past. Relating the changes in have to the Perfective Cycle, we can see that the developments from (101) to (108) follow this cycle. The last stage is for have to be used as a past, and that occurs infrequently in English, for example, in (106) to (108), but it does occur in related languages, as shown in (109). (109) Resultative > Anterior > Perfective > Past have (Old English) have (Old and Middle English) have (Germanic) (109) can be added to (99) showing loss of ge- and renewal by have. Both changes happen during the Old English period. 4.3 Imperfective and perfective renewal in Basque

To end this section, we turn to the history of Basque, which shows a renewal of both the imperfective and perfective, as described by Haase (1994) and de Rijk (1995), among others. There are two conjugation types, an older synthetic pattern, as in (110), and a newer analytic one, as in (111). (110) ba-nago Basque ENC-stay.PR.1S ‘I stay’ (Haase 1994: 284; ENC = enunciative). (111) Ego(i)-te-n naiz Basque stay-NM-IN PRS.1S ‘I am staying’ (Haase 1994: 285). The synthetic present in (110) is aspectually neutral, but the analytic present progressive in (111) consists of an auxiliary with a nominalized lexical verb that is marked inessive. The latter is an innovation that can be seen in many languages (cf. Kuteva et al. 2019: 272–273). Synthetic verbs just have a present and past opposition, shown in (110) and (112), and do not express perfect or progressive aspect, which the analytic verbs do, shown in (111) and (113). (112) ba-nengoen Basque ENC-stay.PST.1S ‘I stayed’ (Haase 1994: 284). (113) Joan da Basque go.PTC PRS.3S ‘S/he has gone’ (Haase 1994: 285).

Micro cycles

73

Haase (1994: 289) concludes that “the analytical construction is an innovation” with the innovation most probably “originating from spoken Latin.” Although Padilla-Moyano (2013) reviews the evidence and argues that both synthetic and analytic verbs are of antiquity, the evidence is that synthetic verbs are now reduced to about a dozen from 68 in the sixteenth century (Mounole 2011; Jendraschek 2003). Section 4 has discussed two aspectual cycles, an Imperfective and Perfective Cycle. In the Imperfective Cycle, a progressive appears, generalizes to imperfective, and a new progressive arises. The progressive can be lost before it is renewed, and many languages lack an obligatory progressive, for example, standard Arabic, Dutch, and French. In the Perfective Cycle, a resultative construction gets used as anterior and perfective. As part of this cycle, a past tense can also appear. Loss and renewal in the perfect in English seem to be at the same time. The interaction between these two cycles has also been noted. For instance, as mentioned previously, Streitberg (1891) and Mossé (1938) link the introduction of a progressive in English to the demise of the perfective prefix, and de Wit (2017) examines the extension of the present tense to perfective to the change of the progressive to imperfective. 5

Mood and modal cycles

Different moods express modalities such as how certain the speaker is about the event, how they’ve acquired the information, and what the illocutionary force of the sentence is. Another word used in this context is irrealis or non-factual mood typical of subjunctives, as opposed to realis mood exemplified by declaratives. Mood and modality can be expressed synthetically or analytically, that is, through markings on the verb or through auxiliaries, respectively. Not much is known about the path of the synthetic subjunctive, but a lot is known about changes in the analytic modal. I’ll first say a little on the subjunctive but mostly concentrate on changes in modals. As for the synthetic mood, the indicative mood in (114) indicates that the speaker is certain about the truth of the utterance, whereas the subjunctive mood, for instance, in Spanish (115), indicates uncertainty on the part of the speaker, which the English translation renders with a modal auxiliary. (114) Genius visited yesterday. (115) Dudo que me visite la genialidad. Spanish doubt that me visit-SUBJ ART genius ‘I doubt that genius will visit me’ (Corpus de Español). Verbal indicative endings originate from pronouns, as we’ll see in Chapter 5. Subjunctives arise from “affixes comparable to those that were selected to mark

74

Micro cycles

preterites and perfects,” according to Lehmann (1993: 181), so a provisional cycle is represented in (116). (116) Mood Cycle preterite/perfect >

subjunctive > zero

Old English still has (synthetic) subjunctives, but these are lost and replaced by (analytic) modals by the end of the Middle English period. Haeberli and Ihsane (2016: 533) write that, “[w]hereas there are clear inflectional distinctions between indicative and subjunctive in [Old English], these distinctions are considerably reduced during the [Middle English] period” and lost around 1500 (see also Mustanoja 1960: 453). Fischer (2003: 20) sees the loss of the subjunctive in Old English “as the end of a modality cycle,” renewed by (analytic) modals, as in (117). (See also Kuteva et al. 2019; Bybee et al. 1994: 191–195; 204–205). (117) Modal Cycles a. V > deontic > epistemic > future =grammaticalization will (118) will (119) will (120) ‘ll (121) b. want (122) > wanna (123) want (124) = renewal Sentences (118) to (124) illustrate the stages. In (118) to (120), I provide the verbal, deontic, and epistemic meanings of Old English willan and in (121) its future use. Sentences (122) to (124) do the same for the renewal of will, namely, want. (118) Ne drincð nan man eald win & wylle sona þæt niwe. =V NEG drinks no man old wine and wants then the new ‘No man drinks old wine and then asks for new wine’ (OED, c1000, West Saxon Gospels: Luke v. 39). (119) Þæt, la, mæg secgan se ðe wyle soð specan. = deontic That indeed may say he REL wishes truth speak ‘He who wishes to speak the truth may indeed say that’ (OED, c1000, Beowulf 2864). (120) þou wolt fur-sake me þrien ar þe coc him crowe. = epistemic You will forsake me thrice before the rooster REFL crow.SUBJ ‘You will deny that you know me three times before the rooster crows’ (OED, a1275, in Brown English Lyrics 13th Cent. 39). (121) You’ll do that = future (122) I want cookies =V (123) I want to (wanna) help you = deontic (124) Feels like it wants to rain (song title) = epistemic

Micro cycles

75

Turning to analytic mood, modal auxiliaries are frequently used in this context. The change from a verb to a deontic modal auxiliary to an epistemic modal is probably the best known chain of grammaticalization (Closs 1965; Lightfoot 1979). As we just saw, all stages are already present in Old English, verbal forms, such as cunnan ‘know’ and wille ‘want’ in (125), the deontic modal motan ‘be allowed to’ in (126), and the epistemic modal mihte ‘might’ in (127). So, their development overlaps with the demise of the subjunctive, mentioned before. The core modals in English originate in verbs of ability, as shown in (125) with cunnan, and volition, as also shown in (125) with wille, and verbs of motion and intention. These verbs “invite” certain inferences of futurity and possibility (e.g. Traugott 1972; Traugott & König 1991). (125) Ic sceal hraðe cunnan hwæt ðu us to duguðum gedon wille. I must swiftly know what you us to benefit do want ‘I must know quickly what you want to do to benefit us’ (DOE, c1000, Andreas 341, Vercelli Book). (126) No þe laðes ma þurh daroða gedrep gedon motan. Not you harm more through spear stroke do may ‘No more injury will they be allowed to do to you through the stroke of darts’ (DOE, c1000, Andreas 1443, Vercelli Book) (127) Þis mihte beon geseald to myclum wurþe & þearfum gedæled. this might be sold for many worth and poor given ‘This might have been sold for much and given to the poor.’ (DOE, c1000, West Saxon Gospels, Matthew 26.9) In sentence-repair tasks, Cournane (2014 and other work) has shown that children take deontic modals, such as will, and use them as epistemic ones, where adults may use clearly epistemic might, in repairing the gap in (128). (128) Because Sam ... (Cournane 2014: 110)

fall in the water

This explains the historical development: children grammaticalize words, for reasons we come back to in Chapter 6. Fischer (2003: 20) coins the term ‘Modality Cycle’ to describe the changes from a main verb to a deontic and epistemic modal and shows a later renewal of the deontic modals, such as must, by semi-modals like have to. She sees a real cycle in that “the original modal verbs may grammaticalize into auxiliaries, clitics, and affixes on the verb” (20) and be renewed by semi-modals. Gergel (2009) uses the ‘Modal Cycle’ for the change from the adverb rather to the modal auxiliary, and I too will be using the term ‘Modal Cycle,’ rather than ‘Modality Cycle,’ because my main emphasis is on verbs and auxiliaries.

76

Micro cycles

As for the recent renewals in the modal system that Fischer refers to, Leech (2003: 223) writes that “English modal auxiliaries as a group have been declining significantly in their frequency of use.” Leech’s corpus analysis looks at modals around the years 1961 and 1991. His data show a decline of 10% in the 30 year period, but modals are not all acting in the same way; for example, may increases its epistemic use, but should does not. Leech looks for a possible increase in the use of semi-modals. This loss of core modals and increase of semi-modals is also mentioned by Fischer (2003), who sees the semi-modals as replacing the deontic ones. The increase occurs with deontic have to, need to, want to, and used to but at different rates. Leech et al. (2009: 78) argue that the picture is more complex and that core modals still outnumber semi-modals. The shift to semi-modals has been attributed to a shift towards more colloquial language (Myhill 1995; Leech 2003; Leech et al. 2009). The core modals are in decline in both American and British English, according to Leech et al. (2009) and Leech (2011), but there is individual difference (Love & Curry 2021). So where their epistemic counterparts flourish, deontic verbs disappear (Kranich 2021) and are replaced by semi-modals (Leech et al. 2009: 90). The class of increasing semi-modals, for example, have (got) to, need to, and want to, are often rendered as hafta, needa, and wanna. These do not behave like other auxiliaries, such as in undergoing Subject Auxiliary Inversion (cf. *Hafta I go?), and phonologically look like the modal with the perfect auxiliary as mighta, coulda, woulda, shoulda, and musta. The raising verbs seemta and liketa are also seen as semi-modals, expressing modality, ending in -(t)a, and not participating in SAI. Bolinger (1980: 297) adds many more to this list, noting that “[t]he moment a verb is given an infinitival complement, that verb starts down the road of auxiliariness.” Quirk et al. (1985: 137) provide a continuum of the stages between a main verb and a modal, as presented in Figure 3.6. In the remainder of this section, I provide examples from Arabic and Slovenian that illustrate variations on the possibilities in (117). I then come back to possible modal renewals in English. An example of a change from a verb to auxiliary that skips some intermediary steps is taken from Arabic. In Syrian, Gulf, Cairene, and Levantine Arabic, the

mono-clausal

bi-clausal FIGURE 3.6

core modals (can, could, will, would, shall, should, may, might, must) marginal modals (dare, need, ought to, used to) idioms (had better, would rather, …) semi-auxiliaries (have to, be about to, be able to, wanna …) catenatives (appear to, happen to, seem to,…) main verb (hope to, begin …)

Verb auxiliary continuum (adapted from Quirk et al. (1985: 137)

Micro cycles

77

volition verb is grammaticalized to a future marker; see Persson (2008), Owens (2018), and Altamimi (2021), who argue this for Saudi Najdi Arabic where, in (129), we see a deontic modal originating as a lexical verb of volition. In the same stage of the language, we see a prefixed future marker in (130). (129) mona baɣ-at

ta-ʔtˤi axt-ah ħalau Saudi Najdi Arabic Mona want.PERF.3P.SG.F 3P-SG.F-give.IMPF sister-her candy

‘Mona wanted to give her sister candy’ (Altamimi 2021: 103). (130) laħaẓat wa bi-ya-ktub-uun el-waʤib Saudi Najdi Arabic moments and B-3P.F/M-writeIMPF-PL the-assignment ‘Moments and they will write the assignment’ (Altamimi 2021: 81). In Slovenian, we see a different renewal of modality, that is, through an adverb. In present-day Slovenian, there is a modal auxiliary in the negative (131a), as well as a modal adverb in positive contexts, as in (131b). The modal adverb derives from the manner adverb, which is still used, and is phonologically weakening in certain varieties of Slovenian. (131) a. Ne morem iti v kino. Slovenian not can.1S go.INF in cinema ‘I can’t go to the cinema.’ b. Lahko grem v kino. easily go.1S in cinema ‘I can go to the cinema’ (Hansen 2005; Greenberg 2006: 130, from Marušič & Žaucer 2016). Marušič and Žaucer (2016) argue that an early stage of the language has a modal auxiliary, which is strengthened by the adverb lahko, after which the auxiliary is lost, except in the negative. The evidence for this scenario is that in the earliest texts of a thousand years ago, the auxiliary is used in the positive, as well as the negative, as (132) shows. (132) a.

Tîge se mosem I mui este like.that PRT can.1P and we still ‘We can still be like them’ b. . . . egose ne mosem nikimse liza whose not can.1P noone face ‘before his face we cannot hide behind 2016: 171).

buiti . . . Old Slovenian be.INF ni ucriti . . . us hide anyone’ (Marušič & Žaucer

In the sixteenth century, both positive and negative modal contexts continue to have the auxiliary, but a renewal appears. The change is that positive possibility

78

Micro cycles

contexts have either an auxiliary and an adverb, as in (133), or just the adverb, as in (134). In this stage, the auxiliary and adverb can be ordered freely. (133)

Leto zapuvid bi Adam inu Eva bila cilu lahku mogla deržati . . . 16th C Slovenian

this commandment COND Adam and Eva were whole easily can hold ‘Adam and Eve could follow this commandment . . . ’ Trubar 557, fromMarušič & Žaucer 2016: 173). (134) . . . , taku iz tiga vsaki lahku zastopi, de 16th C Slovenian so from this everyone easily understands, that . . . ‘so that everyone can understand that . . . ’ (Trubar 557, from Marušič & Žaucer 2016: 173). This situation continues until the nineteenth century when even negative modal auxiliaries cooccur with an adverb, as in (135). (135) Da pa je . . . veliko dela, o tem pač ne more lehko še kdo dvomiti. that PTCL AUX a-lot work of this PTCL not can easily still anyone doubt ‘But that there is a lot of work, no one can doubt’ (1891, Slovenski gospodar, Marušič & Žaucer 2016: 177). From about the second half of the twentieth century, Marušič & Žaucer show that the double modal construction, as in (135), is no longer attested, except in linguistic enclaves. The modal adverb becomes restricted to the positive context. The exact reasons can be found in Marušič and Žaucer (2016: 183); they are due to the nature of the negative particle to which the verb attaches and the changes in the status of the adverb (from phrase to head), which then intervenes. The changes in Slovenian thus show a partial Modal Cycle: renewal of a modal auxiliary by an adverb and then loss of that auxiliary in a certain environment. Now, I turn to the change that is happening in English where, after modals, the perfective auxiliary have is reduced to -a and then renewed by another have. So in addition to the shift away from the core modals to the semi-modals mentioned previously, there is another change going on relevant to the Modal Cycle, namely, a new modal ending in -a, making core modals and semi-modals more uniform in form but not necessarily syntactic behavior. These new modals often appear as musta, shoulda, mighta (or typological variants such as mustav and mustof) and can be accompanied by a separate perfect auxiliary have. They originally include modal and perfective features, but for some speakers, come to select an infinitive instead of a participle, so with only modal features. The change proceeds as in (136). (136) a. must have seen > musta seen b. musta seen > musta have seen c. musta seen > musta see

= grammaticalization = renewal

Micro cycles

79

The first stage, that is, (136a), is the one standard English is: most written texts still spell the auxiliaries separately, but spoken versions contract the perfect. Stage (136b), that is, the reduction of the auxiliary have before past participles, has occurred since the fourteenth century, as (137) shows. (137) a. þat wald a don me o mi lijf that would have done me of my life ‘that would have taken my life’ (c1300, Cursor Mundi, Cotton 5284). b. Syr, in good fayth Y am sory therefor, for and Y had west that ye would a taked so sor y ‘Sir, in good faith, I am sorry for it, if I had known that you would have taken it so badly’ (Cely Letters, William Maryon, anno 1480). c. there xuld not a be do so mykele. ‘There should not have been done so much’ (Paston Letters, I, #205, anno 1469, Margaret Paston). d. þat he myght ansowlde all yowr woll ‘that he might have sold all your wool’ (Cely Letters, George Cely, anno 1478). The current stage of English has counterparts to (137ab), as shown in (138) to (142), where a modal that is contiguous to -a is grammatical. (138) This is like a glimpse of the world America coulda built for itself (Twitter 2020). (139) They shoulda never let us know how much it bothers them (Twitter 2020). (140) Yeah I think I musta seen the same one (Twitter 2020). (141) I think he mighta had a breakthrough in understanding (Twitter 2020). (142) The memories I woulda made with these dudes if I knew (Twitter 2020). The contraction in (138) to (142) occurs with all epistemic modals in spoken and written English. The full pronunciation of have is unusual, so stage (136b) has definitely been reached. At this stage, the modal is like the other core modals and not like semi-auxiliaries, such as have to and ought to in that do-insertion is not necessary with negatives, as in (143), but it is with semi-auxiliaries, as in (144). (143) a. ‘if only he woulda not done one more caper’ (https://books.google. com/books?isbn=1257357123). b. ‘he just coulda not told the media about his disappointment with his teammates’ (Twitter 2018).

80

Micro cycles

(144) a. *He has to not go b. He doesn’t have to go. However, inversion in questions, as from (145a) to (145b), is very infrequent according to a Google search, and native speakers find it awkward; it is much more acceptable with an additional have, as in (145b). (145) a. ‘what coulda i done better to prevent this from happenin’ (Twitter 2014). b. ‘all you can do is think, what coulda I have done better’ (Twitter 2010). This ‘doubling’ of modals ending in -a/-of by have, as in (146) to (150), represents stage (136c) and is judged fine by many native speakers (of American English) with all five of the ‘new’ modals. (146) a. ‘And it’s not like I coulda not have killed a gastropod’ (2015 https://steamcommunity.com/app/105600/discussions/0/53064 6080846937998/). b. ‘I’m not really mad about it, personally I think Madden and Fifa coulda have been passed on’ (COCA Web 2012). (147) a. ‘I shoulda just have kept her as a friend’ (Twitter 2019). b. ‘There is no such thing as a respiratory heart rate. Oh. Maybe I shoulda have went to Columbia’ (COCA TV 2003). (148) a. ‘This research musta have been made just after sehri’ (Twitter 2018). b. ‘Guy comes in here with 2 posts and says the Guvnor he thought was a good guy . . . hhmmm . . . you musta have not watched this show before, seen any previews either that or . . . troll . . . or worse . . .’ (COCA Blog 2012). (149) a. ‘it mighta have created the trolling’ (Twitter 2017). b. ‘Any idea of who he might of have been planning to see?’ (COCA TV 1990). (150) a. ‘Woulda have missed it if not for this tweet!’ (Twitter 2018). b. ‘You would of have slept during anyway’ (COCA TV 1997). With the doubling, there is evidence of Subject Auxiliary Inversion in (151). Their grammaticality is confirmed by native speakers that allow ‘double’ have. This shows the modal and -a are now a real unit. (151) a. ‘But What Message Coulda I Have Sent Behind That’ (Twitter 2020). b. ‘How coulda I have been so stupid to trust the experts!’ (Twitter 2019). c. ‘Shoulda I have a chimichanga or a subway sandwich’ (Twitter 2019).

Micro cycles

81

d. ‘what coulda he have done to prevent someone from just walking tho?’ (www.reddit.com/r/ . . . /to_the_person_who_walked_out_on_their_ 700_tab). e. ‘What coulda I have been thinking of’ (Twitter 2013). The doubling in (146) to (150) paves the way for a new development, namely, for the modal ending in -a (-of) to now be followed by an infinitive with a nonperfect reading. That is stage (136c). The appearance of an infinitival form of a lexical verb to the modal’s right instead of a participle strikes most native speakers as odd but occurs in corpora and social media for all of the five modals, as in (152) to (156). In these, a perfect auxiliary ‘have’ is not implied and some modals are deontic, e.g. (154). This is evidence for a renewed modal class, again in varieties of American English. (152) ‘Yes, I woulda agree its time for Laura Hoffmiester to get off the council’ (COCA – iWEB). (153) ‘i shoulda be in Las vegas’ (Twitter 2019). (154) ‘Then . . . playing with the shots you coulda obtain ‘certain’ effects’ (COCA – iWEB). (155) ‘I mighta get a backpack instead, it’ll make . . . ’ (Twitter 2018). (156) ‘Manholes musta be constructed of two-course brick masonry laid in cement mortar . . . ’ (1911, City Council of the City of Bloomington, Illinois). The new modals precede negatives, as in (157) to (161), which shows they pattern with the core negatives and not the semi-auxiliaries. (157) ‘What would you attempt to do if ypu [sic] knew you wouldanot fail?’ (Twitter 2016) (158) ‘Today I learned two valuable lessons: 1) Petting an animal can lower your heart rate. 2) That animal shouldanot be a wolverine’ (Twitter 2017). (159) ‘What would you attempt to do if you knew you couldanot fail?’ (Twitter 2016). (160) a. ‘I mightanot be the same But that’s not important’ (Twitter 2016). b. ‘Your customers mightanot be on Twitter. Use raplemf to find them’ (Twitter 2016). (161) ‘You mustanot lose faith in humanity’ (Twitter 2017). (162) ‘Harden coulda probably be a pinch runner for the Astros’ (Twitter 2015).

82

Micro cycles

These modals also participate in SAI, as (163) shows. Although this use is rare in corpora, on Google, and on Twitter, native speakers who accept modals in -a/-of don’t think these are ungrammatical. (163) a. ‘When shoulda I expect your call?’ (Twitter 2019). b. ‘If you can change anything about you what woulda it be?’ (Twitter 2020). c. ‘What shoulda he be doing?’ (Carol Adams, Living Among Meat Eaters). Concluding the discussion on the recent renewal in modals, I have shown that mighta, coulda, woulda, shoulda, and musta may be forming a new class of modal auxiliary. This change is possibly helped by the analogy to hafta, oughta, wanna, seemsta, and liketa, although the new modals continue to behave like the core ones in terms of Subject Auxiliary Inversion, unlike the semi-modals. For many speakers, an additional have is possible, as in (146) to (150), and for some, the modal selects an infinitive to its right, as in (152) to (156). In Section 5, I have discussed (parts of) Modal Cycles, where verbs reanalyze as auxiliaries, as in the history of English and in varieties of Arabic. The original verb can then be lost and renewed as happened in the case of willan ‘want.’ The case of Arabic shows that grammaticalization occurs before renewal. The history of Slovenian shows that modal auxiliaries can be renewed by adverbs and then lost. This shows renewal is first. The English semi-modals and the new modals ending in -a show a renewal of the core modals, but the loss and renewal seems to coincide. 6

Voice cycles

In this section, I provide examples of how passive voice markers arise, are lost, and renewed. In a passive, the Theme argument is promoted (to subject) and the Agent argument demoted (to adjunct). Like many other grammatical categories, passives need not be marked morphologically. According to Siewierska (2013a), passive constructions occur in less than half of the languages examined, mostly in Eurasia and Africa, less frequently in Southeast Asia and the Pacific, rarely in Australia, and do not occur in New Guinea. When marked, passives are expressed through morphological means whose sources are pronouns, reflexives, reciprocals, nominalizations, and lexical verbs, according to Givón (1979 [2018]), Haspelmath (1990), and Zúñiga and Kittilä (2019: 226), among others. Antipassive markers demote the Theme (object) and promote the Agent (subject) in ergative languages. In (164a), the subject is marked by ergative and the object in the absolutive; in the antipassive in (164b), the subject is now marked by the absolutive, and the object is demoted to an instrumental case.

Micro cycles

(164) a. ʔaaček-a youth-ERG b. ʔaaček-ət youth-ABS ‘The young men Polinsky 2013).

83

kimitʔ-ən ne-nlʔetet-ən Chukchi load-ABS 3P.SU-carry-aor.3S.OB ine-nlʔetet-gʔe-t kimitʔ-e ANTIP-carry-aor.3S.SU-P load-instr carried away the/a load’ (Kozinsky et al. 1988, from

Since fewer languages are ergative, fewer languages mark the antipassive (Polinsky 2013). When they do, these markers arise from pronouns, reflexives, reciprocals, generic nouns, and nominalizations (cf. Sansò 2017). The middle voice is a mixture between the active and the passive voice in which the Theme is playing the role of subject and has some agentivity, as in the English middle in (165). (165) Donuts sell well. English doesn’t mark the middle morphologically but in languages that do, these markers have sources in common with passives and antipassives, that is, reflexives, spatial verbs, nominalizers and verbalizers, and aspect markers (cf. Inglese to appear). Since antipassives and middles spring from similar origins as the passives, I will mainly focus on the passive markers in this section. Exceptions are a sketch of the Middle Voice Cycle in Hungarian and the grammaticalization of a generic noun into an antipassive because that is not a source for the passive. Haspelmath sees the development of the passive as unidirectional, as shown in Figure 3.7. The sources on the left are renewed if they are lost in that function, and that results in a cycle. Sansò (2017) and Inglese (to appear) find the same unidirectionality for antipassives and middles, respectively. In this section, I will focus on three cyclical sources of passives. First, I discuss transitive constructions reanalyzing as passives and antipassives, via object (reflexive) pronouns to markers of voice in Scandinavian, Indonesian,

causative reflexive causative reflexive plural/general

generalized subject

pronoun FIGURE 3.7

anticausative

Sources of passives

(adapted from Haspelmath 1990: 54)

passive

zero

84

Micro cycles

and Tuvan. Second, I examine unaccusative verbs becoming passive markers, in a change from a lexical verb to a passive auxiliary. A third source is nominalizations, and for this, an example is provided from Western Austronesian. After this, I discuss recent work on a Middle Voice Cycle in Hungarian and the change of a noun to an antipassive. I conclude with a discussion of some counterevidence that change in voice is unidirectional. In the history of Scandinavian, a reflexive object, as in (166) and (167), has been reanalyzed as passive in (168) and (169). The original reflexive is strengthened by ‘self’ in some environments. (166) hann nefndi sik Ola he called self Oli ‘He called himself Oli’ (Faarlund 2004: 149). (167) sumir hofðu sik sjalfa deydda some. NOM had REFL self killed ‘Some had killed themselves’ (Faarlund 2004: 90). (168) kalla-sk call-3S ‘He calls himself; he is called’ (Ottosson 2008). (169) Det dansad-es hela natten it dance-PASS whole night ‘There was dancing the entire night.’

Old Norse Old Norse Old Norse Swedish

Other language families show a similar development. Guilfoyle et al. (1992) have argued that the Indonesian passive di- prefix in (170) has its origin in the third person singular pronoun dia ‘s/he,’ also shown in (170). The construction is therefore limited to third person agents. (170) saya di-jemput oleh dia I PASS-met by 3S ‘I was met by him/her’ (Sneddon 1996: 248).

Indonesian

Currently, Koh (1990) and Gil (2002) have argued that Malay/Indonesian is losing the prefixes meN- and di-, for active and passive, respectively. As far as I know, there is no replacement for these. A third example of a reflexive as a source for a voice marker comes from Tuvan. Sansò (2017) shows that, in this language, antipassives arise from four main sources: agent nominalizations, generic/indefinite objects, action nominalizations, and reflexive/reciprocal morphemes. The last source is the most frequent. An example of this is given in (171), where the reflexive is the same as the antipassive in (172).

Micro cycles

85

(171) ol savaƞna-n -də Tuvan he soap-REFL-3.PST ‘He soaped himself’ (from Sansò 2017: 195, based on Kuular 2007). (172) ava-m am daara-n -əp tur Tuvan mother-my now sew-AP-ConV AUX ‘My mother is sewing now’ (from Sansò 2017: 195, based on Kuular 2007). The Voice Cycle involving a pronoun can be schematized as (173). (173)

a. reflexive/personal pronoun b. new reflexive/pronoun

>

passive >

zero

The second major source of voice markers is from unaccusative to passive (e.g. Michaelis 1997). Haspelmath’s (1990) examples include the Korean passive – ji originates as the unaccusative ji ‘fall’ and, likewise, the Tamil passive – pat derives from paṭu ‘fall, happen.’ In Germanic, the source of the passive is the verb meaning ‘become.’ In the history of English, two passives competed with each other, the beon ‘be’ passive and the wearðan ‘become’ passive. They are both used in (174), and both are, of course, unaccusative in origin. (174) Ær þæm þe Romeburg getimbred wære . . . Ere that REL Rome built was oð him se mæsta dæl wearð underþieded. until him that most part became subjugated ‘Before Rome was built, [Vesoges was fighting . . . ], until to him most of it was subjugated’ (c900, Orosius, 28.22). Many have argued there is a meaning difference between the two: the be-passive would be used for stative situations and the become-passive for dynamic ones. However, since they are often used with the same verb, it is hard to regard that as an absolute rule, and this is probably the reason the become-passive is lost (e.g. Brunner 1962, but see Mailhammer et al. 2013). The actual change from an unaccusative, such as (175), to a passive auxiliary in (174) is complex if done in one step: where would the (second) lexical verb in the passive come from? (175) Hlynn wearð on ceastrum noise happened in fortresses ‘There was tumult in the fortresses’(OED, c1000, Junius MS, Genesis A 2548).

86

Micro cycles

It is therefore more likely that the transition is from unaccusative to copula to passive. Wearðan is a copula as well, as (176) shows, and therefore this transition would be easier because the subject predicate can be a participle, as in (177). (176) he wearþ færinga geong cniht he became suddenly young boy ‘He suddenly became a young boy’ (OED, c950, Blickling Homilies 175). (177) Þæt cweartern wearð afylled mid fulum adelan that prison became filled with foul mud ‘The prison became/was filled with fould mud’ (OED, c1000, Ælfric Lives of Saints (Julius) II. 392). Such a change from copula to passive is happening in contemporary Dutch (178) to (180), where a copula can be used with a participle in a semi-passive function (see Verhagen 1992: 318). (178) Dat staat aangekondigd/vermeld. that stands announced/mentioned ‘That is/has been announced/mentioned.’ (179) Dat ligt begraven/uitgestald. That lies buried/displayed ‘That has been buried/displayed.’ (180) Dat zit beklemd. That sits stuck ‘That is stuck.’

Dutch

This unaccusative to passive cycle can be represented as in (181). The grammaticalization involves a change in semantic meaning and a loss in theta roles. (181) lexical verb > ‘happen’ (Theme)

copula > ‘become’ (Theme)

auxiliary ‘be’

The grammaticalization of the verb ‘finish’ provides another example of an unaccusative source of the passive. Alasmari (2022) shows an interesting bridging context that’s not copular but where the subject is a deverbal noun that can be reanalyzed as the lexical verb. In Modern Standard Arabic, the verb tamma/yatimmu ‘be finished’ is ambiguous between a lexical verb and an auxiliary in (182). Thus, “the writing of the article was finished” is the earlier meaning and “the article was written yesterday” is the grammaticalized one. Alasmari argues that the latter is the correct analysis. (182) Tamm-at kitab-at-u l-maqal-i amsi MSA be.finished-3FS writing-FEM-NOM the-article-GEN yesterday ‘The article was written yesterday’ (Alasmari 2022: 32).

Micro cycles

87

Nominalizations are a third source of voice morphology. Synthetizing work by Wolff (1973), Ferrell (1982), and Starosta et al. (1981), Ross (2002) shows this for Western Austronesian. In one of those languages, Paiwan, the four voices are as shown in (183) to (186), where PV marks Patient passivization, LV Locative passivization, CV Instrument passivization, and AV Agent prominence. (183) təkəl-ən a vaua Paiwan drink-PV SPEC wine ‘The wine will be drunk.’ (184) təkəl-an a kakəsan drink-LV SPEC kitchen ‘The kitchen will be drunk in.’ (185) si-təkəl a kupu CV-drink SPEC cup ‘The cup will be drunk with.’ (186) t-əm -kəl a qała AV-drink SPEC stranger ‘The stranger will drink (something)’ (Ross 2002: 20–21). The nominalizing morphology in the same language is remarkably similar, as shown in (187) to (190) for the verb kan ‘to eat.’ (187) (188) (189) (190)

kan-ən ‘food,’ i.e. something to be eaten (Patient) kan-an ‘place where on eats’ (Location) si-kan ‘eating utensil’ (Instrument) k-əm -an ‘eater’ (Ross 2002: 38) (Agent)

Paiwan

According to Ross, this means that (183) to (186) originate in the nominalizations of (187) to (190). The nominalization in (191) is reinterpreted as voice morphology, in a change as in (192). (191) təkəl-ən a vaua Paiwan drink-PV SPEC wine ‘Something to be drunk is the wine’ (adapted from Ross 2002: 39). (192) V-NOM > V+PASS The reconstructed voice affixes for Proto Austronesian are *-en, *-an, *si-, and *um, respectively, according to Ross (2002: 33). He also provides an alternative reconstruction without deciding on one. Having provided examples of three major sources of voice markers, I now turn to another type of voice marker that is lost and replaced. Halm (2020) argues that the history of Hungarian shows a Voice Cycle. On the basis of linguistic fossils and evidence from related languages, he argues that the middle voice head

88

Micro cycles

is *-v, shown in (193). The subject agreement is special in the case of non-active verbs, the so-called ik-paradigm, which is also shown in (193). (193) *tör-v-ik break-mid-3Smid ‘gets broken’ (Halm 2020: 11).

Reconstructed Proto-Hungarian

This voice marker *-v disappears by the time of Old Hungarian, which marks the active and middle voice just through “allomorphy in the subject agreement suffix conditional on the feature content of a silent Voice head.” This alternation is shown in (194), where -ʃa marks the active and -ik the middle. (194) a. hanēčac hog labait moʃ-ʃa Old Hungarian rather:only that foot.3S wash-3Sdef ‘Rather that he washes only his feet’ (Halm 2020: 7; Munich Codex (1466), 100ra [John 13:10]) b. meǵ-moſ-d-ik uala PRT-wash-freq-3Smid be.pst3S ‘He washed himself’ (Halm 2020: 7; Vienna Codex (mid-15th C), 35 [Judith 7:12]). Modern Hungarian innovates a designated voice marker, as shown by the alternation in (195). (195) a. old- Ø Modern Hungarian loosen-3Sindef ‘Somebody loosens something.’ b. old-ód -ikə loosen-MID-3Smiddle ‘Something gets loosened’ (Halm 2020: 2). This Modern Hungarian -ód- middle marker derives from a frequentive aspect, which was reanalyzed as voice because frequentives are often passive/middle, according to Halm. So this example constitutes a case of voice loss and subsequent renewal. Middles need not be marked, as is clear from the history of Hungarian and the example I gave in the beginning of this section. Middles have a lot in common with passives and unaccusatives in that the object is singled out, for example, the books in (196). Some languages have a special verbal marker, but English lacks this. Thus, Latin has an -r suffix that marks the middle, as in (197). (196) These books sell well.

middle

Micro cycles

(197) Miro-r admire-MID ‘I admire.’

89

Latin

Turning to antipassive markers in Tibeto-Burman languages, Bickel and Gaenszle (2015) propose that the first step in the development [of the antipassive] was in fact the use of the original noun in an antipassive environment where it was assigned nonquantifiable and nonspecific reference, i.e., a notion of ‘people’ in a generic sense rather than a specific person. They also mention a bridge construction, which is attested in Belhare where the noun maʔi ‘person’ is found in both an active and antipassive construction in (198ab) and “in the function of a regular marker for first person (exclusive) objects [in (198c)]” (Bickel & Gaenszle 2015: 67). (198) a. un-na maʔi niu-t-u. (active) Belhare 3s-ERG person[sNOM] [3sA-]see-NPST-3O ‘S/he sees a (specific) person’ or ‘S/he sees the person.’ b. un maʔi ni-yu. (antipassive) 3sNOM person[sNOM] [3sS-]see-NPST ‘S/he sees people’ but not *‘S/he sees the/a (specific) person.’ c. un-na maʔi-ni-yu. (first person object agreement) 3s-ERG eP-see-NPST ‘S/he sees us (e)’ (Bickel & Gaenszle 2015: 67). So the personal pronoun and antipassive markers derive from the same origin, that is, a generic noun. As in the case of reflexive pronouns turning into passive markers, seen in (166) to (169), one of the arguments is reanalyzed as a marker of an intransitive state. Auderset (2021) provides many more instances of overlaps between antipassives and first person plurals that derive from generic nouns. An example of such a change can be found in Saliba, an Oceanic language, where the first person plural kai in (199) is the same as the antipassive in (200). (199) kai-wa ka-matausi palapa Saliba 1P.EXCL-DET 1P.EXCL-be.frightened really ‘We were really frightened.’ (200) ya-lao ya-kai -deuli 1S.NOM-go 1S.NOM-ANTIP-wash ‘I go and wash the laundry/dishes’ (Auderset 2021: 392, based on Mosel 1994: 6 and Margetts 1999: 182).

90

Micro cycles

I will end this section with research that argues that changes in voice are not unidirectional. Grestenberger (2020) argues that changes in voice marking can go in either direction because, due to the ambiguity of deverbal participles and adjectives between stative and eventive semantics, the language learner can incorporate different voice heads, for example, ASP, v, or Voice. She calls this ‘Voice flipping.’ In the change from Ancient Greek ‑menos to Modern Greek -menos, the participle goes from active to passive, and in the change from Proto‑Indo‑ European *-nt- and Hittite -ant‑ to Ancient Greek ‑nt-, it goes from resultative/ stative to active. A voice flip may not count as a counterexample to unidirectionality since these changes do not involve grammaticalization but a lateral change from one type of voice marker to another. More worrisome for the cycle are examples where passive voice morphology is the source for pronouns and reflexives. Bahrt (2021) reviews syncretisms between voice and other markers, for example, reflexive and reciprocal syncretisms. As for unidirectionality, he concludes that, although “sparse,” “there is some evidence suggesting that passive voice marking can potentially develop a reflexive, reciprocal, or anticausative function” (208). The example of passive to reflexive is restricted to Tarahumara, where Bahrt notes a limited use of the reflexive, and to Ts’amakko, an east Cushitic language where this development is also a possibility. The change from passive to reciprocal may have occurred in Sidaama, another East Cushitic language. In conclusion, in this section, I have first discussed three types of sources for Voice Cycles (pronouns, unaccusatives, and nominalizations). Languages can lose voice marking without needing immediate renewal. This is seen in the loss and (later) replacement in Hungarian of a middle. Finally, I discuss a source for an antipassive typical perhaps for an antipassive, namely, an indefinite noun. I have also discussed some apparent counterexamples to unidirectionality. In one case, it can be argued that there is not grammaticalization but just reanalysis as another voice marker; in the other case (voice to pronoun), the evidence is genuine but scarce. The cycles discussed in this section involve grammatical voice; I have not examined where lexical voice such as causative and anticausative come from (see e.g. Song 1990; Jacques 2015; Operstein 2014). 7

Conclusions

This chapter has provided a definition of micro cycles and given several examples of such cycles. The Determiner Cycle takes a demonstrative and reanalyzes it as an article, and the determiner may itself be renewed by a deictic element. The article may also be lost in the process. The Copula Cycle has a number of sources of renewal, namely, demonstratives, posture verbs, unaccusatives, and locational particles. TMA cycles are frequent, and I have given a few examples of the Perfective and Imperfective Cycles and of Modal Cycles. Voice Cycles

Micro cycles

91

finish off the chapter: renewal sources are varied, and the loss of a voice marker is sometimes seen (e.g. Malay and Hungarian). Does grammaticalization or renewal start a micro cycle? Both scenarios are possible. In certain cases, there is a long time gap between the loss of a lexical item and its renewal, for example, when English demonstratives reanalyze as articles, the deictic renewal occurs much later. Copulas are also lost before they are renewed. This is evidence that grammaticalization starts the cycle. In other cases, the two occur at the same time, and then renewal may be starting the cycle, such as when Slovenian modal auxiliaries are renewed by adverbs and then lost. The next chapter continues with four other micro cycles. Suggestions for further reading Determiner Cycles: Greenberg (1978) and Diessel (1999b). Copula Cycles: Katz (1996), Tan (2020; 2022), and Truppi (2021); Gómez Vázquez and Mateu (2022) on the change from posture verbs to copula. Change in TMA: Bybee and Dahl (1989), Bybee et al. (1994), Kiss (2017), and Chatzopoulou (2021). Voice Cycles: Haspelmath (1990) and Zúñiga and Kittilä (2019).

Review questions and exercises

1. 2. 3.

4.

Describe the Copula Cycle in your own words. As we saw in this chapter, some languages use copulas and others don’t. Why do you think the copula is optional, and what implications does that have on Copula Cycles? In Chapter 1 and in Section 2 of this chapter, we looked at changes in English articles and distal demonstratives. Is there a research question you can formulate? What data would you use to answer your question? Take a look at the Old Norse (201) and (202), a precursor of Swedish (3). What could have happened? (201) Hinn er sæll er sér of getr lof . . . Old Norse DEM is happy REL REFL PRT gets fame ‘He is happy who (for) himself obtains fame’ (Edda, Hávamál 8, https://etext.old.no/Bugge/havamal.html). (202) ok hinn siðasta vetr er hann var í Nóregi Old Norse and that last winterthat he was in Norway ‘and the last winter that he was in Norway . . . ’ (Bjarni’s Voyage 41.8, Gordon 1956). In (203) to (205), some sentences from Russian are provided. Is there evidence of a cycle? Which of the forms (sebja and -sja) do you hypothesize is archaic, the morpheme or word, and why? (203) Ivan uvidel sebja Russian I saw himself ‘Ivan saw himself.’

92

Micro cycles

(204)

5.

Ivan moet-sja Ivan wash-REFL ‘Ivan washes himself.’ (205) Eta kniga legko citaet-sja . this book easily read.PR.3S-REFL This book is easy to read. (Ahn 2006: 94) In English, the -ing is also used for habitual aspect, as Kranich (2010) shows. (206) Mary’s working at the library this semester (Kranich 2010: 32). Thinking back to Section 4, is this expected? Looking at Figure 3.4 might help as well.

Advanced exercise

Does the language you have selected in Chapter 1 have the following: Have articles; if yes, are they related to demonstratives, and do the demonstrative show renewal? Make a table to put all forms in, and if you have access to a sister language, compare it with articles and demonstratives in that language. Have copulas and can you see evidence of what the origin of the copula is? Look at the ‘usual suspects,’ for example, demonstratives, prepositions, and adverbs. Have tense and aspect, and how are they marked (morphology or syntax)? Is there evidence of recent changes, for example, in a new progressive? Anything else that strikes you, for example, about mood or voice?

4 MICRO CYCLES Polarity and discourse cycles

In this chapter, I continue the discussion of micro cycles by examining cycles that involve polarity. In language, polar opposites can be found in positive and negative statements and in answers to yes/no questions. The chapter starts with the most well-known of all cycles, namely, the Negative Cycle. In fact, there are several different negative cycles, those whose source is adverbial or nominal and those whose source is verbal. Negatives in turn can be used as a source for the renewal of yes/no questions, as can wh-elements. Two other micro cycles to be discussed are Complementizer and Pragmatic Cycles. The former come about from the preposing of adverbials that are then reanalyzed as complementizers where the latter reanalyze adverbs as discourse markers. Complementizer and Pragmatic Cycles are related to each other in that PPs and adverbs become complementizers and discourse markers for pragmatic reasons. As I have mentioned in the previous chapter, the Copula and Negative Cycles interact with each other: the negative copula can be adopted as a negative, and then a new copula may appear. The same is true for the Perfective and Imperfective Cycles. There are other micro cyclical interactions that will be discussed in this chapter, for example, the interaction of TMA auxiliaries and negation and that of the Negative and Interrogative Cycles. Macro cycles interact as well, and I discuss that for the Pronoun Cycle in the next chapter. Some of the data on the Negative, Interrogative, and Complementizer Cycles can also be found in van Gelderen (2011). The outline is as follows. In Section 1, I discuss three types of Negative Cycles. Section 2 considers another polarity phenomenon, namely, interrogatives. Section 3 examines Complementizer Cycles, and Section 4 turns to the DOI: 10.4324/9781003272564-4

94

Micro cycles

Pragmatic Cycle. Section 5 reviews interactions between micro cycles, and Section 6 is a conclusion. 1

Negative cycles

Negation is universally marked in all languages although the means that languages use to express negation are pretty varied, even in closely related languages. The reason is that negation can change fast, and what’s an independent negative in one language may have turned into an affix in another or have been completely replaced. Yuman languages, for instance, have quite different negative forms: waly . . . ma in Piipaash (1), -mo(t) in Mohave (2), and ma:w in Diegueño/Kumeyaay (3). (1) (2) (3)

waly-‘-tpuy-ma -k Piipaash NEG-1-kill-NEG-REAL ‘I didn’t kill him’ (Gordon 1980: 65). ʔnyeč ʔ-iyem-mo -t-m Mohave 1S 1-go-NEG-EMPH-REAL ‘I didn’t go’ (Munro 1976: 106 with gloss adapted). ʔ-a:m-x ʔ-ma:w Diegueño/Kumeyaay 1-go-IRR 1-NEG ‘I don’t/didn’t go’ (Langdon 1970, cited after Munro 1974).

The diversity occurs not only in the form but also in how the negative attaches, bound in (1) and free in (3), and in whether it is one marker, as in (2) and (3), or more than one, as in (1). In this section, I relate changes in negative marking to cycles of change. There are (at least) three types of negative cycles. The most well-known is also called the Jespersen Cycle, after Jespersen (1917). Another was identified by Givón (1978) using verbs with negative meaning as a source for a grammatical negative marker and a third by Croft (1991), which is known as the Negative Existential Cycle (NEC). Jespersen’s Cycle relies on nominal or adverbial material to renew the negative, whereas Givón’s and Croft’s Cycles rely on verbs for their renewal and work best when these verbs don’t have too many grammatical features, as we’ll see; Jespersen’s Cycle is not affected by these features. In this section, I discuss all three cycles. 1.1 Jespersen’s negative cycle

The Jespersen Cycle involves a negative that weakens and is replaced by another expression, as (4), repeated from Chapter 1, for Old English ne, which is reinforced and then replaced by a negative not.

Micro cycles

(4)

95

The Negative Cycle Old English Middle English later Middle English Modern English ne > ne . . . na wiht/not > not > not/-n’t . . . nothing NEG NEG NP/ADVERB NEG NEG NP

In Jespersen’s words, the original negative adverb is first weakened, then found insufficient and therefore strengthened, generally through some additional word, and this in turn may be felt as the negative proper and may then in the course of time be subject to the same development as the original word. (Jespersen 1917: 4) Ingham (2006, 2013: 130) provides support for the scenario in (1): ne doesn’t only weaken, evidenced by the contration with a verb, in the presence of a second negative, as in (5), but also in contexts without another negative, as in (6) and (7), so it is a general weakening, not a pragmatically induced one. (5)

(6)

(7)

Ac nis nan scild trum[ra] wið ðæt tuiefalde but NEG.is no shield stronger against that twofold gesuinc ðonne mon sie untwiefeald toil than man be sincere ‘But there is no stronger shield against the twofold toil than being sincere’ (c900, Alfred, Pastoral Care 239.9 Hatton). Heo axodon him ða: And hwær is he nu? They asked him then and where is he now He cwæð þæt he nuste. He said that he NEG.knew ‘They asked him ‘where is he now?’ He said that he didn’t know’ (DOE, c1200, Homilies 36). nis þæt seldguma wæpnum geweorðad NEG.is that retainer weapons honored ‘This is not a mere man honored with weapons’ (c1000, Beowulf 249–250).

The earliest doubled negatives are with a doubled negative argument (see van Gelderen 2004b: 82–83), as the object nowiht in (8) shows, repeated from Chapter 1. (8)

And þa æfter micelre tide sæt ic ana in þam and then after much time sat I alone in that westenne and ic ne geseah nowiht buton eorðan. desert and I NEG saw nothing except earth ‘And after a long time, I was alone in the desert and saw nothing except earth’ (DOE, c1000, Lives of Saints, 310–311).

96

Micro cycles

The negative argument in (8) subsequently starts to function as an adjunct, and the question arises how it does so. One possibly bridging context is with optionally transitive verbs, for example, eat and drink, as Lucas (2007), Willis et al. (2013: 148 for Arabic), and Willis (2016) show. For Old English, Willis (2016: 476) examines 121 examples of nawiht and finds that of the 45 cases where nawiht is a direct object, four involve an optional transitive, as in (9), three are objects of modals, and 14 are objects of extent, as in (10), so optional as well. (9) ne con ic noht singan NEG can I nothing sing ‘I can’t sing’ or ‘I can sing nothing’ (c950, Bede 4.25.342, Willis 2016: 476). (10) & he nowiht fromade in his lære and he nothing succeded in his teaching ‘And he didn’t succeed in his teaching’ or ‘And he succeeded not far in his teaching’ (c950, Bede 3.3.162, Willis 2016: 478). These are ambiguous between the renewing negative being an argument or an adverbial and could be the trigger for the change. As Breitbart et al. (2020: 62) note, these numbers render the bridging contexts 17.4%. In addition to the examples of English given in Chapter 1, the Jespersen Cycle has been described for Latin, French, Basque, Bantu (Devos at al. 2010), Mayan (Romero 2012; van der Auwera & Vossen 2016), Quetchuan, and Maipurean (van der Auwera & Vossen 2016), just to mention a few. I’ll just sketch a lesser known example here, that of Mayan, as described in van der Auwera and Vossen (2016). As documented by Romero (2012), in the sixteenth century, there was a preverbal ma (with an optional clitic k’u), as in (11), that is optionally renewed by the irrealis marker ta(x), as in (12). (11) Mak’u ʃucinik ʃetʃ’awik Classical, 16th century K’iche’ ma-k’u ʃ-∅-ucin-ik ʃ-e-tʃ’aw-ik NEG-CL C-3S.A -succeed-PF C-3P.A-speak-PF ‘But they did not succeed in speaking’ (Romero 2012: 82; van der Auwera & Vossen 2016: 194). (12) man kawil tax 16th century K’iche’ man(a) k-∅-a:w-il tax NEG INC-3S.A-2S.E-see IRR ‘You didn’t see it at all’ (Romero 2012: 87; van der Auwera & Vossen 2016: 194). By the beginning of the twentieth century, ta(x) becomes obligatory and man(a) is optional, as in (13).

Micro cycles

97

(13) (man) w-etaman ta u-watʃ Modern K’iche’ NEG 1S.E-know NEG 3S.E-face ‘I don’t know him/her’ (Romero 2012: 85; van der Auwera & Vossen 2016: 195). The stage of obligatory pre-verbal ma is followed by one with an optional reinforcement by tax, which in turn becomes the main negator, a classical example of the Jespersen Cycle. As to how the cycle starts, the irrealis marker tax typically occurs with the negative marker mana, which adds emphasis, so renewal rather than grammaticalization drives this change. Thus, the motivation for the Negative Cycle can also be seen as pragmatically driven, as in (14), so starting on the right of (4), with pragmatic strengthening, and instigating the phonetic weakening that starts on the left side of (4). (14) The Negative Cycle a. new emphatic negative b. weakening/loss of regular negative c. emphatic negative > regular negative d. new emphatic negative For instance, Kiparsky and Condoravdi (2006), in examining Jespersen’s Cycle in Greek, suggest pragmatic and semantic reasons for the renewal. A simple negative cannot be emphatic; in order for a negative to be emphatic, it needs to be reinforced, for instance, by a minimizer. Adapting ideas from Dahl (2001), they argue that, when emphatic negatives are overused, their semantic impact weakens, and they become the regular negative, and a new emphatic will appear, as in (14cd). Larrivée (2010), examining the history of French negation, argues that a specific pragmatic function, namely, accessibility of a proposition to the hearer, plays a role. Thus, as in the case of other cycles we’ve encountered, the Negative Cycle can start either through weakening of the semantics or phonology or by strengthening the negative. In a similar way Breitbarth et al. (2020: 11–13) see neither weakening nor strengthening as driving forces for the (Negative) Cycle because there are languages with weak negatives that see no renewal and languages having reinforced negatives and no weakening. They see the cycle as an interplay of “pragmatic, more narrowly syntactic, and language external forces” (p. 13). 1.2 Givón’s negative cycle

As for the second type of Negative Cycle, Givón (1978: 89) writes “[n]egative markers in language most often arise, diachronically, from erstwhile negative main verbs, most commonly ‘refuse,’ ‘deny,’ ‘reject,’ ‘avoid,’ ‘fail,’ or ‘lack.’”

98

Micro cycles

In earlier work (Givón 1973: 917), he provides example verbs: English fail, Kihung’an – khona ‘refuse,’ and Bemba – bula ‘lack, miss,’ but no actual trajectories and how it starts. The motivation is pragmatic. Veselinova (2013) quotes the Kannada negative illa as derived from a Dravidian root ‘to die.’ In this section, I’ll provide examples from Chinese, English, and Koorete. The Chinese negative mei presents a well-known example of a verb meaning ‘to sink, die,’ as in (15), being reanalyzed as negative in (16) (Croft 1991: 11; Lin 2002; Shi & Li 2004; Yang 2012). (15) Yao Shun ji mo ... Old Chinese Yao Shun since died ‘Since Yao and Shun died . . . ’ (Mengzi, Tengwengong B, from Lin 2002: 5). (16) yu de wang ren mei kunan . . . Early Mandarin wish PRT died person not.EX suffering ‘If you wish that the deceased one has no suffering . . . ’ (Dunhuang Bianwen, from Lin 2002: 5–6). This change connects Givón’s Cycle to Croft’s Cycle and the Copula Cycle. In (17), a new existential is used, which can be contracted with the copula to become just a negative. (17) wo mei (you) shu I not EX book ‘I don’t have a book.’

Modern Chinese

This is indeed what Lam Chit Yu (2022) shows for Hong Kong Cantonese, where the negative mei and existential/possessive you merge phonetically as mou, and a new copula could develop. With the verb mei “to die,” the reanalysis of the negative verb in (15) is to a negative existential in (16) and then to a negative in (17) with the existential being renewed by you and subsequently merging with the negative in Cantonese. Clear cases where a negative verb would be reanalyzed as a negative might be the verb fail in English, as in (18). Here, ‘failed to’ can be replaced by ‘didn’t’ without loss of meaning. (18) [we] became so engrossed in our game of tetherball that we failed to hear the teacher calling us to return to the classroom (COCA magazine). Because the English negative -n’t is in the final stages of the Jespersen Cycle (i.e. it is often inaudible and speakers have to repeat it), we expect such a renewal because the alternative, renewal of the negative by a negative adverb, as in (19), is objected to so much by prescriptivists.

Micro cycles

FIGURE 4.1

99

Failed to and a verb in COCA: no real change

(19) ‘He don’t care about nothing but his car, rims, money’ (COCA spoken). However, the use of (18) is not frequent, and many of the failed to instances still have the meaning of ‘not be successful.’ Figure 4.1 provides some data on fail from American English since 1990. This figure shows little change in 30 years and that the spoken register lags behind, a sign that this change is not really in progress. The use of fail in the British National Corpus is even lower in the spoken registers though not in the written registers. Other negative verbs, for example, lack, don’t show an increase either, however. Negation in Koorete, an Omotic language of Ethiopia, is expressed by means of a negative marker ba, as in (20) and (21). (20) nen-i doro woon-do ba-nna-ko Koorete you-NOM sheep buy-PF NEG-2S-DEC ‘You didn’t buy sheep’ (Binyam 2008: 123). (21) is-i dana ush-iya ba-nni-ko Koorete she-NOM beer drink-PROG NEG-3FS-DEC ‘She is NOT drinking beer’ (Binyam 2008: 151). As Binyam (2008) shows, there is still a lexical verb ba in Koorete with the meaning ‘disappear,’ as in (22a), which can itself be negated, as in (22b). (22) a. is-i ba-d-o Koorete she-NOM disappear-PF-PST ‘She disappeared.’ b. is-I ba-d-o ba-nni-ko she-NOM disappear-PF-PST not.exist-3FS-DEC ‘She did NOT disappear’ (Binyam 2008: 150). Binyam considers ba a negative existential because it also means ‘not exist.’ That would mean it is an example of the Negative Existential Cycle, to be

100

Micro cycles

discussed next. However, because the verbal meaning is also negative in (22a), I consider it an example of Givón’s Cycle. There is also an emphatic negative, as in (23), using an extra adverbial petto ‘never’ indicating what would be a renewal through Jespersen’s Cycle. (23) es-i keele petto han-g-u-waa -s-so Koorete he-NOM Keele never go-IMF-PR-not.exist-3MS-AFOC:DEC ‘He will never go to Keele’ (Binyam 2008: 160). There are some puzzling aspects, for example, why the negative auxiliaries are inflected for person, gender, and number but not the main verbs, as seen in (22b) and (22a), respectively. The grammaticalization and renewals are clear, however. 1.3 Croft’s negative cycle, also known as the negative existential cycle

The third type of Negative Cycle, the Negative Existential Cycle (NEC), was so named by Croft (1991) and is often referred to as Croft’s Cycle. It was added to greatly by Veselinova (e.g. 2013; 2016). The basic cycle is given in Figure 4.2: Type A involves a stage where standard negation and existential negation are expressed by the same morpheme; Type B is (usually) where the negative has attached itself to the existential verb and is no longer the same as the standard negative; and Type C is where the Negative Existential of Type B is used for all negation, often with a null existential. Croft has intermediate stages as well but not in his figure. As for what starts the NEC, it is the cliticization of the negative to an (existential) verb, whose meaning is gradually lost. I’ll give a few examples of this cycle or part of this cycle. I show that it is not restricted to existential verbs but can involve auxiliaries and copulas as well, and this is important for the interaction of the negative and other cycles, that is, the Copula and TMA Cycles. Type A

Type B

One NEG

NEG ˜ NEG Existential

Type C NEG = NEG Existential FIGURE 4.2

The negative existential cycle (NEC)

(adapted from Croft 1991)

Micro cycles

101

The development of the Hindi/Urdu negative goes from the single na in Sanskrit (24) to a stage where na and nehĩ alternate to one where nehĩ is the main negative, as in present-day Hindi-Urdu (25). (24) Raamah vanam na gacchati Sanskrit Rama forest NEG goes ‘Rama doesn’t go to the forest.’ (25) . . . haakii aur krikeT ke maamle mE bhaarat-paakistaan Hindi hockey and cricket of matter in India Pakistan se acchaa anya koii mukaablaa nahII than good other any rivalry NEG ‘In the matter of cricket and hockey there is not a better rivalry than IndiaPakistan’ (from Lampp 2006: 8). In Kellogg’s account, -hĩ is a remnant of an auxiliary verb (1938: 281; 459); simple na therefore remains with non-indicatives, and a prohibitive mat occurs with imperatives in the modern language. I have put the changes involving na and nehĩ in Table 4.1, with the stages from Croft’s Cycle. The last stage is one where a ‘double’ auxiliary is appearing. One piece of synchronic evidence that nehĩ is formed from na and an earlier inflected form of the verb/auxiliary hona ‘to be’ is that copulas and auxiliaries, that is, typical uses of hona ‘to be,’ are not necessary with nehĩ, as (26) and (27) show, and are uncommon. (26) mẽ student nehĩ (hũ) (Hindi/Urdu) I student not am ‘I am not a student’ (data checked with Sakshi Jain). (27) mẽ yehã kam nehĩ karti (hũ) I here work not do am ‘I don’t (generally) work here’ (data checked with Sakshi Jain).

TABLE 4.1 The stages of the NEC from Sanskrit to Hindi/Urdu

Croft A B C C~A

stage Sanskrit Early Hindi/Urdu Hindi/Urdu change in Hindi/Urdu

negative na na or na hĩ [NEG + ‘be’] nehĩ (marginal na and mat) nehĩ nehĩ + hona ‘be’

102

Micro cycles

Currently, the last stage of the cycle is reached, and the copula and auxiliary are used in combination again, as in (28). (28) koi bhi Pakistani bharat me nehĩ rah raha hai (Hindi/Urdu) Any even Pakistani India in NEG live PROGR is ‘No Pakistani is living in India’ (Lampp 2006: 17, her transliteration). This ‘doubling’ of the auxiliary verb (in stage C~A) would be expected although, cross-linguistically, this stage is very rare. Auxiliary verbs typically add tense, mood, aspect, or voice and accompany a lexical verb. They may also agree with the subject and have the negative cliticized to it. However, when auxiliaries are heavily inflected for tense, aspect, mood, and agreement, the loss of that information will hinder their reanalysis as negatives. If the auxiliary is heavily inflected, the contracted forms will be many, and that also stops the reanalysis as a new negative. Knowing that hona is inflected, how was the reanalysis from stage A to B in Hindi/Urdu possible? Numerous scholars have argued there is a second source that may have helped the NEC along. Whitney (1889: 413), Turner (1966: 404), and Bashir (2006: 7), to name a few, have argued that na was strengthened with an emphatic hĩ, which is still around in the language. Since the paradigm of hona ‘to be’ shows many forms, hũ, ho, hẽ [1S, 2S, 3S], etc., it may also be that the presence of hĩ helped solidify the form nehĩ. Different cycles compete, and that is visible in a minimizer that is sporadically used as negative, for example, the one identified by Gul (2009), namely, thoRi ‘little,’ as in (29). When thoRi is negative, emphatic particles like si, hi, and tu cannot follow it, as in (b), according to Gul, and that is a way to distinguish the negative from its adverbial origin, possibly shedding doubt on the emphatic origin of -hĩ discussed previously. This renewal by a minimizer is typical of the Jespersen Cycle. (29) a. Usne thoRi bat ki (Hindi/Urdu) he NEG talk did ‘He didn’t talk’ (Gul 2009). b. wo BASHEER Thori Tha, wo Tou PAPA The. (Hindi/Urdu) he Basheer NEG was he EMPH papa was ‘He wasn’t Basheer, he really was daddy’ (mobiletextsms.blogspot. com/2011/08/wo-basheer-thori-tha.html). The verb hona ‘to be,’ according to Platts (1930: 1242), also means ‘to exist, subsist, be born’ and a variety of other meanings typical of existential verbs. However, in present-day Hindi/Urdu compounds like mowjud hona ‘be present’ or rehna ‘to live’ are used instead. Such renewal of verbs that participate in the NEC is expected.

Micro cycles

103

So far, I have shown that a copula and auxiliary can be the source of a reanalysis but that verbal inflection might be hindering the reanalysis as negative. What probably made it possible to see nehĩ as a negative particle in Hindi/Urdu, also an inflected language, rather than as verb is the independent existence of hĩ in the language. Next, I show how an inflected negative auxiliary can indeed lose the inflection in this process. There are stages in the history of English and present-day varieties where the inflected forms am not, are not, is not, etc. are reduced to ain’t, which is then used as a multipurpose negative copula and auxiliary, as in (30), for all persons. (30) a. I ain’t afeard o’ nyther on you copula ‘I am not afraid of neither of you’ (COHA fiction 1828). b. and when you ain’t got any tanks (BNC spoken) perfective auxiliary c. that we ain’t gonna relet these (BNC spoken) progressive auxiliary This ain’t could in principle be reanalyzed as the negative, but there is no evidence in British English that ain’t is spreading as a standard negative, for example, used with an inflected, finite verb, as in (31). This sentence probably has a meaning of ‘I didn’t see/haven’t seen.’ (31) I ain’t see any because I were with Jacqueline weren’t I? ‘I didn’t see any because I was with Jacqueline, wasn’t I’ (BNC spoken). Of the 1,270 instances of ain’t followed by a verb in the British National Corpus, no verbs are finite, but of the 4,405 instances in COCA, there are 50 or so where ain’t could be a negative particle preceding the finite verb, namely, those in (32). (32) a. When I came to this class, I ain’t know nothing. ‘When I came to this class, I didn’t know anything’ (COCA spoken). b. Nah, you ai n’t want trying to hit the coach in the face. ‘No, you don’t want to try to hit the coach in his face’ (COCA fiction). c. It ain’t have any beer? ‘It doesn’t have any beer’ (COCA fiction). So English copulas and auxiliaries could participate in a NEC when their inflection is neutralized, as with ain’t. For external reasons, ain’t is stigmatized as nonstandard (OED, s.v. have). What is possible in connection to the NEC is that the verb is renewed at the end of the cycle by a new existential or copula, in something we’ve encountered in the previous chapter as the Copula Cycle, where I take a copula in the broad sense as locative, equational, possessive, or existential. This copula can then again be the source to another NEC.

104

Micro cycles

There is evidence of the NEC in families such as Athabaskan, Pomoan, and Finno-Ugric, to name but a few. The Athabascan family has a negative construction derived from a negative copula/auxiliary. For instance, Kari (1990) suggests that the negative ‘ele’ in Ahtna (33) is related to the verb lae ‘to be,’ and one could argue that the suffix -leh is also related to that verb. Kwadacha (34), Dëne Sųłiné (35), and Tlingit (36) have the same forms but no affix, and in Carrier (37), the negative is a prefix. It is thus likely that the negative marker arose from a negative existential, which was itself lost. (33) ‘ele’ ugheli ghi-leh Ahtna NEG good 3-PF.be.NEG ‘He is not good’ (Kari 1990: 272). (34) Edna ?ədu Mary əʔi̢’h Kwadacha/Ft Ware Sekani Edna NEG Mary 3.see ‘Edna doesn’t see Mary’ (Hargus 2002: 110). (35) nεzú-hílε Chipewyan/Dëne Sųłiné be.good-not ‘It is not good’ (Li 1967: 420). (36) ƛéł wusgîd Tlingit NEG fall.IRR ‘He didn’t fall’ (Krauss 1969: 72). (37) lh-e’-z- us-’al Carrier NEG-OM-NEG-1S-eat ‘I am not eating (an unspecified object)’ (Poser 2009: 26). Leer reconstructs an alternative scenario with a Proto-Athabascan *-he suffix, which is “originally an enclitic” (2000: 102), and a Proto-Atabaskan-Eyak-Tlingit particle *(ʔi)łeʔ ‘it is not’ (Leer 2000: 123). One question is the following. Is *(ʔi)łeʔ originally a third person negative of the verb ‘to be’ that is reanalyzed as a negative particle during Proto-Athabascan-Eyak-Tlingit, or is it still an auxiliary? Rice (1989: 1108, n. 1) suggests that the negative yíle in Slave, for example, in Bearlake (38), “may historically be an auxiliary verb in the perfective aspect.” (38) bebí nedá yíle baby heavy NEG ‘The baby is light’ (Rice 1989: 1101).

Bearlake

Athabascan most likely underwent the NEC, where the copula and the negative merged. Currently, the copula verb is absent, and the negative is used for all kinds of verbs.

Micro cycles

105

Mithun (1998) presents data from Central Pomo (Pomoan), where two negatives (generally) specialize for imperfective or perfective aspect. The čhów negative in (39) marks the perfective, whereas the hín in (40) is used for other aspects. (39) ʔaˑ čá-ˑn-ka-w čhów. Central Pomo (Pomoan) 1.AGT run-IMPF-CAUS-PRF NEG ‘I didn’t drive’ (Mithun 1998: 78). hín =ka. Central Pomo (Pomoan) (40) áwhal yhé- -aʔ work do-MULTIPLE-IMPF.PL NEG=INFERENTIAL ‘They must not be working’ (Mithun 1998: 78). In (39) and (40), the negative depends on the mood or aspect of the clause. This is to be expected if the negative derives from a verb or auxiliary, which are marked for TMA and were merged with the negative. Mithun (2021) argues exactly that: hín is used in more functions and in more languages of the family and represents the older use; čhów is the perfective of a verb with the meaning ‘not exist,’ still present in the language, as (41) shows. (41) Míye: ṭʰé cʰó-w. Central Pomo her mother be.absent-PF ‘Her mother wasn’t there’ (Mithun 2021: 689). The origin of the negative auxiliary in Uralic “may well be related to the verb ‘is’ (i-)” (Simoncsics 1998: 594) and more precisely to a negative copula (Honti 1997: 173). That would mean the NEC occurred in earlier Uralic. We cannot be completely sure about this scenario, but the present-day languages in the family show how the NEC proceeds: the negative auxiliary gradually loses inflection to end up as an uninflected particle. An example of an inflected negative auxiliary in the Uralic family appears in (42a). Other varieties of Saami have reduced inflection, as in (42b) and (42c), with the main verb picking up the tense. (42) a. b. c.

Idtj-im (manne) daejrie-h Southern Saami NEG.PST-1S (1S) know-CONNEG ‘I didn’t know’ (Bergsland 1994: 44). (mon) jiõm poor Saami (I) 1S.NEG eat.PR ‘I don’t eat’ (Miestamo & Koponen 2015: 355–356). jiõm poor-râm Saami 1S.NEG eat-PST ‘I didn’t eat’ (Miestamo & Koponen 2015: 355–356).

106

Micro cycles

In Estonian and nonstandard Finnish, the auxiliary has been reduced to a noninflected particle ei for all negation, as in (43), and ei can be deleted if a negative adverb is present (see Honti 1997: 164). (43) Maia ei laula Estonian Maia NEG sing.CONNEG ‘Maia doesn’t sing’ (Veselinova 2016: 151, data from Miina Norvik). The variety among the languages of the Uralic family shows an auxiliary as source for a negative particle. In conclusion, Negatives Cycles come in different varieties. Some negatives are renewed through a nominal that is originally negative or denotes a small quantity and others through verbs and auxiliaries that are either merged with a negative or through an originally semantically negative verb. As pointed out mostly for the Jespersen Cycle, both grammaticalization and renewal can start this cycle. The NEC starts with grammaticalization and Givón’s Cycle presumably with a pragmatic motivation. Negation is one type of polarity, and I now turn to another polarity cycle, namely, the Interrogative Cycle. 2

The interrogative cycle

Negatives and polar interrogatives share an irrealis quality that revolves around being negative or having a choice between positive and negative. It has often been noted that yes/no interrogatives are derived from negatives, for example, Bencini (2003), van Gelderen (2011), Bailey (2013), and Waltereit (2020). The latter points out that Tesnière (1959: 191) groups negation and interrogation together because they do not commit to the truth. Waltereit coins the ‘Interrogative Cycle’ for this change. In addition to the interrogative marker originating as a negative, there is another source, namely, a wh-element (cf. van Gelderen 2011: Chapter 8). In this section, I first provide examples of the first cycle, from Latin, Quechua, and Southern Min, and then of the second, from the history of English. Negatives reanalyze as yes/no interrogatives, that is, to a polarity position in the left periphery. Negatives are specified for the negative value, and if the negative quality somehow weakens, it can be reanalyzed as a marker whose polarity is not specified. An example of this cycle can be found in Latin. Yes/no questions can be introduced by a clitic -ne added to an emphatic word, as in (44). This -ne is the original negative, so the situation in Latin is one where an original negative is now an interrogative in the left periphery without a negative meaning attached. (44) tu=ne id veritus es Latin you=Q that fear be ‘Did you fear that?’ (Greenough et al. 1931: 205).

Micro cycles

107

Regular negatives in Latin show the effect of the Jespersen Cycle because the original negative marker ne is reinforced to non through contraction with oenum ‘even one,’ as in (45). (45) non


Y/N Question > zero wh-element > Y/N Question > zero Walkden (2022) examines two other sources of question-particles, namely, subordinators and modal verbs. I leave these for future work. In this section, I have looked at two Interrogative Cycles, involving Latin, Quechua, Chinese, Southern Min, and English. Because both negatives and whwords are fronted to a focus position, they can be reanalyzed as polarity markers and can themselves, over time, be renewed, as in Latin, or continue in their original function, as in Southern Min. 3

Complementizer cycles

Complementizers derive from PPs, NPs, wh-elements, adverbs, and determiners. In this section, I focus on renewal by the first of these, PPs (see van Gelderen 2009b for the other sources). Their grammaticalization typically depends on first being fronted in their original function and then losing some of their semantic content to link two clauses. In this section, I provide examples from the history of English of such a reanalysis. The preposition and adverb after in Old English, according to the OED, indicate place, order, or time, as in (67). (67) Fand þa ðær inne æþelinga gedriht swefan æfter symble found then there in noble company sleeping after feast ‘He found therein a company of nobles sleeping after their feast’ (Beowulf 118–119). As far as the syntax of these constructions is concerned, in early texts such as Beowulf, after is mainly used as a preposition in a PP situated inside the VP, as in (67), that is, not yet grammaticalized. Of the 65 instances of after in Beowulf, only two occur inside a fronted PP, as in (68). This is typical for Old English use. (68) Æfter þæm wordum Wedergeata leod efste mid elne after those words Weather.Geats chief hastened with courage ‘After those words the Weather-Geats chief hastened with courage’ (Beowulf 1492–1494).

112

Micro cycles

In later (prose) texts, the PPs are fronted more often, as in (69) to (71), and the object is a demonstrative, as in (72) and (73). This is typical for languages like Old English that use the initial position for pragmatic linking. These clauses are still independent, however, so that the PP has no subordinating function. (69) Her Leo se æþela papa & se halga forþferde, & æfter him Stephanus feng to rice. Here Leo the noble pope and the holy died

(70)

(71)

(72)

(73)

and after him Stephen started to rule

‘In this year, Leo the noble and holy pope died and after him, Stephen started to rule’ (Parker Chronicle, entry for the year 814 [816]). & þær wearþ Heahmund biscep ofslægen, & fela godra monna; & æfter þissum gefeohte cuom micel sumorlida. ‘And there was Bishop H. killed and many good men, and after this fight came many summer troops’ (Parker Chronicle, entry for the year 871). Her forðferde Wulfstan diacon on Cilda mæssedæge 7 æfter þon forðferde Gyric mæsse preost. ‘In this year died Wulfstan deacon . . . and after that died Gyric the masspriest’ (Parker Chronicle, entry for the year 963). Þa æfter þam for se here eall up. Then after that went that army all up ‘Then after that went the army all there’ (Parker Chronicle, entry for the year 918). Æfter þysan com Thomas to Cantwarebyri. after this came Thomas to Canterbury ‘After this, Thomas came to Canterbury’ (Parker Chronicle, entry for the year 1070).

In Table 4.2, figures are given for three stages of Old English, (a) the stage with a PP predominantly inside the VP, represented by Beowulf, (b) the stage with some topicalization, represented by the early version of The Anglo Saxon Chronicle up to 891, and (c) the stage with frequent topicalization where the adverb’s VP-origin is less obvious to the language learner, represented by the same chronicle in entries after 892. Before 892, after is followed by a noun or pronoun and rarely (7.7%) by a demonstrative, and the PP is preposed in 27% of the cases. In the later Chronicle (i.e. after 892), many of the objects of after are demonstratives, as in (108) and (109), namely, 17 out of 22 (= 77%). The use of a demonstrative object indicates that the PP is starting to be seen as an adverb linking one sentence to another. This is confirmed by the frequent topicalization of the PP (12 out of 22 = 54.5%). In a Verb-Second language, such as Old English, this preposing (especially with demonstratives) occurs for reasons of textual cohesion, as argued in Los

Micro cycles

113

TABLE 4.2 Numbers and percentages of demonstrative objects (Dem) with after and

fronting

Topicalization Dem objects

Beowulf 2/65 = 3% 2/65 = 3%

Chron A 892 12/22 = 54.5% 17/22 = 77%

(2009). According to Los (2009: 107–108), the loss of V2 causes a typological change from bounded to unbounded, that is, a particular structuring of the narrative and a change from frequent anchoring by temporal markers to little or no such marking. Topicalization makes it possible for the PP to ultimately be reanalyzed as a complementizer. In (74), the status of the PP is ambiguous. It could be a temporal adverb or a complex conjunction introducing a subordinate clause, as indicated in the translation. The same text has subordinate clauses headed by æfter, a demonstrative, and a complementizer, as in (75). (74) Æfter þæm Iulius for to Rome & bæd after that Julius went to Rome and asked þæt him mon brohte þone triumphan ongean. that him man brought that triumph meet ‘After that/After Julius went to Rome, he asked that that triumph would be brought to meet him’ (c900, Alfred, Orosius, 126.11–12). (75) Æfter þæm þe he hie oferwunnen hæfde, after that C he them conquered had he for on Bretanie þæt iglond. he went to Britain the island ‘After he had conquered them, he went to the island of Britain’ (c900, Alfred, Orosius, 126.3–4). The first instances of reanalysis as part of the left periphery include (75), from around 900, (76) from around 1,000, and (77) and (78), from around 1,200. These instances of after are not complementizers but PPs and indicate time. (76) Witodlice æfter þam þe ic of deaþe arise ic cume to eow on galilee Surely after that C I of death arise I come to you in Galilee ‘Surely after I rise from the dead, I will come to Galilea’ (OED, c1000, West Saxon Gospel Matthew 26.32 Hatton MS). (77) for efter þan þet þe mon bið dead me leið þene licome in þere þruh Because after that C the man is dead men lay the body in the tomb ‘After the man is dead, they put the body in the tomb’ (c1200, Lambeth Homilies 51: 4–5).

114

Micro cycles

(78) Affterr þatt tatt he wass dæd Ne toc 3ho wiþþ nan oþerr after that C he was dead NEG took she with no other ‘after he was dead, she didn’t take another (man)’ (OED, c1175, Ormulum 7667). (74) and (75) occur very close to each other in the manuscript and are clearly variants. The same variation occurs in Middle English. For instance, in addition to (77) from the Lambeth Homilies (West Midlands, early thirteenth century), there are other constructions, such as (79) and (80), which represent earlier stages. (79) Efter þon he him sceawede þe sea of helle and innan þan sea weren . . . ‘After that he showed him the sea of hell and in that sea were’ (c1200, Lambeth Homilies 43.2, Morris edition). (80) Sunnendei fond noe lond efter þet ure drihten hefde þet folc adreint. ‘On Sunday, Noah found land after our lord had drowned the people’ (c1200, Lambeth Homilies 139–141, Morris edition). In (79), the PP is clearly preposed and is not conjoining the sentence to another. The orientation of the clause, as in Diessel (2019), is backwards looking: it is referring to something that has happened. In (80), it is only possible to analyze the PP as a conjunction because the drowning occurs before finding the land. So after loses its backward-looking orientation. The order of the two clauses doesn’t follow the chronological order of events. The same shift from less embedded to more integrated can be seen between the different stages of the Gospel glosses. The (earlier) Lindisfarne gloss renders the relevant part of (76) as (81), without the complementizer. (81) æfter ðon uutedlice ic eft-ariso ic forlioro l iowih in galileam after that surely I again-rise I come_before you in Galilee ‘after that surely I arise again and come before you in Galilee’ (c950, Lindisfarne Gospel, Matthew 26. 32). These Gospel glosses are not based on the Latin original, which lacks the complementizer that the later West-Saxon version puts in. The complementizer-less stage represents an earlier variety. This is confirmed by data in Rissanen (2007: 61; 64), who examines the Old English parts of the Helsinki Corpus and finds an increase in complementizers following the PP. The two clauses in (81) are more independent of each other than those in (76), and the PP could be a regular adverb. So far, the development shows that the PP with after is fronted and that its object is often a demonstrative, not a full noun. The demonstratives are still

Micro cycles

115

inflected for case (sometimes dative, other times accusative), which means the PP is still adverbial. The second stage, we have seen, is for a complementizer to follow the PP. This stage involves a change in the temporal orientation of the clause, from backwards looking to forwards looking. The third stage, which we turn to now, is for the preposition to be reanalyzed as a complementizer. The first use of after as a clear head in the OED is in the Late Middle English (82). This is attributed to Wycliff, and as (82) also shows, after that and after are variants. (82) After þat Crist had ordeynid his apostlis, and sent hem to preche; after he assignid seuenty and two disciplis, and sent hem . . . ‘After Christ had ordained his apostles and sent them to preach; after he appointed 72 disciples and sent them . . . ’ (MED, c1360 Wyclif Apology for Lollard Doctrines V). The late Middle English period is indeed one with after as a frequent complementizer, as (83) and (84) show, from the middle of the fifteenth century, with an occasional that, as in (85) and (86). (83) after I met wyth hym in þe strett and spak wyth hym ‘After I met him in the street and spoke with him . . . ’ (Paston Letters 119, anno 1464). (84) and after he had taried there a while he toke a promise ‘and after he had stayed there a while, he took a promise’ (Paston Letters 112, anno 1480). (85) Also Will . . . told me þat he, after þat ye told hym of þis matier lyke as ye wr[i]te, he comuned with Maister Will Swan, and . . . ‘Also, Will told me that he, after you told him of this matter as you write, he communicated with Master Will Swan, and . . . ’ (Paston Letters 3, anno 1425). (86) Plesse it 3ow to wett þat after þat I harde say þat þe parson of Blowfelde wasse com to town I went to hym to hys jn . . . ‘May it please you to know that, after I heard say that the parson of Blowfield had come to town, I went to him in his inn’ (Paston Letters 119, anno 1464). In conclusion, the reanalysis of after from P to C is triggered by a frequent preposing of the PP, which plays a clause-connecting role. The complete changes are indicated in Table 4.3. The development of after shows that, after an initial period of functioning both as VP adverb and sentence connector, the PP headed by after is reanalyzed as part of the CP and then as a temporal complementizer in the head C.

116

Micro cycles TABLE 4.3 Reanalyses involving after

a. b. c.

PP PP C C

pre 900 900 1360

-

present 1600 present

In the case of after, the prepositional function remains intact, and after is not renewed but splits into the original preposition and the new complementizer. This is not the case with the preposition for, for instance, which originally has a causal meaning in (87) and which is replaced by a new PP, because of, in most varieties of English, as the translation in (87) shows. (87) ouþer for untrumnisse ouþer for lauerdes neode ouþer for haueleste ouþer for hwilces cinnes oþer neod he ne muge þær cumon ‘either because of infirmity or because of his lord’s need or because of lack of means or because of need of any other kind he cannot go there’ (c1100, Peterborough Chronicle, entry for the year 675). This PP goes through the same changes as the one headed by after, for example, preposing, as in (87), and then reanalysis as complementizer in (88). (88) Þa þestre sona þas landes, for æuric man sone ræuede oþer þe mihte then dark(ened) soon the land, because every man soon robbed another that could ‘Then, there soon was tribulation in the land, because everyone that could robbed someone else’ (Peterborough Chronicle, anno 1135, 8). A schema for this change is shown in (89). (89) A Complementizer Cycle may involve: a. PP in VP > PP in left periphery b. new PP (e.g. because of)

>

C = grammaticalization = renewal

In this section, I have shown that an adjunct PP headed by after topicalizes from its VP-internal position to the left periphery. After the increase in fronting, the PP can get reanalyzed as complementizer. In the case of after, this word keeps its original prepositional function as well. This is not so in the case of for, whose original causal function is lost and replaced. Here, grammaticalization precedes renewal. 4

Pragmatic cycles

In this section, I first take a look at Hansen’s definitions of a Pragmatic Cycle and argue that these can be seen as covering (a) the renewal side of what I have

Micro cycles

117

considered the Linguistic Cycle and (b) the grammaticalization of pragmatic markers. I then give some examples of the temporal adverbial reanalyzing as discourse markers and of personal pronouns being pragmatically strengthened. 4.1 A defnition

Hansen (2020: 165) considers “semantic – pragmatic cycles [as] a relatively recently discovered phenomenon” that contrast with the negative and other cycles. In these cycles, according to Hansen (2020: 167), “pragmatics may in fact be the driver, such that the innovative items or constructions are initially specialized for certain types of pragmatically loaded context, and are only gradually extended to all contexts.” An example of such a cycle is given in Hansen and Visconti (2009), namely, the Old French negative renewals pas and mie. These forms were first used for special pragmatic functions, as evidenced by their negating discourse-old information and are therefore considered different from cycles where grammaticalization sets the process in motion. However, as far as I understand von der Gabelentz, this pragmatic start of a cycle is part of the linguistic cycle, namely, the renewal part. As I have argued in previous chapters, that cycle can start either on the grammaticalization side or the renewal side, that is, be pragmatically motivated, and sometimes it is hard to determine which of the two is first. As was just mentioned, Hansen uses two different definitions of (Semantic)Pragmatic Cycles. According to one such definition, Pragmatic Cycles proceed as in (90). (90) Semantic-Pragmatic Cycles take place as follows: A given linguistic expression e that originally has a content-level function may, from a certain point in time, develop one or more context-level functions, and thus become “pragmaticalized.” At some point, a new expression e′ may begin to fulfil the contentlevel function originally associated with e. Depending on whether or not e has lost its source meaning, it may thus find itself in competition with or replaced by e′ at the content level. In turn, e′ may itself become pragmaticalized, developing context-level uses that strongly resemble those of e. Eventually, e may disappear from the language altogether, or at least find itself increasingly confined to specific genres or registers. In some cases, more than two expressions may be involved in a semantic – pragmatic cycle, either within a given language or across a mother language and a daughter language, such that the same cycle is repeated with a new expression e″, and so forth. (Hansen 2020: 168)

118

Micro cycles

In (90), the change from a content item to a pragmatic function is first described. This happened in the case of pas and all the tea in China, discussed in Chapter 1. The lexical/content meaning of these may be replaced by something that itself may start to be used in a pragmatic sense. Because the two uses are competitors, one may disappear, and it is assumed to be the original one. I assume these changes are called semantic-pragmatic because the semantic content is lost and replaced by a pragmatic use. However, if we think of the case of the negative or demonstrative or copula, the content is retained but used in a different way, for example, with more emphasis. The description in (90) seems to be very similar to that in von der Gabelentz, and the cycles discussed so far do fit there. The difference is that it attributes the start of the change to pragmatic needs. Another angle on this cycle, mentioned by Hansen, that I think is slightly different, is given in (91). This is what I will consider as a Pragmatic Cycle. (91) Semantic-Pragmatic Cycles, or cycles of pragmaticalization are “cycles of change affecting linguistic items that express more subjective, context-dependent and non-truth-conditional meanings, such as the speaker’s stance, or how the utterance fits into the surrounding discourse” (Hansen 2020: 168). Such cycles include the development of discourse markers. A discourse marker is a syntactically optional and non-truth-conditional marker serving to mark speaker attitudes, hearer-speaker interactions, and textual organization (e.g. Traugott 1998; Abraham 1991; Schourup 1999; Heine et al. 2021). Modal particles are not discourse markers because they are integrated in the syntax and prosody of the sentence. Examples of new discourse markers appear in Mithun (2008a), where Navajo -go changes from a complementizer to a discourse marker. Navajo (92) shows the use of the clitic – (g)o as a subordinator and (93), which is the beginning of an anecdote, includes the use of -(g)o in independent sentences. (92) áko jineezjéé’=o shį́į́ da’jiłhxaazh. Navajo (Dolly Hermes Soulé) so recline=GO perhaps all.PF.sleep ‘And then apparently [when everybody was in bed], they all went to sleep’ (Mithun 2008a: 71, glosses adapted). (93) a. Hádą́ą́léi’ya, hastiin léí’ nihaanííáya=ǫ́ akwe’é one_day man some us.visit=GO here ‘One day a man came to visit us here.’ b. haashį́į́ t’áao nhik’éí=go . somehow PRT our.relative=GO ‘He’s our relative somehow’ (Mithun 2008a: 71, glosses adapted).

Micro cycles

119

Mithun demonstrates that earlier stages of Navajo and other Athabaskan languages do not use -go in this independent function and that therefore the direction has gone from dependent, as in (92), to insubordinate, as in (93). 4.2 Temporal sources

A complete cycle is provided by Fedriani and Molinelli (2020), who show that tum ‘then’ functions as a temporal adverb in (94) and also as a connective in Early and Classical Latin. (94) Cumae, quam Graeci tum urbem tenebant Classical Latin Cumae REL Greeks then city held.3P ‘Cumae, a city which the Greeks at that time held’ (Titus Livius 4, 44, 13, Fedriani & Molinelli 2020: 187). Tum is subsequently replaced by tunc, which has the same basic meaning although originally used in an emphatic way, as in (95) – note also the extra alliteration. Tunc is a renewal of tum, having had a deictic -ce (derived from ecce ‘see’) reinforce it. (95) ut tibi tunc toto pectore sollicitae Classical Latin To you then whole breast anxious ‘At that time how from your whole breast (did your anxious spirit fail!)’ (Catullus 66, 24, p. 189, Fedriani & Molinelli 2020: 189). Fedriani and Molinelli (2020: 190) show that a classical writer like Catullus still uses tum more frequently and that the shift comes in the post-classical period, after the first century ce, for example, with Ovid. In this period, as shown in (96), tunc loses its “emphatic nuance” and is displaying the “neutral meaning of earlier tum.” (96) memini Safinium: tunc habitabat ad arcum veterem, me puero ‘I remember Safinius: he used to live then by the old arch when I was a boy’ (Petronius 44, 6, Fedriani & Molinelli 2020: 190). In turn, tunc is replaced by illa hora ‘that hour’ for temporal meanings and by dumque ‘and while’ for “a resultative-connective discursive function” (p. 194). These result in allora ‘so’ and dunque ‘so’ in Italian and alors ‘then’ and donc ‘so’ in French. Spanish entonces retains both temporal and discursive meanings of the original. Ghezzi and Molinelli (2020) describe the cycle of illa hora in more detail; it starts with grammaticalization and proceeds to pragmaticalization. As just seen

120

Micro cycles

in the data by Fedriani and Molinelli, Latin tum and tunc are temporal adverbs that survive in Modern Italian as discourse markers in the form of dunque. In Late Latin, the temporal illa hora (‘that hour’) preceded by prepositions such as ad becomes allora, adlora, etc., and takes over some of the functions of tum and tunc. “Progressively the Latin system disappears due to a long process of erosion, and the synthetic forms were substituted by popular analytic innovations. One of these, namely, illā horā (literally, ‘at-that.ABL moment/hour.ABL’), routinised also in written varieties in Late Latin” (Ghezzi & Molinelli 2020: 214–215). In Old Italian, the form has a temporal and connective meaning, and it keeps these in later Italian, in addition to the newer function of discourse marker, as in (97). (97) Gigliola: . . . Ha preso sonno. Iddio sia lodato. ‘Gigliola: He has fallen asleep. God be praised.’ Fortunato: Allora, mammà, noi arriviamo a casa, mangiamo un boccone e torniamo. ‘Fortunato: So, mum, we come home, we have a bite to eat, and we go back’ (De Filippo, Gli esami non finiscono mai, III, 1973, from Ghezzi & Molinelli 2020: 227). These cycles can be schematized, as in (98). (98) Cycles of temporal ‘then’ involve tum > tum+ce > tunc > neutral ‘then’ emphatic ‘then’ neutral ‘then’ ad illa hora > allora neutral ‘then’ pragmatic use

dumque/donc pragmatic use

Scivoletto (2020) studies the evolution of the Sicilian discourse marker mentri, from temporal complementizer to contrastive adversarial ‘though’ to discourse marker. Originating “from Latin dŭm ĭntĕrim (‘while, in the meantime’), the temporal conjunction develops – like its Romance cognates – an adversative function meaning ‘whereas,’ which further evolves from an oppositional to a counter-expectational contrast value meaning ‘though,’ The latter serves as a bridging context for the emergence of discourse-pragmatic uses” (Scivoletto 2020: 236) such as in (99). (99) ora cu P. vi vir-iti menṭṛi? Sicilian now with P. you see-PRS.2PL while ‘By the way, now are you and P. going to meet?’ (Scivoletto 2020: 247). A relatively new development in the study of discourse markers is Heine et al. (2021). They first discuss Degand and Simon-Vandenbergen’s (2011) edited

Micro cycles

121

volume and other work that favors grammaticalization as the correct term for the changes that lead to new discourse markers and argue that the term pragmaticalization is not the right one (293). Heine et al. see grammaticalization as crucial in the development of discourse markers but they add a second mechanism, cooptation, “a cognitive-communicative operation whereby a text segment such as a clause, a phrase, or a word is transferred from the domain or level of sentence grammar and deployed for use on the level of discourse organization” (Heine et al. 2021: 26). They then discuss the development of discourse markers in four languages, English, French, Japanese, and Korean. The case studies for English include after all, anyway, I mean, if you like/ will, instead, like, no doubt, right, so to say/speak, well and what else. I’ll just mention one to illustrate the notion of cooptation. After all creates a link between two discourse segments, creating a concessive relation. Based on Lewis (2000) and Traugott (2018), Heine et al. describe the origin of this marker as a temporal adverbial, as in (100). The cooptation to discourse marker takes place in the seventeenth century where it starts appearing after but or yet, as in (101). (100) And after all they whent to his plasse to dener (Machyn, anno 1559, from Heine et al. 2021: 94). (101) And yet, after all, there has been so little ground for his jealousy of me (Pepys, anno 1663, from Heine et al. 2021: 94). 4.3 Emphatic pronoun cycles

A last case of a Pragmatic Cycle I discuss is emphatic pronoun renewal and person shift. In Section 3 of the next chapter, I will discuss pronouns that renew after having been reanalyzed as agreement markers. There are also pronoun renewals for pragmatic reasons that I consider as micro cycles since they affect individual words, not pronoun systems. They are pragmatically motivated cycles because they are caused by external factors such as “hierarchical social organization and status” (Heine & Song 2011: 687). Heine and Song (2010, 2011), Shibasaki (2014), Bouzouita and Breitbarth (2019) have described such cases of pronoun shift, from third to second and second to first and from plural to singular. Jespersen (1941: 39–40) spoke of them as “obnoxious” polysemy. For instance, Japanese is a language where pronouns are continually renewed because of pragmatic depreciation and semantic bleaching, according to Ishiyama (2019: 61–62). He writes “nouns are the major source of personal pronouns, demonstratives and location expressions give rise to personal pronouns only in limited circumstances.” In Japanese, “speakers have to constantly introduce new forms for politeness purposes” (61). Nouns can be used for third, as well as for first and second person pronoun renewal, and pronouns can shift their person, in what has been called ‘person shift.’ One such case is that of watakushi, whose

122

Micro cycles

meaning is originally ‘private affairs,’ as in (102), and the opposite of ‘public.’ It is then extended to ‘individual/self,’ as in (103), which triggers the first person in (104). (102) Makotoni watakushi o wasurete chuu o sonzuru hito truly watakushi OM forget loyalty OM think person wa kayooni koso aru beki ni TOP like.this EMPH exist MOD but ‘Truly, this is the way the person who disregards personal interests (watakushi) and has high regards for loyalty should be, but . . . ’ (Stage III: 14C, Taiheiki, Book 26, Ishiyama 2019: 22). (103) Iro koso sonjite-soorae domo imada watakushi ni sooroo color EMPH damage-HUMBLE but still watakushi LOC exist ‘Although the color is damaged, it is still by me (watakushi)’ (Stage III: 14C, Taiheiki, Book 18, Ishiyama 2019: 22). (104) Watakushi wa wakai toki ni Harima no kuni Akashi ni arishiga watakushi TOP young time at Harima of country Akashi in live ‘When I (watakushi) was young, I lived in Akashi in the Province of Harima’ (Stage IV: 1686, Koushoku Ichidai Onna, Book 3: Chapter 2, Ishiyama 2019: 23). Ishiyama, quoting Whitman (1999), calls the person shift ‘personalization.’ It is most frequent in languages without verbal agreement and is ‘caused’ by ‘empathetic identification’ (taking the perspective of another, e.g. doctorese we) or extravagant politeness (vuestra merced ‘your grace’ > Spanish usted, Haspelmath 1999: 1056). A plural is more distant and can be seen as increasing or decreasing politeness (Ishiyama 122–123). Another instance of person shift is in German der Herr ‘the gentleman,’ which can be used for second person and Dutch deze dame ‘this woman’ for first person. These are used in pragmatically marked situations (cf. Heine & Song 2011) and have not become regular pronouns, as in Japanese. Concluding, I have discussed a few Pragmatic Cycles, that is, those that take adverbials to create new discourse markers. I have given examples of grammaticalization from Navajo and of discourse marker cycles from Latin, Italian, Sicilian, and English. Finally, I consider the notion of cooptation and pronoun renewal or person shift. 5

Interactions between micro cycles

In this short section, I list some interactions between different micro cycles. Interestingly, many of these interactions involve negation. For instance, copulas and auxiliaries contract with negatives to eventually just express negation.

Micro cycles

123

This means that the Negative Existential Cycle and the Copula Cycle interact because a copula is lost and may be renewed again. Copulas are also lost because they come to be seen as agreement. The Jespersen Cycle is frequently used to renew a negative weakened in the NEC. Negatives and interrogatives have much in common so that the former provide input to the latter cycles. Negatives also interact with the Subject Cycle, as does the Object Cycle, as we’ll see in the next chapter. First, the Negative Existential Cycle (Croft 1991) reanalyzes a negative and a copula to combine as a negative marker and, since a copula is lost, may thus set the Copula Cycle in motion. We have seen negatives and copulas interact in varieties of Arabic, Chinese, and Athabascan. The example from Arabic from the previous chapter is reviewed abstractly in (105). (105) a. demonstrative > copula b. copula + negative > negative c. new copula

}

= Grammaticalization = Renewal

The Jodï-Sáliban family provides evidence how copulas can be recruited to mark subject agreement. In (106), which involves a class I verb, the subject is marked by a prefix that Rosés Labrada (2018) argues to date from Proto-Sáliban times and by a suffix that is more recent. (106) ʧ- ɑ̃dĩt- æ̃ˈkʷ- ɑ̃sæ̃ Piaroa 1S work FUT M 1 ‘I will work’ (Rosés Labrada 2018: 33). The agreement suffix has a cognate in the related language Mako, where it is a copula, as in (107) and (108), and extended to habitual and other verbs, as in (107). (107) maestro-ʦa teacher-1.COP ‘I am a/the teacher’ (Rosés Labrada 2018: 43). (108) ɨtʰɨ-ma ʤulewa hoho-ʦa 1S-topic Yureba person-1.COP ‘I am a Yureba person’ (Rosés Labrada 2018: 44). (109) waiʦuhu- ʦa know- F1.COP ‘I know’ (Rosés Labrada 2018: 45).

Mako Mako Mako

The Jespersen Cycle frequently renews negatives that have weakened in Givón’s Cycle and the Negative Existential Cycle. Examples of the former include

124

Micro cycles

Koorete (23), and for the latter, we find a renewal of the negative auxiliary in Uralic. Thus, in Northern Saami, a negative auxiliary appears but less fully inflected than in sister languages, for example, (110) (Lagercrantz 1929: 195; 203–204). It can be accompanied by a reinforcing element, glossed as ‘never,’ which works together with the negative auxiliary since the end result is a negative meaning. (110) In leat goassege dahkan dan NEG.1S be never do.PART it ‘I have never done that’ (Trosterud p.c.).

Northern Saami

In Finnish, the negative auxiliary has become a negative particle en and is also reinforced with a negative polarity adverb koskaan, as (111) shows. In Finnish varieties, the auxiliary can be deleted if this adverbial is present (see Honti 1997: 164 who quotes Savijärvi 1977). This loss is, of course, expected if one considers a typical Negative Cycle. (111) En ole koskaan maistanut sellaisia leipiä NEG have ever tasted such bread ‘I have never tasted such bread’ (Sollid 2002).

Finnish

Negatives are also often affixed onto tense and aspect markers, and these markers may then renew. Examples of this are negatives and TMA auxiliaries that merge and are renewed in, let’s say, Hindi/Urdu. Finally, negatives provide a source for interrogatives because negatives and polar interrogatives share an irrealis quality that revolves about being negative or having a choice between positive and negative. This happens in Latin, Quechua, Chinese, and Southern Min. 6

Conclusions

This chapter has continued the discussion of micro cycles, starting with the best known of all cycles, namely, the Negative Cycle. In fact, there are several different Negative Cycles, those whose source is adverbial or nominal or minimizing (also known as Jespersen Cycles), and those whose source is verbal (also known as Croft’s and Givón’s Cycles). Negatives in turn can be used as a source for the renewal of yes/no questions. Two other cycles conclude the list of micro cycles discussed in this book, the Complementizer and Pragmatic Cycles. The section on the Complementizer Cycle focuses on adverbial PPs being fronted and reanalyzed and then being renewed, and the Pragmatic Cycle is based on work by Hansen (2020) and focuses on the change from adverbial to pragmatic marker.

Micro cycles

125

The last section examines the interactions between cycles that have been noted in this chapter and the previous one. These often involve the negative. As it grammaticalizes, it may contract with a verb (Jespersen Cycle) or a copula/ auxiliary (NEC), and then one of the two elements may disappear. Looking at the features that are lost and renewed provides a window on the universal set of features that human languages employ. Cycles involve all grammatical categories, D, C, Neg, and TMA, as well as the lexical ones, V, N, P, and adverbs (adjectives also renew, as is mentioned in Chapter 7). Thus, cycles give insight in the typological inventory of categories. Micro cyclical changes are interesting from the point of view of the morpho-syntactic features that are found universally. Nouns need definiteness or case or aspect marked on the verb, and verbs need tense, mood, aspect, voice, and pronouns or agreement (or all of these). Copulas are not generally needed, unless they are used to express a certain tense or aspect, but can easily be lost, and their renewal is frequent. Cycles give us insight in the features that are needed and how they are expressed. Suggestions for further reading On Negative Cycles: van der Auwera (2009), Croft (1991), Hoeksema (2009), Breitbarth (2014), Chatzopoulou (2019), and Veselinova and Hamari (2022). For a slightly different view on Athabascan negation: Lovick (2020). On Interrogative, Complementizer, and Pragmatic Cycles: Waltereit (2020), Eckardt and Walkden (2022), Hansen (2020), respectively.

Review questions and exercises

1. 2.

Describe one of the cycles discussed in this chapter in your own words. The following sentence is from Sanskrit. What do you think may be going on? (112) kim aryamno mahas pathaati kramema dudhyo? what Aryaman.GEN great.GEN road.INS surpass.1P inferior.P.ACC ‘Should we overcome the base people on the path of the great Aryaman?’ (Rigveda I, 105.6cd, Davison 2006, quoted from Etter 1985).

3.

Kamassian, an extinct Northern Samoyedic language in Siberia, had an auxiliary for general negation, as in (113), and one to negate existence and possession, as in (114). Given what you know about Croft’s Cycle, what could the expected next stage be? (113) (man) e-m nere-? Kamassian I NEG-1S fear-CONNEG ‘I will not be frightened’ (Künnap 1999: 25). (114) bilä kuza man naFa-m Kamassian bad man I NEG-1S ‘I am not a bad man’ (Simoncsics 1998: 594).

126

4.

Micro cycles

Given what you know from Jespersen’s Cycle (see also Chapter 1), which of the following varieties of Italian is the earlier and which the later? The data are from Parry (2013: 78–79). (115) a n dorum 1S NEG sleep.FUT.1S ‘I will not sleep.’ (116) non dormirò NEG sleep.FUT.1S ‘I will not sleep.’

5.

briza NEG

Emilian Italian Standard Italian

In Section 2 of this chapter, I mention (117). Look at the first 50 instances of sentence-initial how in COCA (or any other corpus) to see if this is frequent or occurs at all. (117) How would you like your eggs cooked? Yes, I would. (cartoon)

6.

7.

In light of our discussion on pronoun shift, comment on the second person pronouns in Table 4.4, taken from Bouzouita and Breitbarth (2019: 1). The T stands for the familiar form and V for the polite (these abbreviations are based on the French tu and vous). Please comment on two recent examples of change in negative expressions. (118) The world ain’t never been ready and ain’t never gonna be ready, but here we are! (social media 2022) (119) I’ve never been there for a long time. ‘I haven’t been there for a long time’ (reported by John Powell 2022).

TABLE 4.4 Changes in the history of Dutch

S

1150–1500 1500–1700 1700–1800 Present-day

P NOM ACC/DAT NOM du di ghi du, ghi di, u ghi gij, je/jij u, je/jou gij je/jij (familiar) je/jou (familiar) jullie (familiar) u (formal) u (formal) < gij lieden u (formal)

(adapted from van Leuvensteijn 2002: 289 and Vermaas 2005)

ACC/DAT u u u jullie (familiar) u (formal)

Micro cycles

127

Advanced exercise

In the language you have selected in Chapter 1: What does sentential negation look like: affix or independent word(s)? Is there evidence of a cycle? Are there special words for polar questions, and if so, are they homophonous with anything? What do the complementizers look like, and what are they homophonous with? Are there discourse markers that originate in temporal adverbs?

5 MACRO CYCLES

In this chapter, I turn to macro cycles. Unlike the cycles described in the previous chapters, these cycles shift the typology of a language. For this kind of change, the focus has traditionally been on the change from analytic to synthetic and then to analytic again. This typology depends on how many morphemes a word has and is usually seen as a continuum on a scale from most synthetic to most analytic. There are other typologies. Languages can also be seen as head marking, for example, when the (synthetic) marking shows agreement on the verb, or dependent marking, for example, when case is marked on the noun. The cycles that involve head and dependent marking involve agreement and case, respectively. Some of the data on these two cycles has also appeared in van Gelderen (2011). Languages have been characterized in terms of word order (Greenberg 1966), with a set of related characteristics. Languages frequently change word order, for example, from SOV to SVO, and although this has been described as a cycle in Vennemann (1973), I do not consider them (macro) cycles in this book because the trigger is not grammaticalization, and the change is not unidirectional. The outline of this chapter is as follows. In Section 1, I elaborate on the definition of macro cycles from Chapter 1 and expand on the discussion there. In Section 2, I discuss the terms analytic and synthetic and provide some examples of a cycle involving these stages. Because the terms analytic and synthetic are not well understood, I suggest using head and dependent marking. In Section 3, I therefore examine a cycle of head marking from pronoun to agreement marking on the verb to zero with subsequent renewal of the pronoun. In Section 4, I discuss Case Cycles, which lose and create dependent marking morphology. Section 5 looks at cyclic interactions between macro cycles, and Section 6 is a conclusion. DOI: 10.4324/9781003272564-5

Macro cycles

1

129

Defnition and controversies

As defined in Chapter 1, repeated as (1), macro cycles involve changes such as those described for Egyptian in Hodge (1970), namely, as changing the typological character of a language, for example, from analytic to synthetic. (1)

Macro cycles involve grammaticalization and renewal in ways that affect entire languages and which change language typologies.

One major typological characterization involves the difference between analytic and synthetic languages, granting that it is hard to use this typology, as I show in Section 2. A change from synthetic to analytic in the history of English is given from Old English (2) to Modern English (3). In (2), we see that almost all nouns and verbs have endings (in bold) that have been replaced in (3) by a stricter word order and the use of articles, pronouns, and auxiliaries (again in bold). (2)

(3)

hwæt we garden-a in geardag-um þeodcyning-a þrym gefrun-on indeed we spear.danes-GEN in yore.days-DAT kings-GEN glory hear-PST

hu ða æþeling-as ellen freme-don how those nobles-NOM courage do-PST ‘Indeed, we have heard of the courageous deeds of the Danes (and) their kings in earlier times, how the noble ones accomplished courageous deeds’ (c1000, Beowulf 1–3). ‘So. The Spear-Danes in years gone by and the kings who ruled them had courage and greatness. We have heard of those princes’ heroic campaigns’ (Seamus Heaney’s 1999 translation).

Note that contentful prepositions, such as in, occur in both synthetic and analytic languages. Heaney’s translation in (3) strives to keep the synthetic character of Old English, for example, by using the synthetic those princes’ heroic campaigns instead of the analytic the campaigns of those princes. The latter would be a more standard translation, with a preposition that replaces the genitive case. Modern English cannot, however, be characterized as a completely analytic language since modals and negatives are merging into single words, as in (4a). It is even possible to see the first person I as a bound element (clitic or affix), typically occurring before a verb and renewed by an oblique me, as in (4b). (4)

a. I couldna done that ‘I could not have done that.’ b. Me, Icouldna done that

Colloquial English

130

Macro cycles

Based on the change between (2) and (3)/)4), the Synthetic to Analytic to Synthetic Cycle can be given as in (5). (5)

Synthetic > case endings, e.g. – a [GEN], -um [DAT] past, e.g. – on

analytic > of, to have

synthetic – couldna

Not addressed very often is the question of derivational morphology. Especially since the 1500s, English has acquired many morphemes from loans (trans-, anti-, micro-, ex-, -ion, -able, -ian, -ity). Should derivational morphology be counted towards synthetic as well? For the purposes of this chapter, my answer will be negative. Grammaticalized derivational affixes, such as -ly, are counted as synthetic. Another typology, articulated by Nichols (1986), distinguishes between synthetically marked verbs, that is, head marking languages, and synthetically marked nouns, that is, dependent marking languages. An example of a head marking language is given in (6) and of a dependent marking language in (7). In (6), both the subject and object are marked on the verb, whereas in (7), the case is marked on the nouns (with only tense marked on the verb). (6)

(7)

jar aak’aalaa7 x-Ø-kee -k’aq aab’aj Tzutujil the boys comp-3S-3P-throw rock pa rwi7 ja jaay on top.of the house ‘The boys threw rock(s) on top of the house’ (Mayan; Dayley 1985: 282, 75, from Bickel & Nichols 2013). wutpu-nku uma-Ø ute-n Uradhi old.man-ERG firewood-ABS pick.up-PST ‘The old man picked up some firewood’ (Paman; Crowley 1983: 339, from Bickel & Nichols 2013).

Languages can also mark both nouns and verbs, as in (2), or neither, as in completely analytic languages. Using head and dependent marking provides a more precise typology than just synthetic or analytic languages, and I will therefore regard changes in head or dependent marking as macro cycles. (Earlier, in van Gelderen 2013, I regarded Pronoun to Agreement Cycles as micro cycles, but I think that was incorrect because these cycles change the typology of a language.) An example of a change in head or dependent marking is taking place in Romance where, for instance, in French, head marking is increasing on the one hand, as I show in Section 3. However, on the other hand, dependent marking

Macro cycles

131

is being lost between older varieties of Romance, as in Latin (8) and Modern French (9), as argued by Ledgeway (2012). This change can be seen by the heavy use of case endings in (8) and the use of prepositions in (9). (8) (9)

pro uita hominis nisi hominis uita reddatur Latin for life.ABL man.GEN unless man.GEN life.NOM is.returned (Caes. B.G. 6.16) à moins que la vie d’ un homme ne soit nécessaire pour racheter French at less that the life of a man not be necessary for buy.back.INF la vie d’ un homme the life of a man ‘unless for the life of a man a man’s life be paid’ (from Ledgeway 2012: 423).

This section has defined the macro cycle as a cycle that affects the typology of an entire language, as in the change from analytic to synthetic and back to analytic. Examples of languages that are synthetic or analytic have also been given, as well as languages that change from one stage to another. These cyclical changes will be developed more in the next section. Since head and dependent marking signal synthetic languages, they have been introduced here and will be elaborated on in Sections 3 and 4, respectively. 2

Analytic and synthetic cycles

In this section, I provide a little more background to the notions of analytic and synthetic, the history of the terms, and how these words have been used. I finish with some instances of full cycles. Analytic languages have words with few morphemes, with the most analytic showing a one-to-one relationship between word and morpheme. Chinese, as in (10), is often cited as a good example of an analytic language. Words in synthetic languages contain more than one morpheme, and Latin (8) is often cited as a good example. (10) wo he le cha 1S drink PF tea ‘I drank tea.’

Chinese

According to Schwegler (1990: 10), it is Du Ponceau (1819) who proposes a third type of language, namely, a polysynthetic kind, although von Humboldt (1836) may be more famous for using the term. As Sapir (1921: 128) puts it, polysynthetic languages are “more than ordinarily synthetic.” An example of a polysynthetic language is given in (11), where the entire sentence consists of a

132

Macro cycles Navajo

Latin

Old English

French

Chinese

Synthetic …………..…………………………………………………………………………………………………………… analytic

FIGURE 5.1

Synthetic analytic continuum

verb tsą́ ‘see.PST’ with three prefixes, expressing the subject and object of the sentence and a classifier specific to the verb. (11) yiyiiłtsą́ Navajo yi-yii-ł-tsą 3–3-CL-saw ‘S/he saw him/her.’ Figure 5.1 shows the continuum that can be formulated for a few select languages. von Schlegel (1818) seems to be the first to use the terms analytic and synthetic where languages are concerned. As Schwegler (1990: 5) points out, from the beginning, the terms were not used in precise ways since they include gradations. Thus, the labels don’t fit the Germanic languages very well in that, according to von Schlegel, these languages “penchent fortement vers les forms analytiques” [“lean strongly towards analytic forms”], and at the same time, they have “une certaine puissance de synthèse” [“a certain power of synthesis”]. von Schlegel’s reasons for postulating the terms may have been to distinguish the more ‘perfect’ synthetic languages from the less perfect ones. He sees the reason for change towards an analytic language to be “les conquérans barbares” [“the barbarian conquerors”] (1818: 24) who acquired Latin imperfectly. McWhorter (e.g. 2007) engages this question of change towards analyticity in languages that are spoken by a majority of non-native speakers, and I will come back to this issue of external influence in the next chapter. In Figure 5.1, an approximation of the degree of analyticity or syntheticity is given, but how do we make that a more concrete calculation? A number of methods have been proposed. Here, I mention four different types of answers, namely, by Greenberg (1960), Easterday et al. (2021), Nichols (1992), and Bickel and Nichols (2013), and Szmrecsanyi (2012, 2016). I find the latter’s method the most helpful in that it calculates both syntheticity and analyticity. Greenberg (1960) provides a system where words are assigned values depending on their complexity. A completely analytic language (one word, one morpheme) would have the value 1.00, a mildly synthetic language would be 2.00 (two morphemes in a word), and a polysynthetic language would average above 3.00. This is shown for a few languages in Table 5.1.

Macro cycles

133

TABLE 5.1 Greenberg’s index of syntheticity

Synthesis

Sanskrit OE Persian English Yakut Swahili Vietnamese 2.59 2.12 1.52 1.68 2.17 2.55 1.06

Inuit 3.72

(adapted from Schwegler 1990: 21) TABLE 5.2 Locus of marking

Head marking Dependent marking Double marking No marking Other Total

71 63 58 42 2 236

(from Nichols & Bickel 2013a) TABLE 5.3 Whole-language typology

Consistently head marking Consistently dependent marking Consistently double marking Consistently zero marking Inconsistent or other Total

47 46 16 6 121 236

(from Nichols & Bickel 2013b)

There are some drawbacks to this system, for example, the ubiquitous problem of what constitutes a word, as Schwegler (1990: 22) points out, and that may be the reason Greenberg stops pursuing it. This method also doesn’t show where the syntheticity is in the system, on the noun or on the verb. Easterday et al. (2021), in examining the relationship between complex syllable structure and degree of synthesis, employ the same method as Greenberg in dividing the number of morphemes by the number of words in texts that show morpheme segmentation. The length of their texts is longer than Greenberg’s samples, but the results are not widely different. For instance, Kalaallisut, a polysynthetic Eskimo-Aleut language, shows an index of 3.02 and Koho, a very analytic Austronesian language, scores 1.01. Nichols (1992) is interested in where the synthesis occurs, on the head or on the dependent, and formulates a point system to determine how head marking or dependent marking a language is. She is less concerned how extreme a language is in its marking but more whether it is head marking or dependent marking consistently. Nichols and Bickel (2013a) provide the breakdown for where the marking occurs in the clause and nominal phrase in 236 languages, which is shown in Tables 5.2 and 5.3.

134

Macro cycles

Both tables illustrate that head and dependent marking languages are relatively evenly divided; not marking heads or dependent occurs but is rarer. Examples of the four consistent types, adapted from Nichols and Bickel (2013b), are Tzutujil, as in (5), Chechen (12), Jivaro (13), and Pirahã (14), respectively. (12) oox-a cha-ca txou-Ø qeelira Chechen 1P-ERG straw-INSTR roof-ABS cover.pst ‘We thatched the roof/we covered the roof with straw.’ (13) tarách-ru-mi-n nawánt-an su-Ø-sá-ja-I Jivaro garment-poss-2sg-obj son-obj give-3-asp-1sg-decl ‘I gave your garment to (my) son.’ (14) ti gí kapiigaxiítoii hoa-í Pirahã 1 2 pencil give-prox ‘I give the pencil to you’ (all examples from Bickel & Nichols). Bickel and Nichols (2013) use the method outlined in (15) for calculating the syntheticity of verbs. (15) Within the same language, verbs can be used with more or less synthesis: the English past, for example, is more synthetic than the future. For surveying purposes, we looked for the maximally inflected verb form, i.e. the one form that is most synthetic, and determined its category-per-word value (“cpw value”). In English, the maximally inflected verb form expresses two categories: agreement (in the present: -s) and tense (past: -ed). The English verb has therefore a synthesis degree of 2 cpw (=categories per word). At one extreme in the sample is Vietnamese, where we found no evidence for any synthetic inflectional category in the verb. The maximally inflected verb in Vietnamese therefore has 0 cpw. At the other extreme are languages like Koasati, whose inflected verb forms can include up to 13 cpw. Easterday et al. (2021) also differentiate between verb and noun synthesis. In one set of languages, it finds verb synthesis varies between 1.52 and 4.14 but noun synthesis only between 1.24 and 2.57. This shows that verbs are more synthetic than nouns and that the complexity of verbs varies more than that of nouns. Szmrecsanyi’s (2012, 2016) approach is novel in that it measures both syntheticity and analyticity. It defines analyticity as using “coding strategies that convey grammatical information via free grammatical markers” and syntheticity as “those coding strategies where grammatical information is signaled by bound grammatical markers.” It also proposes a measure in terms of number of free or bound morphemes per 1,000 words. Analytic words include conjunctions, determiners, auxiliaries, negators, expletives, and comparatives; synthetic markers

Macro cycles

135

consist of clitics, verbal agreement and tense, -er/-est, and plurals. Words can also be simultaneously analytic and synthetic, as in an inflected auxiliary and a pronoun. In the Appendix to this chapter, I show how to use this system. As is obvious from this brief review of the four approaches, it is relatively easy to decide on a purely analytic language but hard to decide on what counts as a synthetic language: is it having words that contain three morphemes or words with five morphemes? As discussed in Chapter 1, it is notoriously hard to decide what counts as a word. As just mentioned, Szmrecsanyi includes clitics in synthesis, but Bickel and Nichols (2013) exclude clitics from synthesis: “words can phonologically cliticize to other words, but this is not synthesis.” In many cases, the distinction is subtle. Another challenge is that register plays a role in determining the analyticity of languages such as English. If, for instance, verbs are more inflected than nouns are, and if verbs are more frequently used in a certain register (Biber & Conrad 2019), that would skew the results. Szmrecsanyi (2016) shows that this is in fact the case, making claims about the Analytic-Synthetic Cycle very hard to evaluate. Apart from morphemes per word or categories per word (cpw), a second distinction is made as to whether the morphemes in the synthetic languages are agglutinative, with clearly distinguished morphemes, as in Jivaro (13), Inuktitut, and Korean, or (in)flectional/fusional, with fused categories, as in English and Latin (8). From a diachronic perspective, there is a cyclical relationship between these stages, for instance, as formulated by Crowley (1992: 170) and reproduced as Figure 5.2. Thus, in the change from isolating to agglutinative, separate words are reanalyzed morphologically as a part of another word but with their own grammatical features connected to the morpheme. In the change from agglutinative to inflectional/fusional, the features of the two morphemes may be combined (in portmanteau morphemes) or may no longer be distinguishable from the root at all. Both changes increase the degree of syntheticity of a particular language. The question also arises as to how to count the number of morphemes in that case.

Isolating

Inflectional/ Fusional FIGURE 5.2

Attachment type

Agglutinative

136

Macro cycles

The development of portmanteau morphemes (also known as cumulative exponence) in French involves a change from agglutinative to fusional and also an increase in degree of syntheticity. Modern French therefore presents an interesting case for an analytic to synthetic change, as Ashby (1974, 1977) has shown in detail. Tesnière (1932) points out that standard French is a synthetic language with an analytic orthography. With this, he means that the pronouns are written separately from the verb but are not in fact independent from these verbs. Bahtchevanova and van Gelderen (2016) show that the Subject Cycle has turned the pronouns into verbal agreement but that this has an effect on the preverbal object pronouns, which can be reanalyzed as agreement as well, turning the language into a really synthetic language. These developments show a macro cycle in progress. Analytic and synthetic stages can co-occur in a language; languages can be in one stage for agreement and in another for TMA and negation. What conspires to make them super-synthetic or super-analytic is an open question. Chinese is analytic in that mood, negation, and aspect are expressed as separate words but might be becoming more synthetic because, for instance, the perfective marker -le in (16) cannot be on its own and has grammaticalized from the verb liao meaning ‘to complete,’ among other meanings (Sun 1996: 85; 178). (16) ta ba wenjian-jia qingqingde fang zai le zhuo shang Mandarin she BA document folder gently put on PF table up ‘She put the documents gently on the table’ (Hui-Ling Yang pc, from yahoo.com.tw). There are many other such words that can no longer be independent in presentday Chinese, for example, the question marker ma and the object marker ba, also shown in (16). LaBarge (2016) explores a change towards the synthetic as the Chinese word yao changes from a full verb meaning ‘desire’ to a future marker and is increasingly limited to occur before another verb. So if we look carefully, highly analytic languages are grammaticalizing as well (see also Post 2007). Haspelmath (2018) writes that it is “quite unclear” if complete Analytic to Synthetic to Analytic Cycles “are all that common,” but he describes the one in Egyptian referring to these as anasynthesic spirals. Change “can in some sense be seen as cyclic alternations between synthetic and analytic patterns” (2018: 1). He also doubts that the changes in attachment need to include fusional in the change from isolating to agglutinative to fusional, as in Figure 5.2. Thus, it is very hard to give a precise definition of synthetic and analytic. A language with one (grammatical) morpheme per word is analytic, and one with more than one per word is synthetic. A language where most arguments are marked on the verb and where nominals are optional is polysynthetic. As already

Macro cycles

137

mentioned, it is an open question as to what factors contribute to a language developing analyticity or syntheticity. I’ll now discuss cycles involving changes in analytic and synthetic typology. First, I outline the full cycle that Hodge provides for Egyptian that was mentioned briefly in Chapter 1. Then, I turn to Chinese, where it’s been argued that earlier stages are synthetic and are reanalyzed as analytic, and currently there are some signs towards syntheticity again. Finally, I describe a change in Indo-European that goes from synthetic to analytic but then doesn’t show signs of further change. Hintze (1947: 96) sees the use of auxiliaries and articles in later Egyptian as an analytic tendency and notes “daß dieser analytischen Tendenz sofort wieder die entgegengesetzte Tendenz zur Univerbierung entgegensteht” (98) [“that this analytic tendency is immediately countered by the opposite tendency towards univerbation”]. Based on the latter’s work and on Polotsky (1960), Korostovcev (1963), Hodge (1970: 3), and Haspelmath (2018), the following changes from Old Egyptian to Middle Egyptian to Coptic can be observed. Old Egyptian is synthetic, Middle Egyptian analytic, and Coptic is synthetic again, both on the verb, as in (17), and in the nominal system, as in (18). The time period between these stages is over 4,000 years. (17) Old Egyptian Middle Egyptian Coptic sḏm-n-f > jr-f sdm > a-f-sôtm hear-PF-3SM do-3SM hear PRET-3SM-hear ‘He heard’ (Loprieno 1995: 220; Haspelmath 2018: 13). (18) rn-k > pɜj-k rn > p-ek-ran name-2SM this-2SM name DEF-2SM-name ‘your name’ (Hagège 1993: 160–162; Haspelmath 2018: 13). In (17), the perfect first appears on the lexical verb and then develops to a stage where an auxiliary is used which in turn develops into a marker on the lexical verb. In (18), the noun is inflected for possessive, and no determiner is used, after which a determiner appears and subsequently disappears. As for Chinese, Karlgren (1920) shows that the pronominal forms are marked for case, as in Old Chinese (19), and that they consolidate as one form for each throughout the history of Chinese. (19) Old Chinese pronouns NOM GEN ACC 1S wú wú wŏ 2S rŭ rŭ ěr 3S yī qí zhī (Karlgren 1920: 207; quoted in Hagège 1993: 159)

138

Macro cycles

Conrady (1915: 262) and Karlgren (1920) also see remnants of synthetic morphology in Old Chinese disyllabic words, initial tone change, and reduplications. According to Karlgren (1920: 204), the initial consonant modifications are remnants of old prefixes serving to form derivational morphology, in particular causative and denominative. Handel (2012) similarly argues that the voicing alternation from Middle Chinese, as shown in (20), is the reflection of a morphological word formation in Old Chinese. (20) intransitive (voiceless) cháng ‘to be long’ shé ‘to bend’ -

causative (voiced) zhāng ‘to stretch’ zhé ‘to break’ (Handel 2012: 61–62)

The prefix is either voiced (*N-) in which case the intransitives are more basic and the transitive derived or voiceless (*s-) in which case the intransitive is derived. Evaluating the work of a number of scholars (e.g. Mei 1989; Sagart 1999), Handel concludes that the voiced prefix makes the most sense as being basic. Without mentioning Karlgren but based on Mei (1989, 2012), Huang (2014: 27) argues that the absence of light verbs in Old Chinese (21), for denominal and causative constructions, shows the presence of a zero prefix and hence the synthetic nature of the language. (21) zhu mu piao you yi mu jian Xin ji, fanXin Old Chinese various women wash, have one woman see Xin hungry, rice Xin ‘Various women were washing clothes. One woman saw Xin being hungry, so she riced Xin [= fed him with rice]’ (Shiji, Huang 2014: 28). Based on (19) and (20), one could argue that Old Chinese has some case marking on pronouns and some derivational morphology on verbs. The question is if that is enough evidence to call it synthetic, as Huang does (“significant synthesis” 2014: 3). Currently, as I have mentioned in relation to (16), there is some evidence that the TMA system may be becoming more synthetic; Huang’s evaluation is “mild synthesis” (2014: 3). If that is true, there is a full Synthetic to Analytic to Synthetic Cycle. I personally think it is debatable how significant the evidence in Chinese is. The development from Sanskrit to Hindi/Urdu shows a shift from synthetic to analytic. In Sanskrit (22a) and (23a), both the noun and verb are marked, for case and agreement with the subject respectively. In Urdu/Hindi (22b) and (23b), case and agreement are marked through prepositions, pronouns, and auxiliaries. (22) a. nagar-at vana-m city-ABL forest-ACC

gaccha-ti goes-3S

Sanskrit

Macro cycles

b. wo shehr see jungel ko jaata he 3S city from forest to go PR.3SM ‘He goes from the city to the forest.’ (23) a.

Ayodhya-yam vasa-ti Ayodhya-LOC lives-3S b. wo Ayodhya mee rehta 3S Ayodhya in live ‘He lives in Ayodhya.’

139

Urdu/Hindi

Sanskrit he PR.3SM

Urdu/Hindi

Is there evidence in present-day Hindi/Urdu of cyclical change towards a more synthetic stage? As in Chinese, there is a subpart of the grammar that is synthetic in the case of Hindi/Urdu that is the future tense, which has gone through a Future Cycle. The future is formed synthetically from a conditional stem of the verb. Thus, ker-ũ-g-i [do-1S-FUT-FS] ‘I will do (it)’ consists of the stem ker ‘doINF,’ the first person ending ũ, a future marker -g- derived from ga ‘go,’ and an inflection for singular number and feminine gender. Lahiri (2000) and Butt and Lahiri (2002), based on Kellog (1893), suggest the development in (24), which I have slightly adapted. (24) a. [mar + ũ] [ga + tah] hit 1S go MS ‘I go to hit.’ b. [mar + ps/#] + [ga + gender/#] c. [mar + ps + fut + gender/#] Sanskrit derives from the Indo Iranian branch of Indo European, as does Old Iranian, and the two are quite close in their synthetic grammar: “Old Iranian preserved much of the rich (though irregular) inflectional morphology of Proto-Indo-European,” according to Haig (2018b: 58). Similar to the changes in the history of later Sanskrit, this synthetic system is lost in the transition to “Middle Iranian [which] witnessed a collapse and levelling of much of the inherited morphology” (Haig 2018b: 58). However, since the great levelling of mophology some two thousand years ago, Iranian languages have been gradually reacquiring morphosyntactic complexity through, for example, univerbation of erstwhile copulas with lexical verbs, the grammaticalization of lexical verbs into modal, aspectual, and voice auxiliaries, the grammaticalization of adpositions to phrasal affixes with case functions, but also in the restructuring and recreation of person agreement systems on the verbs. (58)

140

Macro cycles

There is, for instance, currently some evidence of increased syntheticity in a change from pronoun to clitic to agreement marker from Old to Middle Iranian to Contemporary Central Kurdish (Haig 2018b: 65). So both sides of the Indo Iranian family have undergone a cycle from synthetic to analytic to synthetic. In this section, I have sketched how to calculate if a language is analytic or synthetic, pointed out some problems, and given some examples of these macro cycles. Haspelmath (2018) and Seržant (2021a) point to the paucity of evidence for full Analytic to Synthetic to Analytic Cycles, and I agree. The case of Egyptian is a solid one, as are the changes occurring in Indo Iranian, but the case of Chinese is perhaps less so. Because languages can lose synthetic marking in one area and gain it in another, a better way to think about the Analytic to Synthetic to Analytic Cycles may be to consider them as either involving the verb through agreement/head marking or the noun through case/dependent marking, as in Nichols (1986) and Bickel and Nichols (2013), and I will turn to that now. 3

Agreement cycles or head marking cycles

In this section, I first provide a formulation of the Subject Agreement Cycle and a brief overview of the typological distribution. I then discuss some terminology followed by examples of Subject Agreement Cycles in French, Turkish, Navajo, Agul, and Tabasaran. The second part discusses Object Agreement Cycles in French and Austronesian. Object Cycles are slower and, according to Haig (2018a, b), typically get halted halfway into the cycle. 3.1 Subject agreement cycles

Givón, arguing that agreement markers arise from pronouns, writes “agreement and pronominalization . . . are fundamentally one and the same phenomenon” (Givón 1978: 151). Givón (e.g. 1971, 1976, 1978, 1984) has been the first to emphasize this insight as a cycle, as in (25), prompting Seržant (2021a) to call it Givón’s Cycle. (25) The Subject Agreement Cycle a. demonstrative > third person pronoun > clitic > agreement > zero b. noun/oblique/emphatic > first/second person pronoun > clitic > agreement > zero As is clear from (25), elements other than personal pronouns are a source for agreement; the reason is that any word with semantic or grammatical person, number, and gender features can start to mark the subject on the verb. Siewierska (2004: 248) mentions there are also cases of nouns turning into 3rd person

Macro cycles

141

pronouns, for example, in Zande and Mixtec. Pronoun renewal occurs when a pronoun is reanalyzed as agreement but also for pragmatic reasons, as was discussed in Chapter 4. The typical stages of the Subject Cycle are given in a hypothetical example in (26), where English words have been used for convenience. (26) a. b. c. d.

They (often) eat lasagna. They’eat lasagna. Them th’eat lasagna. Them (often) eat lasagna.

pronoun clitic agreement pronoun

As discussed in Chapter 1, it is hard to distinguish between these stages, on occasion. Pronouns and clitics are arguments and cannot be doubled by another argument; agreement is obligatory and can be doubled, as (26c) shows. Dryer (2013b) provides the distribution of pronominal subject as opposed to subject agreement languages that is shown in Table 5.4. As is clear from those numbers, agreement is much more prevalent (437/711 = 61%) than obligatory, and optional subjects are (242/711 = 34%). From Siewierska (2013b), as shown in Table 5.4, we also learn that agreement with both subjects and objects is frequent (296/378 = 78%), as opposed to zero-marking (82/378 = 22%). TABLE 5.4 Distribution of pronominal subjects

Obligatory pronouns in subject position Subject affixes on verb Subject clitics on variable host Subject pronouns in different position Optional pronouns in subject position Mixed Total:

82 437 32 67 61 32 711

(Dryer 2013b)

TABLE 5.5 Frequency of verbal agreement

No person marking of any argument Person marking of only the A argument Person marking of only the P argument Person marking of the A or P argument Person marking of both the A and P arguments Total: (Siewierska 2013b)

82 73 24 6 193 378

142

Macro cycles

Putting these percentages in the scale of (25) arrives at (27), where Dryer’s “mixed” category has been left out and the optional pronouns have been considered as zero. (27) pronoun and clitic 181 (27%) not compiled

> agreement > zero 437 (64%) 61 (9%) 296 (78%) 82 (22%)

Dryer Siewierska

What (27) shows is that the agreement stage is the most frequent and that the beginning and end stages of (25) are less so. However, as mentioned, the terms pronoun and agreement have been difficult to differentiate on occasion. Most linguists agree that pronouns are arguments of a predicate and that agreement marks an argument on that predicate but is not itself the argument. Clitics are usually seen as arguments but without the phonological independence typical of pronouns. Since the work of Jelinek (1984), it has been acknowledged that agreement markers also function as arguments; this position is also known as the pronominal argument hypothesis. All of this prompts Haspelmath (2013) to give up the terms pronoun and agreement in favor of ‘indexing.’ In what follows, I continue to use agreement and pronouns because they are useful in understanding cycles. Seržant (2021a) argues that the Pronoun Cycle is prevalent in Northwestern Europe, Eastern Southeast Asia, Oceania, Mid Africa, and Northwestern South America, but that 93% of languages are relatively stable in their subject marking. Turkic, Athabaskan, and Lezgic are not mentioned among the unstable ones, but there is evidence of grammaticalization and renewal in these languages, as I show at the end of this section. In what follows, I first look at the Subject Cycle, mainly in French and Turkish, and towards the end of the section at the Object Cycle. French has gone through a complete Subject Agreement Cycle. The Modern French data are well-known from Ashby (1974), Lambrecht (1981), and Zribi-Hertz (1994): there is a strong form of the pronoun (moi, toi, lui, elle for the singulars) and a weak form (je, tu, il, elle for the singulars), the latter of which has been argued to be verbal agreement. I’ll start this discussion with Old French. The subject pronoun in Old French is a fully phrasal pronoun since it can be separated from the verb, as in (28) and (29), and be coordinated, as in (30). (28) Il nen i ad ne veie ne senter Old French He NEG there have.3S.PR NEG road nor path ‘He doesn’t have a road or path there’ (Chanson de Roland, 2399, LRC Old French Online).

Macro cycles

(29) Que jo ne fui a l’estur That I NEG be.1S.PST at the.battle ‘That I wasn’t there at the beginning of the battle’ (Chanson de Roland, 2413, LRC Old French Online)

143

cumencer! beginning

(30) Quant je et mi chevalier venimes hors de l’ost When I and my knights come.1P.PST from the.east ‘When my knights and I came from the east’ (Histoire de St Lois, XLVI, Google Books). However, another emphatic pronoun may appear, as in (31). (31) Renars respond: ‘Jou, je n’=irai’ Old French Renard answers 1S 1S NEG=go.1S.FUT ‘Renard answers “Me, I will not go”’ (Coronnement Renart, A. Foulet (ed.) 1929: 598, from Roberts 1993: 112). Old French, which need not have a subject (i.e. is pro-drop), has nominative and oblique case forms for first and second person singular and nominative, accusative, and dative forms for the third person (also differentiations for gender which are not shown). Table 5.6 provides the data. Early on, these pronouns are optional, but they become obligatory as verbal agreement is lost on regular verbs. Table 5.7 provides the endings for a verb TABLE 5.6 Pronouns in Old French

1 2 3

NOM OBL NOM OBL NOM ACC DAT

S gie, jo, je mei, moi, me tu, te tei, toi, te il, ele, el lui, lo, le, li, la, ol lui, li, lie, lei

P nos, nous nos, nous vos, vous vos, vous il, eles el(e)s, les lor, lour, leur

(after Nyrop II: 392)

TABLE 5.7 Loss of verbal agreement from Old to Modern French

1S 2S 3S 1P 2P 3P

Old French chant chantes chante chantons chantez chantent

Formal French, writing only je chante tu chantes il/elle chante nous chantons vous chantez ils/elles chantent

Colloquial French je chant tu chant il/elle chant on chant vous chantez ils/elles chant

[ʃɑ̃t] [ʃɑ̃t] [ʃɑ̃t] [ʃɑ̃t] [ʃɑ̃te] [ʃɑ̃t]

144

Macro cycles

in Old French and those for how verbs are written and pronounced in Modern French. As is clear from the last column, agreement is mostly lost in current French. Simonenko et al. (2019) examine the relationship between the loss of agreement and the presence of null subjects in Medieval French. They argue that the loss of verbal distinctions “eventually led to the disappearance of grammar-generating null subjects” (280). After the loss of pro-drop, je, tu, il, and elle become the regular singular clitic pronouns, according to Adams (1987), and moi, toi, lui, and elle become the emphatics for both nominative and accusative in the singular, according to Harris (1977). In the modern period, the situation is different from (28) because the subject pronoun is obligatorily attached to the finie verb, as (32a) shows. If it doesn’t immediately precede a verb, it is ungrammatical, as in (32b). (32) a. Je lis et j’-écris I read and I-write b. *Je lis et écris I read and write

Colloquial French

As a result of this obligatory attachment of je to the verb, the pronoun needs to be adjacent to the verb, as in (33a), and it is no longer possible to separate them, as (33b) shows. In contrast, a full noun can be separate from the verb, as (33c) shows. (33) a. J’ai heureusement vu ça French 1S.have fortunately seen that b. *Je heureusement ai vu ça 1S fortunately have seen that ‘I’ve fortunately seen that.’ c. Kurt, heureusement, a fait beaucoup d’autres choses. Kurt fortunately has done many other things. ‘Fortunately, Kurt did many other things’ (Google search of French websites). Lambrecht (1981: 6) also mentions the elimination of clitic-verb inversion, as in (34). Instead, one hears (35). (34) Où vas-tu Standard French where go-2S (35) tu vas où Colloquial French 2S go where ‘Where are you going?’

Macro cycles

145

There are other reasons for arguing that the ‘pronoun’ is agreement, for instance, pronouns such as je ‘I’ and tu ‘you’ cannot be coordinated or modified, as pronouns in other languages can or could in earlier French (30). Where and why does the Subject Cycle start? It could be that pragmatic/ semantic strengthening, for example, by jou in (31), triggers the weakening of the je, or the other way round, namely, that the phonetic weakening of je triggers the presence of an emphatic. Let’s take the Old French in (28) in the previous section, repeated as (36). In (36), the emphatic pronoun is il; it is optional and need not immediately precede the verb ad. If Old French is a pro-drop language (cf. Adams 1987; Vance 1997), it is the null subject that provides the features for the verb’s agreement. (36) Il nen i ad ne veie ne senter Old French He NEG there have.3SPR NEG road nor path ‘He doesn’t have a road or path there’ (Chanson de Roland, 2399, LRC Old French Online). In many languages, pro-drop is lost earlier for first person and later for second and third (van Gelderen 2011). Once pro-drop is lost, the agreement features on the verb need a new pronoun to provide them this information, which they find in the emphatic pronoun. This stage is then reached for all persons, and je ‘I,’ tu ‘you,’ and il/elle ‘he/she’ and others are reanalyzed as valuing the agreement features of the verb. Currently, the last stage of the cycle is taking place because the first and second person pronouns are being reanalyzed as agreement on the verb. This is the reason the topicalization, as in (37), is frequent. (37) Moi, je dit toujours me, I say always ‘I always say that . . . ’

que . . . that . . .

Spoken French

Detges (2018) looks at the erosion of moi in spoken French, especially before the subject marker je, as in (37). For instance, moi preceded by a preposition is stronger (longer and higher pitch) than one preceding je. His conclusion is that, even though moi “has undergone heavy prosodic weakening” (p. 1060), it is still a strong pronoun. Now we have come full cycle where agreement is lost (Table 5.7), a pronoun becomes agreement, as in (33a), and a new pronoun is introduced, as in (37). Three other examples of full or partial Subject Cycles involve the Turkic, Athabascan, and Lezgic families, which I now turn to. According to Kóna-Tas (1998), Turkish has gone from a stage with pronouns, as in (38), to one with agreement marking, as in (39), to one with optional

146

Macro cycles

pronouns, as in (40). As can be expected with a renewing pronoun, Lehmann (1976) says that the optional subject appears when it is prominent. (38) ölteči sen die.FUT you ‘You will die’ (Kóna-Tas 1998: 75).

Old Turkish

(39) ol yola ben eyü qulavuz-van Old Turkish DEM road 1S good guide-1S ‘I am a good guide for that road’ (Adamović 1985: 29, Veled 11). (40) (Sen) ölecek-sin You die.FUT-2S ‘You will die’ (Kóna-Tas 1998: 75).

Modern Turkish

Let’s look at this development in a bit more detail. Adamović (1985: 27) mentions that the subject precedes the verb in Old Turkish except in the case of first and second person pronouns where it follows, as in (41). As in (40), an additional pronoun could be added, as in (42). (41) bay ben Old Turkish rich 1S ‘I am rich’ (Adamović 1985: 27). (42) ben bay ben Old Turkish 1S rich 1S ‘I am rich’ (Adamović 1985: 27). The most important changes, according to Adamović, take place before there are written sources. The latter start in the thirteenth century, and in these, we already see the postverbal subject as a suffix, as in (43) and (44). Quite early, there are other forms of this suffix, for example, -am/-em, as in (44), which is retained in modern Turkish dialects. (43) sen diri-sin Old Turkish 2S alive-2S ‘You are alive’ (Adamović 1985: 38, Veled 15). (44) ben hazir-am Old Turkish 1S ready-1S ‘I am ready’ (Adamović 1985: 31, Zade 111). This suffix becomes -um/-üm (Adamović 1985: 50), as in (41), in the fifteenth century. Currently, the first person endings are -um/-üm/-ïm/-im.

Macro cycles

147

(41) seuinér-um Old Turkish rejoice-1S ‘I rejoice’ (Argenti II 267r). So Turkish shows a change from pronoun, as in (38), to agreement marker, as in (39), to optional pronoun, as in (40). In many languages that show evidence of the Subject Cycle, the pronoun is cliticized onto the verb but remains available as replacement. We saw this in Turkish, and it is also evident in Athabascan, for example, Navajo (44) and (45), where the first and second person agreement markers -sh- and -ni- are the grammaticalized forms of the pronouns shi and ni, respectively, as was also mentioned in Chapter 1. (44) (Shi)

Navajo

(45) (Ni)

Navajo

yáshti yá-sh-0-ti I PRT-1S-CL-talk ‘I am talking.’

yániłti’ Yá-ni-ł-ti’ You PRT-2S-CL-talk ‘You are talking.’

Unlike French, where the oblique form is used as renewal, but like Turkish, the reenforcing pronoun in Navajo is the same word as the one that grammaticalized as an agreement marker. Fuß (2005: 3) provides many other examples, for example, from Basque and the Semitic, Mongolian, Tibeto-Burman, and Bantu families. Bogomolova (2018) examines the rise of agreement in two Lezgic languages, Agul and Tabasaran. Agul presents the earlier situation and lacks verbal subject marking but has pronouns, as in (46). Tabasaran has both, as in (47). (46) čin t’alab aq’-u-f-e allah teʕla.jifas. Agul We.EXCL request do-PF-S-COP Allah powerful ‘We asked powerful Allah’ (Bogomolova 2018: 831, ignoring case marking). (47) uvu dina aˁr-uˁ=va . Tabasaran you there go-PR=2S ‘You will go there’ (Bogomolova 2018: 825). The forms uzu ‘I,’ uvu ‘you,’ and others in Tabasaran are fuller forms of the clitic forms, za/zu and va/vu, respectively.

148

Macro cycles

Bogomolova also provides a bridging context on how the clitic may have come about in Tabasaran. In Agul, subjects can appear pre- and postverbally, and some verbs, especially verbs of ‘saying,’ have a doubling, as (48) shows. (48)

Agul

zun

aʁ-a-a

zun,

zu

wun

qu-χ-a-s

I

say-IPF-PRS

I

I

you

RE-carry-IPF-INF RE.come.PF-PART2-S-COP

qaj-naje-f-e.

‘But I say that I came to pick you up’ (Bogomolova 2018: 838).

Such a doubled subject could have triggered the reinterpretation of one as subject marker, in a postverbal position. So postverbal pronouns, as in (48), could have resulted in the agreement in (47), especially in such a suffixing language. In Section 3.3, I’ll discuss the reasons behind prefixing or suffixing of agreement more. So far, we have discussed subject pronouns reanalyzing as agreement, and we now turn to objects. Compared to subjects, objects are less tightly connected to verbs. In many languages, such as Athabascan, the subject affixes are closer to the verb root, which shows that they were attached to the verb before the object affixes. I’ll first show the stages in an Object Cycle and then provide actual examples. 3.2 Object cycles

A typical Object Cycle is given in (49), another fictitious case for ease of exposition. Let’s say that a language has a fully independent object pronoun, as in stage (49a). Since this pronoun can be coordinated and modified and need not be close to a verb, it is a full phrase. A possible next stage is for speakers to optionally analyze this object pronoun as a head, as in (49b). This head cannot be coordinated or modified and is phonologically dependent on the verb. The next stage might be for the object to be reanalyzed as an agreement marker. Once it has these features, it could be renewed through an emphatic or some other form, as in (49c). The last stage, as in (49d), is similar to the first with the agreement lost and the emphatic counting as the regular argument. (49) a. b. c. d.

I saw yesterday her (and him). I saw ‘r (*and him). I saw’r HER. I saw her.

As in the case of the Subject Cycle, if languages reanalyze pronouns as agreement markers, one expects cross-linguistic surveys of languages to reflect this. Siewierska’s (2013b) data suggest that verbal object marking occurs in 57% of

Macro cycles

149

the languages surveyed, but not much is known about the obligatory appearance of object pronouns. The Object Cycle, identified by Givón (1976: 157), in which a pronominal object is reanalyzed as verbal agreement is more complex than the Subject Cycle in a number of ways. As Haig (2018a, b) shows, this cycle is often disrupted before the pronoun becomes agreement. The object pronoun mostly remains a Differential Object Marker (DOM). Also, person agreement is rare but not gender and number. Haig (2018a: 795) shows that in Old and Middle Iranian, object clitics are in complementary distribution with full NP objects and have case; they are therefore not agreement. Two thousand years later, the situation is still the same in Western Iranian, for instance, in modern Persian, with clitic objects extremely rarely occurring with full object pronouns (as against van Gelderen 2011). Having sketched what an Object Cycle involves, I’ll now provide some examples from the Austronesian family, i.e. from Malayo-Polynesian and Micronesian. As for Malayo-Polynesian, Indonesian has object pronouns and is in stage (49a). Kambera has object agreement for definite objects, and Tukang Besi has agreement with definite, affected objects, which means these are in stage (49b)/ (49c). As for Micronesian, Mokilese is in stage (49a), Marshallese is transitioning to (49b), and Gilbertese is in stage (49b). Indonesian, a western Malayo-Polynesian language, has many types of pronouns, most of which look like they are in stage (a), especially since there are many (honorific) nouns that substitute for pronouns. In a few cases, an object is affixed on the verb, as in (50a); this usually happens with verbs marked by the active voice marker men-, but not with other verbs, as (50b) shows. (50) a.

Dia men-gambil kue itu lalu mem-akan-nya he ACT-took cake that and TR-ate-it b. Dia men-gambil kue itu lalu makan-(*nya) he ACT-took cake that and ate ‘He took the cake and ate it’ (Sneddon 1996: 165).

Indonesian Indonesian

In (50b), the object is dropped probably because an object needs to be licensed by a voice marker, men-. First and second person can also be attached to the verb, but since there is no doubling by a nominal, and the process is optional, all these forms (including (50a)) are still arguments and not agreement. Kambera, a Central-Malayo-Polynesian language, has obligatory object marking on verbs when the object is definite, as in (51), but not when it is indefinite, as in (52a). It could be that it had an earlier stage with only animate definites, but that is no longer the case. A definite object without the object marker nyà is ungrammatical, as (52b) shows.

150

Macro cycles

(51) Mbàda manahu-da-nyà -ka na already cook-3P-3S-PF the ‘They have already cooked the rice.’

uhu rice

Kambera

(52) a. Mbàda manahu-da-ka uhu Kambera already cook-3P-PF rice b. *Mbàda manahu-da-ka na uhu already cook-3P-PF the rice ‘They have already cooked some rice’ (Klamer 1998: 68). Klamer (1997: 918) writes that [t]he cliticization of transitive objects is determined by definiteness, unlike subjects. In other words, definite objects of simple transitive verbs MUST be cliticized on the predicate and their coreferent NPs are always optional, while indefinite objects cannot be marked with clitics but must be expressed by full (indefinite) NPs. In Tukang Besi, a Nuclear-Malayo-Polynesian language, object agreement occurs when the object is affected and the aspect is perfective, as in (53a). In this case, the object is preceded by the nominative article, that is, “is assigned pragmatic prominence” (Donohue 1999: 153). In the non-affected, there is no object marker on the verb, and the article preceding the object is non-nominative; the sentence is non-perfective, as in (53b). (53) a. No-‘ita-aku na iaku 3realis-see-1S.OBJ NOM 1S ‘He saw me.’ b. No-‘ita te iaku 3realis-see CM 1S ‘He is looking at me’ (Donohue 1999: 135).

Tukang Besi

The use of the object markers is determined by givenness rather than by definiteness, according to Donohue, and is perhaps a type of voice-marking, as in the Philippine languages. The source of the object marker is pronominal, that is, aku and iaku in (83a) are clearly related, but so are the others, except perhaps the third person, as Table 5.8 shows. This may show that the cycle started with third person since that form is the least recognizable. Summarizing the three stages in Malayo-Polynesian Indonesian, Indonesian has object pronouns and is in stage (49a), but Kambera and Tukang Besi have object agreement for certain objects, that is, DOM, and are in stage (49b)/(49c). If Haig’s observations for Persian carry over, these languages may never fully develop object agreement.

Macro cycles

151

TABLE 5.8 Pronouns and OMs in Tukang Besi

1S 2S 3S 1(non-S) 2P 3P

Free iaku iko’o ia ikami/ikita ikomiu amai

OM -aku -ko -’e -kami/-kita -komiu -’e

(from Donohue 1999: 113) TABLE 5.9 Pronouns in Marshallese

1S 2S 3S 1P.INCL 1P.EXCL 2P 3P

Emphatic na kwe e kōj kōm kom(i) er

Object eō eok e (human only) kōj kōm kom(i) er/i

Subject i kwō e kōj kōm kom re

(adapted from Willson 2008)

In Micronesian, another branch of the Austronesian language family, there are a number of changes that are represented by different stages in various languages. I will describe the pronoun situation in some detail in Marshallese and then compare it with what Harrison says about the objects in the Micronesian family in general. In Marshallese, subject pronouns are clitics, but a free object pronoun is used in (54) for humans; for non-humans, a marker that looks more like agreement is used, as in (55). This shows that Marshallese is in stage (49a) possibly moving to stage (49b). (54) E=ar denōt er Marshallese 3S=PST slap 3P.H ‘He slapped them (the boys)’ (Willson 2008: 19). (55) E=ar denōt-i Marshallese 3S=PST slap-3P.non-H ‘He slapped them (the fish)’(Willson 2008: 19). All pronouns are provided in Table 5.9. Notice that only first and second person singular object forms are distinct. This suggests that the first and second person singular pronouns are innovative, unlike the situation in Tukang Besi, where third person forms are.

152

Macro cycles TABLE 5.10 Pronouns in Proto Micronesian

1S 2S 3S 1P.INCL 1P.EXCL 2P 3P

Emphatic *ngai *koe *ai *ki(t,t’)a *ka(ma)mi *kamiu *ira

Object *ai *ko *a *ki(t,t’)a *ka(ma)mi *kamiu *ira

(adapted from Harrison 1978: 1081)

Harrison (1978: 1081) reconstructs the pronouns for Proto Micronesian as in Table 5.10 and argues that in the precursor of Proto Micronesian, the emphatic and object pronouns were the same in function and syntax, as they still are in Mokilese: “all pronominal object marking was by means of absolute pronouns” (Harrison 1978: 1082). As Table 5.10 shows, in Proto Micronesian, the emphatic pronoun is also used as the object in the plural, but in the singular, the two types of pronouns are distinct. First and second person are enclitics, according to Harrison (1978: 1082), and the third person is a verbal suffix, which suggests the third person drives this change. The current situation in Micronesian is therefore a continuum, as shown in (56). (56) ABSOLUTE only ABSOLUTE+CLITICS CLITICS only Mokilese, Pingelapese Ponapean, Marshallese, Gilbertese, Trukic Kosraean e.g. (59abc) e.g. (54), (55) (49a) > (49ab) > (49b) Harrison explores the differences among languages and argues that, in some languages, the verb-object attraction “was impeded . . . by the process of final vowel deletion, which created canonical shapes unamenable to the eventual suffixation of object pronouns” (1978: 1100). As mentioned, some transitive verbs have optional Object Markers (OM). There are two kinds: one for pronouns and DPs, as in (57a), and another one for plural non-human objects, as in (57b). (57) a. E-ar pukot-e (kōj) Marshallese 3S-PST look.for-OM 1P ‘He looked for us’ (Willson 2008: 32). b. E-ar denōt-i (kweet ko) Marshallese 3S-PST pound-OM octopus the ‘He pounded the octopuses’ (Harrison 1978:1075).

Macro cycles

153

When the OM is present, the pronominal and DP objects (if the latter are nonhuman) are optional. However, sentences with topicalized objects must have an overt OM on the verb, as in (58); this indicates that the OM is an argument, not an agreement marker, at least in Marshallese. (58) a. Juuj eo a-ō, e-ar shoe the POSS-1S 3S-PST ‘My shoe, he saw.’ b. *Juuj eo a-ō e-ar shoe the POSS-1S 3S-PST ‘My shoe, he saw’ (Willson 2008: 34).

lo-e see-OM

Marshallese

lo see

Marshallese

Harrison (1978: 1068) reconstructs transitive verb markers for Proto-Micronesian that resemble the Marshallese OMs: *-a for singular nominal objects and *-i for pronominals, plural objects, and plural non-human objects. According to him, these are reflected in Gilbertese (59). (59) a. I noor-a te ika 1S see-TV the fish ‘I saw the fish’ (Harrison 1978: 1068). b. I noor-i waa akanne I see-TV canoe those ‘I saw those canoes’ (Harrison 1978: 1068). c. I noor-i-a 1S see-TV-3S ‘I saw her’ (Harrison 1978: 1073).

Gilbertese

Harrison (1978: 1069) further says the transitive markers have been lost in Marshallese, Kosraean, and Ponapeic and that “Marshallese evidence for an earlier *-i/*-a contrast is far from straightforward” (1072) since the reconstructed *ia does not correspond to the third person e [ε]. He excludes other possible sources. Thus, the object pronouns in the Micronesian languages illustrate various stages of the Object Cycle. In some (Mokilese), there are only free pronouns, in others (Marshallese), there are pronouns as well as clitics, and in yet another group (Gilbertese), there are only agreement markers. As far as I can tell, the sources for object markers in Austronesian are pronominal, but the start is unpredictable: possibly first with voice markers in Indonesian and Tukang Besi, with definite DPs in Kambera, and with first and second person pronouns in Marshallese. In all these languages, the order of the verb and clitic/affix reflects the original order of verb and free object. Thus, Ross (2004) shows that canonic Oceanic languages have SVO structure and preverbal subject markings and postverbal object markers, as in (60) and (61).

154

Macro cycles

(60) ko=mese wehe=ra Ambae 2S=DEHORT hit=3P ‘Don’t hit them’ (Hyslop 2001: 235, from Ross 2004: 500). (61) na=ni hui=ko mwere vake 1S=IRR ask=2S like too ‘I will ask you this as well . . . ’ (Hyslop 2001: 237, from Ross 2004: 501). I now turn to cases where the subject and object markers do not follow the order of the free forms. 3.3 Morpheme order

Narrog (2017: 158), among others, asks: why are subjects not more often marked as prefixes since SV order is common? Why is it “unstressed pronominal subject pronouns postposed to the verb, rather than pronouns in their normal position that get grammaticalized”? Kirby (2022) identifies this problem as ‘Wrong-side Affixation.’ In this section, I will discuss the expected and unexpected order of clitics and affixes. For instance, in SVO and SOV languages, one expects prefix subject markers. However, suffixed subject markers are more common, especially in V-last languages, as Table 5.11 shows. I will first discuss subject markers and then turn to object markers. For V-initial languages, the conclusions depend on how those languages are derived. If, as in Celtic and Arabic, the V-initial order derives from an SV order with subsequent verb movement, the prefixes (60%) in Table 5.11 are expected. The suffixes (40%) would have grammaticalized in the V-initial order. For V-medial languages, the data in Table 5.11 are mostly as expected because the subject pronoun precedes the verb and would grammaticalize as a prefix. We have seen such a prefix developing in the history of French. The cases where it is a suffix (27%) could be explained by SVO languages having optional VS word orders and the suffix grammaticalizing in that position. There is some evidence from the history of English for this. In the stage of (62), SVO is common, but all verbs can invert and move to a pre-subject position. Indeed, in (62), subjects are cliticized after a preposed verb. TABLE 5.11 The order of V and subject agreement affix

A(gent) agreement Prefix Suffix (Siewierska 2004: 165)

V-initial 15 (60%) 10 (40%)

V-medial 41 (73%) 15 (27%)

V-last 29 (30%) 70 (70%)

Total 85 95

Macro cycles

155

(62) a. That art so pale and deedly on to see? Why crides=tow? REL are so pale and deadly on to see. Why cried=2S ‘Who is so pale and looks so deadly. Why did you cry?’ (1400, Chaucer, Knight’s Tale, 224–225). b. What wiltow seyen? thou woost nat yet now Wheither she be a womman or goddesse! What will=2S say? 2S know.2S not yet now whether she is a woman or goddess ‘What do you want to say? You don’t yet know if she is a woman or goddess’ (1400, Chaucer, Knight’s Tale, 298–299). What is not expected is the tendency for V-last languages to have suffixed subjects (70%), as Table 5.11 shows. Navajo, which is V-last, is an exception to this, as is Yuman because they have subject marking prefixes. For other V-last languages, there is a reason to be suffixing, and the developments in Turkic and Lezgic, discussed in Section 3.1, provide that reason. Agreement in Turkish and Lezgic is post verbal, that is, suffixal, because some pronouns occur in that position, for example. in (38) and in (48), and can be reanalyzed as such. If grammaticalization is of phonologically adjacent items, suffixes should be rare in both SVO and SOV languages, unless there are post verbal subjects. As is the case of Turkish and Lezgic, Comrie (1980: 211) writes that in Mongolian languages, “it is possible to place the subject, especially a pronoun, after the verb to de-emphasize that pronoun, thus giving rise to a VS order” and the possibility for suffix agreement markers to arise. Fuß (2005) also discusses a number of scenarios for subject agreement suffixes to arise, as does Givón (1976), of course, who argues they arise from earlier postverbal inflected auxiliaries. Concerning object agreement, we’d expect suffixes for the V-initial and V-medial languages and prefixes for the V-last languages. That is true in the majority of cases but not completely, as Siewierska (2004) shows. See Table 5.12. Siewierska (2004: 166) offers Givón’s (1976) solution, which is that the suffix came about because of a prefix on an auxiliary following a nonfinite form can be reanalyzed as a suffix on the nonfinite verb. The reasons for suffixing tendencies given in this section are framed in terms of contexts where the verb precedes the pronoun and the latter can reanalyze. How come pronouns preceding the TABLE 5.12 The order of V and object agreement affix

P(atient) agreement Prefix Suffix (Siewierska 2004: 165)

V-initial 7 (33%) 21 (67%)

V-medial 19 (36%) 34 (64%)

V-last 42 (56%) 31 (42%)

156

Macro cycles

verb don’t do the same? Reinöhl & Casaretto (2018) discuss this in terms of ditropic clitics, but this still describes the phenomenon rather than explaining it. In Section 3, I have given a definition of the Subject and Object Agreement Cycles keeping in mind the difficulties of distinguishing pronouns, clitics, and agreement. After that, examples of various languages have demonstrated various stages. In French, we see a complete Subject Cycle, with a loss of agreement, the rise of a new agreement marker, and a renewal of the subject pronoun. In Turkish, we see a Subject Pronoun to Agreement Cycle with renewal of the pronoun but not with a loss of the agreement, and the same is true in Lezgic languages. I added Athabascan because the optional subject pronoun has the same shape as the agreement marker, something that is not uncommon in the pronoun cycles. The Object Cycle shows the same stages as the Subject Cycle in (49), and examples have been given by regarding languages in the same family as being in different stages of the cycle. Finally, I discuss the question of why suffixed subjects would occur, especially in V-last languages. 4

Case cycles or dependent marking cycles

In this section, I first provide some background to case marking and a definition of the Case Cycle. As languages lose and gain case, we will see that the circumstances under which this happens are not always clear, nor are the mechanisms. The sources from which case is developed are clearer, but again the conditions that require new case markings are not. Examples of (partial) Case Cycles are subsequently offered. Case is frequent in the world’s languages, according to Nichols and Bickel (2013a, b). As shown in Table 5.12, repeated here as Table 5.13, dependent marking is a little less frequent than head marking but not by much. If we add the double marking languages, 51% of the languages (121/236) in the sample have dependent marking. Case can be divided into the case that marks grammatical relations, such as subject and object, and the case that marks semantic relations, such as the instrumental, comitative, and locative. These are sometimes referred to as core/grammatical and peripheral/semantic case, respectively. For both types, case gain and loss can be formulated, as in (63). TABLE 5.13 Dependent marking

Head marking Dependent marking Double marking No marking Other Total

71 63 58 42 2 236

(from Nichols & Bickel 2013a)

Macro cycles

(63) Case Cycles involve DET/P/V/N > case

>

157

zero

As for the loss of case marking, we have seen this change in Section 2 between Old and Modern English, between Sanskrit and Hindi/Urdu, and between Latin and French, all Indo European languages. Case loss has been explored a lot in this family, for example, Allen (1997) and Trips (2001) for English, and Sundquist (2006) for Norwegian. Although there is often a relation between case loss and stricter SVO word order, this is not a necessary condition as, for instance, Western Iranian languages have lost case but retain SOV (John Powell p.c.). In addition, Sundquist states that “fluctuation between [O]V and V[O] is more attributable to information structure and interpretive restrictions on leftward movement than presence of rich morphological case” (2006: 107). Others share that sentiment; for example, Hróarsdóttir (2000) shows there is a loss of OV in Icelandic without a loss of case, Pinkster (1990) shows that only 5–10% of the cases in Latin are functional in determining the grammatical function, and Parodi (1995) shows that Spanish loses case before it moves to a VO word order. Dimmendaal (2005) describes a cycle of a loss and a gain of dependent marking in the history of Nilotic. Nilotic is verb-initial in origin with SVO as an alternative word order, according to Dimmendaal (2005: 75–76). When the subject follows the verb, it is case marked, but when it precedes, it is not. Therefore, as some of the languages shift to SVO, the case marking of the subject is lost. In addition, there is a lot of evidence that the locative, directional, and instrumental cases are also lost in favor of a system where these functions are marked on the verb. However, also according to Dimmendaal (2005: 81), the Eastern-Nilotic languages of the Teso-Turkana dialect cluster have developed new case distinctions, possibly under the influence of the neighboring Surmic languages. So these languages have experienced a cycle of loss and gain of grammatical case. As for the origin of case markers, Blake (2001: 161) says “[t]here are two common sources for lexical markers, one verbal and the other nominal, of which the verbal is probably the more fruitful. Adverbial particles also provide a source.” The sources are different for core and peripheral cases, and I start with the latter, shown in (64). When morphological case endings disappear, adpositions may take over (e.g. with and through replace the instrumental), and they may originate in adverbs, nouns, and verbs. For instance, the locative preposition in (64) has the Latin noun via ‘road’ in (65) as its origin. (64) Semantic Case Adv N P > V

}

affix

>

zero

158

Macro cycles

(65) a. I went via Paris. b. via Paris Latin road Paris ‘The road to Paris.’ Other instances of nouns that reanalyze to adpositions include the Italian senza ‘without’ from Latin absentia, baka in Sranan from ‘back’ (see Plag 1998), tp ‘on’ from ‘head’ in Egyptian (Gardiner 1957 [1988]: 130), various Zapotecan languages (Lillehaugen 2006), Ewe ta from ‘head,’ as in (66), and Swahili juu from ‘top,’ as in (67). (66) a. e-fe ta Ewe he-of head ‘his head’ b. e-ta he-on ‘on him’ (from Heine & Reh 1984: 257). (67) a. juu ya mlima Swahili top of hill b. juu ya mlima on hill (from Heine & Reh 1984: 101) Kahr (1975: 45–46) and Heine and Kuteva (2002) provide numerous other examples, like the Tzotzil ‘ear’ becoming a locative preposition, the Finnish noun ‘earth’ reanalyzing to the adverb/preposition ‘down,’ ‘shoulder’ becoming ‘up,’ and ‘heart’ becoming ‘in.’ Finnic originally has an -ine comitative that is lost in many of the family’s languages and is renewed by various forms of a postposition, for example, in (68). The latter, however, in its turn is derived from a noun kansa ‘people, company’ (Oinas 1961: 12–13) and also used as an adverb, as in (68) from the now extinct Salis Livonian. (68) Utak mind kazu Salis Livonian take me with ‘Take me along’ (Oinas 1961: 13). This comitative surfaces as -kaas as early as the sixteenth century in North Estonian, as -kaan in Southern Estonian, and later as -ga in South Estonian. Different dialects still preserve various forms, as in (69). (69) lapsien-ka Sippola Finnish children-with ‘with children’ (Oinas 1961: 44). Oinas discusses a range of relatively recent case markers, for example, the kecomitative, originally from a noun kerta ‘time’ (Oinas 1961: 60); the mö-prolative/comitative, from myö ‘back’; and a sa-terminative, from a verb meaning “to

Macro cycles

159

receive, get, come, arrive” (Oinas 1961: 139). So adpositions develop into case markers but are themselves derived from nouns and verbs. Turkish has a postposition ile ‘with’ that is starting to be suffixed to the noun, as in (70). This postposition too may have a more lexical origin. (70) Mehmet-le Turkish Mehmet-with ‘with Mehmet’ (Lewis 1967: 88, from Kahr 1976: 121). As for the grammatical cases, their origin is often more grammaticalized already, for example, a demonstrative or a bleached preposition such as of and to. Greenberg’s Definiteness Cycle, discussed in Chapter 3, of course fits here since definiteness develops into case in his formulation of that cycle. Tauli (1956) provides an in-depth overview of the debate on the (pro)nominal sources of these case particles that has gone on at least for the last century and a half. I will argue for the developments shown in (71). (71) Grammatical Case D > V > case P >

}

>

zero

The change of D to grammatical case was widely accepted by the early IndoEuropeanists such as Bopp (1833), Uhlenbeck (1901), Pedersen (1907), Delbrück (1919), and Specht (1947) but is currently not discussed much (exceptions are Greenberg 1978 and Kortlandt 1983). For early Indo-European, it has been argued that the nominative -s ending is a demonstrative. Bopp (1833: 248) is the most definite in (72) about case endings in general. He finds the origin of both the nominative and accusative in the pronoun. For the latter, he lists the forms ima ‘this one’ and amu ‘that one’ (1833: 323). Delbrück puts a ‘probably’ in and is more specific that it holds only for the nominative in (73). (72) Ihrem Ursprunge nach sind sie, wenigstens grösstentheils, Pronomina. [‘In origin, they are for the most part pronouns’] (Bopp 1833: 248). (73) Es ist wahrscheinlich, daß das s aus dem Demonstrativum stammt, mit dem auf das hervorragende Substantivum hingewiesen wurde. [‘It is likely that the s derives from the demonstrative, which referred to the preceding nominal’] (Delbrück 1919: 215). The most common Sanskrit nominal declension is that of noun stems in -a, given in Table 5.14 for the singular of deva ‘god’ and for the masculine singular demonstrative. These paradigms show the similarities between the endings of the nouns and

160

Macro cycles

TABLE 5.14 Cases for Sanskrit -a stems for deva ‘god’ and the demonstrative pronoun

NOM ACC INST DAT ABL GEN LOC VOC

S devas devam devena devaaya devat devasya deve deva

MS Demonstrative sa(s) tam tena tasmai tasmaat tasya tasmin –

those of most of the demonstratives. The argument would be that the noun combined with an original demonstrative (as, am, etc.) and that the original demonstrative was in turn strengthened with a new deictic s-/t- (and with tasm-) for the locative cases. Greenberg (1978: 73–74) maintains that the origin of nominative case is often a definite marker (since subjects are most often definite), and the same is claimed by König (2008: 117) for the origin of ergative Case in West Nilotic. Sasse (1984) has argued for a demonstrative origin of the cases in Berber, and Kulikov (2006: 29–30) provides a review of languages for which this has similarly been argued, for example, Kartvelian, Georgian, and Caucasian. For instance, the ergative -man in Georgian (Lomashvili p.c.) is a postposed pronoun, and the same is probably true of the absolutive -I, going back to -igi ‘that’ (see Kulikov 2009: 447). McGregor (2008) shows that some ergative suffixes in Australian languages derive from pronouns. Mithun (2008b: 215), based on work by Anderson (1992), argues that the Wakashan language Kwakw’ala’s subject marker derives from a proximal demonstrative and the object marker from a distal one. The demonstrative origin of the nominative is widely accepted, certainly in the nineteenth century, but the demonstrative origin of the accusative is less widely agreed upon than that of the nominative. As mentioned, Bopp thinks the accusative has a demonstrative origin. Similarly, the accusative ending -m in Uralic is originally argued to be a demonstrative pronoun by Tauli (1956: 176; 198; 210) and others. That -m is the same element as is present in Finnish tämä ‘this’ and Khanty tam ‘this’ and tom ‘that.’ R. Baker (1985: 134–135) provides an overview of the different hypotheses. The accusative counts as a DOM rather than as a case marker, according to Baker, as befits its demonstrative origin. However, Finck (cited in Royen 1929: 891–892) sees the Latin accusative -m as originating in a locative noun *medhio ‘middle.’ Locative nouns often develop into adpositions, which in turn change into case. Having considered the demonstrative origin of (nominative) case, I turn to verbs as sources, and these are more often the source of accusatives. The Chinese marker ba is perhaps the best known example of a verb being reanalyzed as an object marker, with an enormous literature explaining its properties. Later in

Macro cycles

161

this section, I will mention the possibility that the Urdu/Hindi object marker ko originates from a nonfinite verb. The possibility is remote, but if so, the account is roughly the same: a verb and its object are reanalyzed as object marker and object. Ba is originally a lexical verb ‘to hold,’ as in (74a) from the fifth century bce, but is also a frequent serial verb, as Li and Thompson (1974: 202) show with sentences such as (74b). (74) a. Yu qing ba tian zhi rui-ling . . . Old Chinese Yu himself take heaven PRT mandate ‘Yu himself took the mandate of heaven . . . ’ (Me-zi, Li & Thompson 1974: 202). b. . . . yin ba jian kan should hold sword see ‘I should take the sword and see it’ (Tang dynasty poem, Li & Thompson 1974: 202–203). In Modern Chinese, ba in (75) has been analyzed as verb, co-verb, preposition (A. Li 1990), or case marker (C. Li & Thompson 1974: 203; Koopman 1984; Yang 2008; Blake 2001: 164). Li and Thompson (1981: 465) mention the definiteness and the affectedness of the object in this construction. (75) wo ba shu mai le Chinese I PRT book buy PF ‘I bought the book’ (Li & Thompson 1981: 21). The ba is obligatory when the object is animate and highly affected. This has led many to see it as a differential object marker. Having provided examples of nominal, verbal, and determiner sources of case, I now turn to adpositions and adverbs. It is, however, very hard to find clear cases of adverbial origins. Blake (2001: 170–171) expresses similar feelings. Kuryłowicz (1964: 201) states “all oblique [peripheral] cases and even the acc[usative] go back to expressions of spatial relations,” and that may be correct. The “secondary semantic functions” betray this origin and “[t]he polysemy of case-forms, due to the increasing range of their use, calls for a constant renewal of their etymological spatial value” (p. 202). Kahr (1976) states that adpositions are the only source for case markers; a noun or verb or adverb always has to go through the stage of adposition before it is reanalyzed as case. Kuryłowicz (1964: 171; 179) suggests a change of adverbs to (prepositional) case markers, as in (76a), noting that some adverbs also can become preverbs, as in (76b). (76) a. [Verb + Adverb] + oblique noun > Verb + [P + noun] b. [Verb + Adverb] > [affix-Verb] (+ Adverb)

162

Macro cycles

Only (76a) is relevant to case/dependent marking; (76b) marks the aspect on verbs. Unfortunately, Kuryłowicz gives ample examples of (76b) but not of (76a). Vincent (1999) provides a few instances where a later preposition is functioning as adverb, for example, úd in (77), and Gary Miller (p.c. 2009) provides (78), with the Latin adverb foras ‘out’ as the precursor of the French preposition hors ‘outside.’ Note the doubling through ex-. (77) úd usríyā jánitā yó jajāna Sanskrit out cows creator who created ‘who as creator created forth the cows’ (Vincent 1999: 1119, quoted from Hock 1996: 221). (78) omnīs ex-ēgit forās Latin all.ACC.P.M out-drove.3S out ‘all (of us) he drove out’ (Aulularia 414). Vincent also provides a scenario as to how this adverb could have been reanalyzed. Because preposing of the adverb was frequent, for instance úd in (77), the reanalysis of an independently marked nominal (in this case usríyā) could occur. Based on work by Bréal (1924) and Lehmann (1958), Fairbanks (1977) reconstructs the Proto Indo European case system as having five cases for the singular with the three additional endings in Sanskrit having been added through adverbial particles. According to Fairbanks, Sanskrit has a suffix -tas, which derives from PIE -tos (cf. Greek and Latin ‘from’), but is used only as an adverbial. In Pali, it is used for the ablative case. “[It] must have derived from some noun form obscured by the passage of time” (Fairbanks 1977: 117; 122). So the evidence for adverbial origins of case markers exists but is not extremely frequent. In conclusion, this section has shown that dependent marking is common among the world’s languages. The sources of this marking are varied: demonstratives are common, as known from Greenberg’s Cycle, but so are nouns and verbs. Nouns and verbs develop into adpositions, which are frequent case markers. Less is known about adverbs as a source, possibly because they typically modify verbs, not nouns. Case loss is known from Indo European, Nilotic, and Finnic. 5 Interactions involving macro cycles

In this section, I examine how macro cycles interact. Typologically, the most interesting situation is when the head and dependent marking cycles act together. The change from Latin to French certainly involves an increase in the former and a loss of the latter. However, it is not clear these changes actually interact. A language can do without either agreement or case, as shown by creoles and very

Macro cycles

163

analytic languages. Therefore, I will discuss instead the most obvious interaction that involves the Subject Cycle in French. As we’ve seen in Section 3, the subject pronoun merges with the verb to become an agreement marker. In this process, other elements may interfere: the object clitic, the locative, and the negative. I will mostly discuss the clitic object and locative pronouns here since the preverbal negative has almost disappeared in spoken French. In colloquial French, as we’ve seen in Section 3, first and second person preverbal subject pronouns function as agreement markers on the finite verb because they are obligatory and adjacent to the finite verb. In other spoken varieties of French, third person pronouns are also agreement markers, having lost gender and number (Fonseca-Greber 2000 for Swiss French). However, preverbal pronominal objects are a factor complicating the changes affecting subjects. Subject agreement is not adjacent to the verb if an object, or a negative or adverbial, appears in between. Indeed, Bahtchevanova and van Gelderen (2016) show some replacements of these preverbal objects by postverbal objects, some deletion, and some reanalysis to agreement markers. That means that the Subject Cycle accelerates other cycles, in this case the Object Cycle, and they may have contributed in the loss of the negative ne in the Negative Cycle. In colloquial French, the preverbal ne is almost absent, but Ashby (1976) notes that full subjects and nous/vous subjects retain ne more. The reason is that these are full phrases, and here the negative does not interfere with the Subject Cycle. The remainder of this section focuses on the interaction of the Subject and Object Cycles and is based on Bahtchevanova and van Gelderen’s data. Three responses are possible to the problem of an (object) pronoun that is located between an incipient agreement marker and the verb, namely, (a) the object pronoun’s loss, and thereby a loss of transitivity; (b) a change in its status, from pronoun to agreement; and (c) a change in position, from preverbal to postverbal object. Much is known from the literature about the phenomenon in (a): object loss, as between (79a) and (79b), has been documented by Lambrecht and Lemoine (1996), Larjavaara (2000), and Noailly (1997) for adult French and for L1 acquisition by Jakubowicz et al. (1997) and Grüter (2006a) and for L2 acquisition by Grüter (2006b). (79) a. Je l’ai trouvé 1S it.have found b. J’ai trouvé hier. 1S.have found yesterday ‘I found it yesterday.’

hier yesterday

Object loss is, of course, the least complex way to remove the object pronoun, but since it loses a lot of the underlying structure, it probably cannot occur full scale.

164

Macro cycles

TABLE 5.15 Portmanteau morphemes with first and second person subjects and other

objects

1S subject je and 1S object and 2S and 3SM and 3SF and 1P and 2P and 3P and 3SM.DAT and 3P.Dat and en and y

je me je te je le je la je nous je vous je les je lui je leur j’en j’y

[ʒәm] or [ʒmә] [ʃtә] [ʒәl] or [ʒlә] [ʒәla] or [ʒla] [ʒnu] [ʒvu] [ʒәle] or [ʒle] [ʒәlɥi]], [ʒlɥi], [ʒɥi], or [ʒi] [ʒlœʁ] [ʒɑ̃] [ʒi]

2S subject tu tu me [tym] tu te [tyt] tu le [tyl] tu la [tyla] tu nous [tynu] tu vous [tyvu] tu les [tyle] tu lui [tylɥi] tu leur [tylœʁ] tu en [tɑ̃] tu y [tɥi] or[ti]

The second change is related to the status of the pronoun. There are specific changes in the phonology of preverbal markers, as outlined in e.g. Morin (1979). There is variation in how two preverbal syllables with a schwa vowel are pronounced in spoken French: either, as in (80a), the second schwa deletes or, as in (80b), the first schwa deletes. Table 5.4 provides some of these combinations. (80) a. [ʒәlvwa] b. [ʒlәvwa] je le 1S him ‘I see him.’

vois see

If both subject and object pronouns are reanalyzed as agreement markers, we could analyze [ʒlә] and [ʒәl] in (80) as portmanteau morphemes spelling out the features of the subject and object. The evidence for such a reanalysis would be if le, y, and en became obligatory with the verbs that select PPs. Instances such as (81), where the clitic and the PP are both present, have indeed become frequent in colloquial French but haven’t reached the stage where they are obligatorily used. (81) a. J’y vais à 1S.there go to ‘I am going to the pool.’ b. J’en parle de 1S.about.it speak about ‘I am talking about the film.’

la the

piscine pool

ce that

film film

There is evidence in the corpora for this change in some speakers, as shown in (82), from the Orléans corpus.

Macro cycles

(82) a. mais en FRANCE ils but in France 3P ‘But in France, they eat bread.’ b. j’ en parle de ça 1S.about.it talk about that ‘I talk about it at the same time.’ c. comme des fois on like the times 1P ‘Like when we discussed it.’

en it

mangent du eat PRT

en at

même same

165

pain bread

temps time

en discute de ça about.it discusses about that

The third reanalysis that could occur is a change in the position of the pronoun, as from (83a) and (84a) to (83b) and (84b) respectively. This situation is complicated because a myriad of clitics can appear, both locative and argumental y, partitive en, and various object clitics. What follows only examines y and le, l’, and la. (83) a. J’y travaille 1S.there work ‘I work there.’ b. Je travaille 1S work ‘I work there.’

là there

(84) a. L’agressivité, j‘y travaille Aggression 1S.on.it work ‘Agression, I am working on it’ (from www.leparisien.fr). b. Je travaille/pense à ça 1S work/think on that ‘I’m working on it.’ This shift in position is a change we can check for using corpus data. In the Orléans corpus, there are 844 instances of y- we have excluded the existential il y a ‘there is/are,’ idiomatic expressions, and y preceding nonfinite verbs. This leaves 45 instances of y preceding the verb, as in (85). (85) a. j’ y ai laissé pour ainsi dire IS.there have left for so say ‘I have left my health there so to speak.’

ma santé my health

b. les cours on y va on y the classes 1P there go 1P there ‘We go to classes, we don’t go to classes.’

va go

pas NEG

c. ça choque pas on y est habitué maintenant this shocks NEG 1P to.it is accustomed now ‘This doesn’t shock (us), we are used to it now.’

166

Macro cycles

d. j’ y suis un peu habituée du fait que MICHEL 1S.to.it am a little accustomed to.the fact that MICHEL euh il cause anglais très bien eh 3S speaks English very well ‘I am somewhat used to the fact that Michel speaks English very well.’ In comparison, the postverbal là, là-dedans, là-dessus, là-bas, and à ça, as in (86), which can be replaced by y, are much more frequent, numbering 70. (86) a. qui veulent que leurs enfants aillent jouer là-dedans who want that their children go play over-there ‘who want for their children to go play over there.’ b. il y a presque personne qui va là-dedans ‘ 3s there is almost nobody who goes there ‘Hardly anybody goes there.’ c. ils se basent là-dessus they REFL base there ‘They are based on that.’ Bahtchevanova and van Gelderen (2016) also look at le, l’, and la objects before finite verbs in the Orléans Corpus, as in (87), of which there are 196 occurrences. In comparison, there are 106 cases with a postverbal object ça, as in (88), that could be le, l’, or la, that is, 35% of the combined numbers. The data appear in Table 5.16. (87) la langue de Bretagne je ne l’a comprends pas the language of Brittany 1S NEG it.have understood NEG ‘I don’t understand the language of Brittany.’ (88) le gouvernement aura compris ça the government should.have understood that ‘The government should have understood it.’ These data indicate that, although preverbal markers are more frequent, the postverbal pronouns are not rare. Objects occur between the subject and verb between 9% (Orléans Corpus) and 16% (CdES). In this section, we have seen that preverbal objects occur between subjects and verbs in French (possibly between 9 and 16%) and may interfere with the Subject Cycle. However, some of these objects are lost, reordered, or reanalyzed. These are alternative strategies to avoid preverbal objects. We don’t have spoken data from earlier periods so it is hard to see a change in progress. The

Macro cycles

167

TABLE 5.16 Some pre- and postverbal objects in the Orléans corpus

preverbal locative argument Total

postverbal y le, l’, la

45 196 241

locative argument Total

là, etc ça

70 106 176

change in the behavior of pronominal objects triggered by the grammaticalization processes affecting the subject markers can be understood in terms of an Object Cycle. 6

Conclusions

This chapter has studied macro cycles. Unlike the cycles described in the previous chapters, these cycles shift the typology of a language. For this kind of change, the focus has traditionally (e.g. Hodge) been on the change from analytic to synthetic and then to analytic. The Analytic Synthetic Cycle is the crudest of all macro cycles, however, because it doesn’t specify where the syntheticity is located. There are other typologies that are more precise. Thus, languages can also be seen as head marking, for example, when the (synthetic) marking shows agreement on the verb, or dependent marking, for example, when case is marked on the noun. Heads and dependents can both be marked, or neither can, and again, this is a sliding scale. I argue that these macro cycles are more usable. This chapter has first looked at the difficulties with the terms synthetic and analytic and provided three examples of (almost) complete Analytic Synthetic Cycles. However, full macro cycles are rare: Egyptian is a good example of a full cycle where both verbs and nouns go through the same cycle. The chapter then shifts attention to changes towards agreement, that is, head marking, in the Pronoun Cycle and changes towards case, that is, dependent marking strategies, in the Case Cycle. Macro cycles also interact with each other and with micro cycles. For instance, the change from pronoun to agreement can be interrupted by other material intervening between the subject and the verb. Suggestions for further reading On classifying languages: Greenberg (1960), Plank (1999), Szmrecsanyi (2012). On macro cycles: Haspelmath (2018) On agreement: Siewierska (2004); Fuß (2005) On adpositions and cyclical change: Givón (2021) On variations in synthetic/analytic stages: Mukherjee (2019)

168

Macro cycles

Review questions and exercises

1. 2. 3.

4. 5.

6. 7.

If you were going to calculate the syntheticity of English, as in Bickel and Nichols (2013), what would the maximum categories-per-word (cpw) be for the English noun? How does Bickel and Nichols’ (2013) method compare with that used by Szmrecsanyi (2012, 2016)? Jamaican is an English-based creole and Tawala, a language of Papua New Guinea. How would you characterize their analytic/synthetic nature, respectively, based on the following sentences? (1) Yu dash we mi gud bran nyuu sipaz we mi 2S dash away 1S good brand new slippers REL 1S jos bai laas wiik? just buy last week ‘You threw away my perfectly new slippers which I just bought last week?’ (Farquharson 2013: 89). (2) Kedewa kamkam i-uni-hi Tawala dog chicken 3s.A-kill-3P.P ‘A dog killed the chickens’ (Ezard 1997: 99, from Siewierska 2013b). How would you characterize them in terms of head and dependent marking? In Chapter 1, I mentioned that English subject pronouns, such as I, are bound to their verbs (clitics or affixes) and cannot be used on their own. This claim is controversial because they can occur in coordination with a full NP, although that sounds hypercorrect to many speakers. To help solve this issue, use a corpus and think of a method to determine if English is undergoing a Pronoun to Agreement Cycle. Note that the spoken part of a corpus might be more helpful than the academic part. Look at the pronoun paradigm of Arabic (table adapted from Albuhayri 2013) and note some of your observations. Speculate on the anomalies between the independent pronouns and agreement. Dryer (2013d) has the following table. How does this table compare with Table 5.11?

Macro cycles

169

TABLE 5.17 Pronouns and subject agreement in Arabic

Pronoun Ɂanaa naħnu Ɂanta Ɂanti Ɂantuma Ɂantunna huwa hiya hum hunna

1S 1P 2SM 2SF 2PM 2PF 3SM 3SF 3PM 3PF

Agreement – tu – naa – ta – ti – tum – tunna – (a) – at – wa/-uu/‫ـ‬ – na-

1S1P 2SM 2SM 2PM 2PF 3SM 3SF 3PM 3PF

TABLE 5.18 Suffixing and prefixing

Little affixation Strongly suffixing Weakly suffixing Equal prefixing and suffixing Weakly prefixing Strongly prefixing

141 406 123 147 94 58

Advanced exercise

Is the language you have selected in Chapter 1 more synthetic or analytic? Does it mark heads or dependents, both or neither? What might the sources of this marking be? Is there any evidence from earlier stages or sister languages that an earlier stage was different?

APPENDIX TO CHAPTER 5

Using Szmrecsanyi’s method, I will measure the analyticity and syntheticity of a year from The Peterborough Chronicle at the stage where we see articles being introduced, the year 1140. I will be using 100 words of the early Middle English text and 100 of the Modern English translation of that text. The words that have endings or stem alternations are in bold and the analytic words are underlined. The differences are minor: the earlier text is, as expected, more synthetic and less analytic. Compare that with the earlier Beowulf, which is predominantly synthetic. Early middle English Synthetic vs analytic = 41/100 vs 29/100

& fuhten suythe on Candelmassedæi agenes heore lauerd. & namen him, for his men him suyken & flugæn. & læd him to Bristowe & diden þar in prisun. & feteres. þa was al Engleland styred mar þan ær wæs. & al uyel wæs in lande. þerefter com þe kynges dohter Henries þe hefde ben Emperice in Alamanie. & nu wæs cuntesse in Angou. & com to Lundene & te Lundenissce folc hire wolde tæcen. & scæ fleh & forles þar micel. þerefter þe biscop of Wincestre Henri þe kinges brother Stephnes spac wid Rodbert eorl & wyd þemperice & suor

Macro cycles

171

Modern English Synthetic vs analytic = 35/100 vs 36/100

And they fought strenuously on Candlemas day against their lord, and took him; for his men forsook him and fled. And they led him to Bristol, and there put him into prison in close quarters. Then was all England stirred more than ere was, and all evil was in the land. Afterwards came the daughter of King Henry, who had been Empress of Germany, and now was Countess of Anjou. She came to London; but the people of London attempted to take her, and she fled, losing many of her followers. After this the Bishop of Winchester, Henry, the brother Beowulf, c1000 Synthetic vs analytic = 77/100 vs 18/100

hwæt we gardena in geardagum þeodcyninga þrym gefrunon hu ða æþelingas ellen fremedon Oft Scyld Scefing sceaþena þreatum monegum mægþum meodosetla of teah egsode eorlas syððan ærest wearð feasceaft funden he þæs frofre gebad weox under wolcnum weorðmyndum þah oð þæt him æghwylc þara ymbsittendra ofer hronrade hyran scolde gomban gyldan þæt wæs god cyning ðæm eafera wæs æfter cenned geong in geardum þone god sende folce to frofre fyrenðearfe ongeat þe hie ær drugon aldorlease lange hwile Him þæs liffrea wuldres wealdend woroldare forgeaf Beowulf wæs breme blæd wide sprang Scyldes eafera Scedelandum in swa sceal geong guma gode

6 EXPLANATIONS AND MECHANISMS

So far, we have encountered examples of micro and macro cycles. The explanation behind cyclical change given in Chapter 1 involves the interplay between comfort and clarity. That discussion will be the basis for subsequent accounts to be considered in the current chapter. In historical linguistics, accounts often emphasize an external or internal cause, and selecting one cause over the other will result in a different focus. External accounts of grammaticalization and the cycle are often descriptive in showing that languages that are in contact may be in similar stages of a particular cycle, especially a micro cycle, or that contact accelerates grammaticalization and cyclical change. Internal accounts focus on cognitive reasons behind change and postulate universal economy principles based on cycles. A question frequently asked, when using strictly internal accounts, is why some languages change faster than others and why some languages do not change at all. From an internal change perspective, the ultimate answer to the question regarding the speed of cyclical change (and change in general) remains elusive. Attractor states provide a way to focus on this problem, and I will finish the chapter with a discussion of that topic. Some of the discussion in Sections 4 and 5 is based on van Gelderen (2011, 2022). The outline is as follows. In Section 1, I start with the traditional account given in von der Gabelentz and focus on the use of ‘Bequemlichkeit/Deutlichkeit’ and ‘comfort/clarity.’ A major question we have considered is which of the two tendencies sets the cycle in motion. In the sections that follow, I turn to external and internal explanations. In Section 2, I examine several studies that acknowledge the role of contact, that is, external influence, one on negatives in varieties of Arabic, one on the perfect in European languages, one on grammatical replication, and one that shows evidence of accelerated grammaticalization DOI: 10.4324/9781003272564-6

Explanations and mechanisms

173

in a historical text from English. I continue with internal explanations in Sections 3 to 5. Section 3 considers accounts of grammaticalization and cyclical change using Construction Grammar, and in Sections 4 and 5, two (different) generative accounts of cyclical change are given, one based on the structures and features that are relevant in earlier Minimalism (Chomsky 1995) and another based on labeling from later Minimalism (Chomsky 2015). Section 6 discusses attractor states, and the conclusion follows in Section 7. 1

Comfort and clarity

As discussed in Chapter 1, the most frequently cited description and explanation of cyclical change is that given in von der Gabelentz (1901), namely, as a balance between ‘Bequemlichkeit’ and ‘Deutlichkeit,’ that is, between comfort and clarity. In this section, I examine where these terms come from – they are not unique to linguistics – and who continues to use them in a linguistic context. One of the bones of contention remains which of the two processes sets cycles in motion. Von der Gabelentz is not the first to use the two opposites, Bequemlichkeit and Deutlichkeit ‘comfort and clarity’ in linguistics, but his is the clearest use of these terms in opposition to each other. When, for instance, Tiedemann (1772: 135) uses the terms in connection to the presence or absence of the dual number, they work in concord for him. Similarly, in the work of Flathe (1803: 60), “[e]inige Pronomina werden theils ihrer Natur nach, theils der Bequemlichkeit, Deutlichkeit und Zierlichkeit wegen anstatt anderer gebraucht” [“some pronouns are used in place of others, partly from their nature, partly for comfort, clarity, and elegance”]. The terms are also frequently used (together) from the early nineteenth century in the fields of pedagogy, science, philosophy, and medicine, according to Google books. For instance, (1) appears in a mathematics text from 1836, but again, the terms do not seem complementary. (1)

wodurch aber die nun bestehende Bequemlichkeit und Deutlichkeit im Nachschlagen sehr verloren haben würde, . . . ‘through which, however, the now existing comfort and clarity in looking [something] up would have been very much lost’ (Adam von Burg, Compendium der höhern Mathematik).

The same is true with the English terms ‘comfort and clarity,’ which are used in medicine, sociology, religion, and the air force, to name but a few areas. Again, their use is not oppositional, as (2) shows, but complementary. (2)

It might very well be that they could show claim at once, and in that case the discourse would proceed with an added comfort and clarity on both sides (Harper’s Weekly 1911, 55.1, p. 32).

174

Explanations and mechanisms

In the remainder of this section, I provide some typical uses of the terms in linguistics and summarize the debate as to which of the two starts the cycle. Langacker (1977) paints a very similar picture in (3) as von der Gabelentz. However, Langacker’s terms for ‘comfort and clarity’ are ‘signal simplicity and perceptual optimality,’ respectively, as shown in (3). (3) This tendency toward perceptual optimality will of course often conflict with that toward signal simplicity. . . . The tension between signal simplicity and perceptual optimality does not manifest itself basically as an ebb and flow in the erosion of established expressions; I have noted that the processes contributing to signal simplicity are largely unidirectional. Instead the central mechanism for achieving perceptual optimality in syntax is a process I will call “periphrastic locution,” which is simply the creation by ordinary or extraordinary means of periphrastic expressions to convey the desired sense. As these new locutions become established in a language, they too gradually fall prey to the processes leading toward signal simplicity, and the cycle begins again. (Langacker 1977: 105, quoted in Haspelmath 1999: 1051, my bolding) Langacker stresses that signal simplicity, that is, comfort, is unidirectional and that perceptual optimality, that is, clarity, may start a new cycle. The type of periphrastic expression that renews is not predictable, of course. Hagège (1993) discusses grammaticalization and the linguistic cycle. His emphasis in the first part of the chapter devoted to cycles is expressive renewal (Hagège 1993: 153), and this suggests that he, like Langacker, would see change as starting with innovation. Öhl (2009) uses ‘minimal effort in computation’ for comfort and ‘maximal explicitness’ for clarity. Other terms that are used for ‘comfort and clarity’ are mentioned in Geurts (2000: 783), namely, ‘efficiency and effectiveness,’ ‘force of unification and force of diversification,’ ‘speaker’s economy and hearer’s economy,’ to name but a few. “[T]his is a terminological free-for-all, apparently” (783), according to Geurts. In this article, Geurts also criticizes Haspelmath (1999) for using the ‘maxims’ of extravagance and conformity because maxims provoke associations with pragmatics, and grammaticalization is not a result of a conscious decision on the part of the speaker. Switching to the controversy as to where the cycle starts, Geurts’ position is given in (4) where he privileges comfort over clarity. (4) A linguistic form may start to erode as long as it serves well enough for most purposes, but once erosion has begun it may cease to be

Explanations and mechanisms

175

sufficiently effective for some purposes, and the need for an alternative form of expression arises. But as soon as an alternative has been found, it will start to compete with the older form, because efficiency doesn’t like grammars in which the same or closely related functions are performed by different devices, and in many cases the older form will duly disappear. (Geurts 2000: 784) This view cannot be completely correct because we have seen that some cycles start for pragmatic reasons and that new and old forms remain in the language side-by-side, for example, the demonstratives and articles in Tohono O’odham and pronouns and agreement markers in Athabascan. Around the turn of the last century, Haspelmath published a number of articles on grammaticalization (Haspelmath 1998, 1999, 2000). One of the questions he is interested in is explaining why this change is unidirectional. By having one principle, either comfort or clarity but not both, “we are more likely to detect the causes of unidirectionality” (Haspelmath 2000: 791) and why it goes on in the first place, as he formulates it in (5). (5)

The real problem is to explain why the conflicting tendencies do not cancel each other out, leading to stasis rather than change – why doesn’t erosion stop at the point where it would threaten intelligibility? Or alternatively, why doesn’t the tug-of-war between the two counteracting forces lead only to a back-and-forth movement? (Haspelmath 1999: 1052, my bolding)

The bolded question in (5) is often asked, and with certain categories, for example, negation, there is never a complete disappearance. However, copulas can disappear completely before they are renewed. As for what sets grammaticalization in motion, Haspelmath, unlike Geurts, sees extravagance as the main reason and explains this as in (6). (6)

Grammaticalization is very similar to these processes: a grammatical construction is initially used for a special communicative effect that gives a short-term advantage to the innovator . . . but as more and more people are trying to get their share of this advantage . . . the advantage disappears, and the system has undergone a change. Thus, the parallel with other inflationary processes again demonstrates the importance of the maxim of extravagance. The extravagance effect is the short-term gain that sets the whole process in motion. Dahl (1999) thus aptly coins the term “rhetorical devaluation.” (Haspelmath 1999: 1061)

176

Explanations and mechanisms

Replying to Geurts, Haspelmath (2000) provides some arguments for that chain of events, that is, having pragmatic strengthening first, the most important arguments of which I will address. The first argument for the primacy of extravagance is that grammaticalization creates novel categories, for example, the werden future in Old High German is “completely new” as are the articles in Romance (Haspelmath 2000: 791) that we’ve discussed in Chapter 3. I have similarly shown this creation of new categories in Chapter 3 for the article in English, the copula in Arabic, and the progressive in various languages. Haspelmath argues that “[t]he creation of novel categories would seem to require the periphrasis-first model” (Haspelmath 2000: 791). A response to that argument would be that the cycle has to start somewhere and can definitely start with a new form, but that doesn’t mean all cycles have to start that way, for example, the Determiner Cycle in English that we’ve encountered in Chapter 3 starts with the grammaticalization of the demonstrative into an article. Haspelmath’s second argument is one I have mentioned in Chapter 1, namely, that grammaticalization can occur without phonological reduction. The response to that argument, given there, is that grammaticalization can occur when it is just semantic bleaching that takes place. This is often the case in languages of East and Mainland Southeast Asia, as shown in Bisang (2008) and Post (2007). The reason for the absence of phonological weakening may be due to these languages being non-stress based, as Schiering (2010) argues. A third argument in favor of extravagance setting the cycle in motion involves multiple grammaticalization. This happens when a language has multiple forms for a future or perfective aspect, for instance. According to Haspelmath, “such cases are quite unexpected in the reduction-first model” (Haspelmath 2000: 792) because only one new form would be needed to repair the old one. Under extravagance, we could have more than one renewal. My answer to this is that, even when the cycle starts with phonological reduction, new forms can be in competition for a while. Concluding this section, I have provided more background on how the terms ‘comfort and clarity’ continue to be used and which of these is seen as starting the cycle. That cycles involve both concepts is uncontroversial, but which starts them remains a topic of debate. This book has provided examples of both starting points. I now turn away from that debate and shift to the question whether it is external or internal pressure that sets cycles in motion. Both language internal and external reasons exist behind language variation and change. Language internal reasons have been argued as responsible for grammaticalization (and the cycle) in Roberts and Roussou (2003) and van Gelderen (2004a, 2011). These authors have suggested that the child, in acquiring its language, makes choices based on economy principles. External reasons include language contact,

Explanations and mechanisms

177

language attitudes, and matters pertaining to language and identity, among other sociolinguistic issues, and that’s the subject of Section 2. 2

External infuence

Claiming that external influence triggers grammaticalization and cyclical change is not a recent development. In Chapter 3, I mentioned Raynouard (1816) in connection with the Determiner Cycle, and he already comments on the possible influence on Romance determiners by Gothic. In this section, I first provide a review of three recent studies that attribute cyclical change and grammaticalization to language contact, Arabic negation, the perfect in Standard Average European, and contact with head and dependent marking languages. I then move to a case study of a twelfth century early Middle English text that, possibly due to Norse contact, started or accelerated various cycles, that is, the Determiner, Perfective, Complementizer, and Aspect Cycles. Lucas (2020) gives a comprehensive overview of the varieties of Arabic in contact with other varieties and languages in specific stages of the Negative Cycle. Although Standard Arabic has a preverbal negative ma, many varieties show a bipartite ma . . . š, as in (7) from Egyptian Arabic, and a few a postverbal -š, as in (8) from Palestinian Arabic. These look like typical stages in the Negative Cycle. (7)

(8)

ma-bəhibb-iš migiyy-u hina ktïr Egyptian Arabic NEG-like.IMPF.1SG-NEG coming-his here much ‘I don’t like his coming here a lot’ (Woidich 1968: 33, from Lucas & Lash 2010: 383). (ana) baḥibb-iš il-fūl Palestinian Arabic I like.impf.1s-neg the-fava beans ‘I don’t like fava beans’ (Lucas 2010: 173).

Lucas’ evidence points to “wide-spread contact-induced change”: “everywhere an Arabic variety with bipartite negation is spoken, there is (or was) a contact language that also has bipartite negation” (2020: 643; 653). For instance, in North Africa, Egyptian Arabic had contact with Coptic, which has bipartite negation as well, as in (9). (9)

en ti-na-tsabo-ou an e-mənte Coptic NEG 1S-FUT-teach-3P NEG on-hell ‘I will not teach them about hell’ (Lucas & Lash 2010: 389).

Lucas and Lash (2010) argue that Coptic learners of Arabic reanalyzed the adverb ši/ šay ‘at all’ as a negative. Another case of external influence on the

178

Explanations and mechanisms

negative cycle in Arabic is that the Berber varieties with bipartite negation are in contact with Arabic varieties with bipartite negation, for example, Moroccan Arabic and thus suggest a role for contact-induced change. In a second recent study, language contact is seen as crucial in the development of the periphrastic perfect in the ‘Charlemagne Sprachbund’ (a term coined by Van der Auwera 1998: 824 for similarities among the languages spoken in the area that was at one point ruled by Charlemagne). It is the core of that is currently the Standard Average Eurpean area (SAE). Drinka (2017) focuses on this Sprachbund and shows in much detail that early Germanic does not have a periphrastic perfect but that it develops this under the influence of Latin, which was the lingua franca in the Charlemagne Sprachbund. Core languages in this Sprachbund are French, German, and Dutch, and these are the languages where the periphrastic perfect has flourished, over the use of the preterite, as (10) shows, and where a have/be distinction, as in (10) and (11), developed and was retained. (10) Ik heb gisteren een boek gelezen Dutch I have yesterday a book read ‘I read a book yesterday’ (interpretation is past rather than perfect). (11) Ik ben gisteren aangekomen I am yesterday arrived ‘I arrived yesterday’ (be is used with unaccusative verbs). Drinka is very categorial in claiming that “the actual trigger to change is contact” (Drinka 2017: 407). Drinka’s focus is grammaticalization, but as we’ve seen in Chapter 3, the grammaticalization of the perfect can result in cyclical change. Heine and Kuteva (2005) examine grammaticalization and cyclical morphological changes, as set in motion by language contact. They call this process ‘grammatical replication.’ Again, their main interest is not in cycles, but their work is certainly compatible with cyclical change. Some case studies they provide involve an extension of clefting in English under the influence of Irish, the increased use of an incipient form of a progressive in Pennsylvania Dutch under the influence of English, and the loss of pro-drop in US varieties of Spanish. Contact-induced grammaticalization is also responsible for new grammatical categories, as Pipil shows (Heine & Kuteva 2005: 84–85). Pipil had no adpositions of its own, but under the influence of Spanish prepositions, Pipil has grammaticalized relational nouns into adpositions. A term sometimes used for this process by Heine and Kuteva is polysemy copying. Heine and Kuteva’s work provides some evidence of cyclical change under contact. For instance, contact with languages that have head marking verbs affected Ritharrngu into adding (enclitic) pronouns in most clauses, as in (12). This results in an increase in head marking and a step in the Pronoun Cycle of Chapter 5.

Explanations and mechanisms

179

(12) wa:ni-na=ŋai ḍaramu ya Ritharrngu go-PST=he man the ‘The man went’ (Heine & Kuteva 2005: 166, from Heath 1978: 127). Evidence for renewal is presented by a Slavic variety, Molisean, that is strengthening its case marked nouns with prepositions under contact with Italian, as in (13). (13) hiža do one žene Molisean House of that woman.GEN ‘The house of that woman’ (Heine & Kuteva 2005: 169, from Breu 1990: 54). As a final example of contact induced cyclical change, we’ll survey the twelfth century early Middle English text we’ve encountered in an earlier chapter and in the Appendix to the previous one, the Peterborough Chronicle, in which, possibly due to Norse contact, various cycles start or accelerate their speed (see van Gelderen 2007b). The stage of the language of the last part of that text is almost as analytic as present-day English. The cyclical changes in that last part involve the Determiner Cycle, with the increased use of demonstratives followed by the introduction of articles, and the Perfective Cycle, with a fast loss of perfective ge- marking and an increase of other aspectual markers. We’ve come upon both cycles in Chapter 3. Another cycle, seen in Chapter 4, is the Complementizer Cycle, in which prepositions reanalyze as complementizers, for example, for, till, and at. Looking at the last part of the Peterborough Chronicle, there is a tremendous increase of complementizers (and sentence embeddings). In particular, the grammaticalization of for accelerates after 1135, a word first found as complementizer in this text, and the use of till as a preposition and complementizer is typical for Scandinavian influence. The town of Peterborough was in an area under the ‘Danelaw.’This area in the east of England was occupied by invaders from present-day Denmark and Norway who spoke Scandinavian languages and who arrived starting in the eighth century. At one point the Danish armies had most of England under control, except the south. This situation causes Domingue (1977), Bailey and Maroldt (1977), Poussa (1982), and Thomason and Kaufman (1988), among others, to argue that such extreme language contact resulted in creolization. More recent, though controversial work on external influence in that area includes Emonds and Faarlund (2014). Allen (1997: 77) is more cautious and thinks the changes start in Old English but may have just been accelerated by the contact, and that is my opinion also. Chronicles were maintained at monasteries to record historical events as they unfolded, but since there had been a fire at Peterborough in 1116 that destroyed its manuscript, the earliest part of the Peterborough Chronicle was copied from

180

Explanations and mechanisms

TABLE 6.1 Changes in the Peterborough Chronicle, especially after 1130

a. b. c. d. e. f. g.

Introduction of the definite article the (cf. Chapter 3) Loss of perfective prefix ge- (cf. Chapter 3) Introduction of telic particles (cf. Chapter 3) Introduction of present perfect have (cf. Chapter 3) Loss of genitive case Introduction of for and till as complementizers (cf. Chapter 4) Relatively analytic stage (cf. Appendix Chapter 5)

the chronicle of another abbey. After 1122, the style becomes unique, indicative of a single scribe entering the events (which certainty is indicated by using ‘anno’ rather than by ‘entry for the year’). This First Continuation continues to 1131, after which a second scribe takes over, in the Second or Final Continuation. In Table 6.1, I summarize the phenomena that are telling of fast cyclical change in the Peterborough Chronicle. Leiss (2000) argues for a close connection between definiteness, aspect, and case. Languages such as Russian can ‘do without’ articles because they distinguish perfective and imperfective aspect and the object of a perfective verb is definite by default. That insight connects changes (a)–(d) in Table 6.1: the loss of the perfective aspect necessitates the introduction of the article and the use of telic particles and the auxiliary have. How does the loss of genitive case in (e) play into this? First, the loss of case and the introduction of the article are linked since articles show no case, unlike demonstratives (at least in Germanic). Second, Finnish and other languages show that a special case on the object marks aspect. Throughout the history of language, Leiss (2000: 185) says “[d]ie Artikelexplosion findet . . . im Genitiv statt” [‘the articles start in the genitive’] because the genitive is the case that is no longer marked aspectually. The change in (f) is independent of the five others and indicates an accelerated grammaticalization in sentence connections, possibly typical of language contact, as Heine and Kuteva (2006) have argued. The change in (g) affirms what the introduction of articles suggest, namely, an increase in the analytic character of the language. I’ll now say something more on each of the seven changes. As for articles and demonstratives, Pysz (2007) compares expected Old English demonstrative endings with attested ones. In the pre-1121 Peterborough Chronicle, there are few unexpected forms. For instance, the masculine nominative se is used for masculine nominatives 275 times and only 2 times for a dative, and the innovative þe is used only 10 times. In the 1122–1131 section, se is used as masculine nominative 103 times but 14 times as dative or accusative, and þe/ðe is used 7 times. In the last section, that is, 1132–1154, se is used as masculine nominative only once. The form te is used 19 times, the once, and þe

Explanations and mechanisms

181

55 times. Pysz (2007: 73) provides percentages for when all Cases are indicated ‘correctly,’ that is, archaically: 85% in the pre-1121 period, 46% between 1122 and 1131, and 13% in the last period. This shows the case on demonstratives and the use of articles by the second scribe is definitely innovative. As for aspectual changes, the ge- prefix appears in the Peterborough Chronicle only three times after 1130 (and not with a perfect but with a passive). One sign of the weakening of perfective ge- is the change to a-, as in (14) to (16), and its strengthening by telic adverbs, such as up in (14) to (17). (14) til he aiauen up here castles ‘till they gave up their castles’ (Peterborough Chronicle anno 1140, 52). (15) 7 ælc unriht for gode and for worulde up aras ‘and every wrong in the sight of God and of the world rose up’ (Peterborough Chronicle entry for the year 1100). (16) asprang up to þan swiðe sæ flod ‘sprang up to such height (the) sea flood’ (Peterborough Chronicle entry for the year 1099). (17) Sum he iaf up ‘Some (castles) he gave up’ (Peterborough Chronicle anno 1140). As for the loss of genitive case, Allen (2007: 86) writes that even in the pre-1121 Chronicle “we find a trend towards replacement of the genitive objects” and that in the additions “no certain examples of genitive objects” occur. Allen (1995: 177) provides a few examples from the First Continuation, that is, after 1121, of genitive loss. I repeat two as (18) and (19), where the accusative object would have been genitive in Old English. (18) benam ælc ðone riht hand took every that.ACC right hand ‘deprived each of their right hands’ (Peterborough Chronicle anno 1125, line 9). (19) him me hit beræfode him man it.ACC bereaved ‘He was deprived of it’ (Peterborough Chronicle anno 1124, line 51). The genitive case of the object disappears in English in the twelfth century, at the same time as aspectual prefixes, and when specificity/definiteness markings increase. As for (f) in Table 6.1, for is used as a preposition of causation, as in (20), in the Peterborough Chronicle. The preposition followed by a demonstrative, that is, forðæm also functions as a complementizer in the meaning of ‘because,’ as in (21).

182

Explanations and mechanisms

(20) ouþer for untrumnisse ouþer for lauerdes neode ouþer for haueleste ouþer forhwilces cinnes oþer neod he ne muge þær cumon ‘either from infirmity or from his lord’s need or from lack of means or from need of any other kind he cannot go there’ (Peterborough Chronicle entry for the year 675). (21) forþam Trumbriht wæs adon of þam biscopdome ‘because T had been deprived of his biscopric’ (Peterborough Chronicle entry for the year 685). The earliest instance of for as a finite complementizer in English is also in the Peterborough Chronicle, if the OED is correct, and is from the entry for 1135, as in (22). There are two others from the entry for 1135, as in (23) and (24). (22) for þæt ilc gær warth þe king ded ‘because (in) that same year was the king dead’ (Peterborough Chronicle, anno 1135, line 6). (23) for æuric man sone ræuede oþer þe mihte because every man soon robbed another that could ‘becasue everyone that could robbed someone else’ (Peterborough Chronicle, anno 1135, line 8). (24) for agenes him risen sona þa rice men ‘because against him soon rose the powerful men’ (Peterborough Chronicle, anno1135, line 18). This locates the first use of complementizer for with Scribe II, that is, during the Final Continuation. Between 1135 and the end of the Chronicle in 1154, the use increases dramatically compared to the period before 1135, as (25) to (31) show for the year 1137. (25) for he hadded get his tresor ‘because he had got his treasure’ (Peterborough Chronicle, anno 1137, line 3). (26) for æuric rice man his castles makede ‘because every powerful man made his castles’ (Peterborough Chronicle, anno 1137, lines 13–14). (27) for ne uuæren næure nan martyrs swa pined alse hi wæron ‘because never were martyrs as tortured as they were’ (Peterborough Chronicle, anno 1137, line 20). (28) for nan ne wæs o þe land ‘because none was in that land’ (Peterborough Chronicle, anno 1137, line 42). (29) for ouer siþon ne forbaren hi nouther circe ne . . . ‘because nowhere did they forbear a church nor . . . ’ (Peterborough Chronicle, anno 1137, line 46).

Explanations and mechanisms

183

(30) for hi uueron al forcursæd ‘because they were all accursed’ (Peterborough Chronicle, anno 1137, line 53). (31) for þe land was al fordon mid suilce dædes ‘because the land was all fordone by such deeds’ (Peterborough Chronicle, anno 1137, lines 54–55). Another innovation is the complementizer till. Mustanoja (1960: 408–409) states that the use of till is “typically northern.” Two instances of til occur in the Chronicle, as in (32) and (33). According to the OED, the use of till is the earliest in the Peterborough Chronicle. Again, it is in the recordings of Scribe II. Interesting is that till occurs only as a complementizer, not as a preposition. This may mean that use was borrowed rather than grammaticalized out of an existing preposition. (32) til hi iafen up here castles ‘until they gave up their castles’ (Peterborough Chronicle anno 1137). (33) til hi aiauen up here castles ‘until they gave up their castles’ (Peterborough Chronicle anno 1140). The suddenly increasing speed of article use, aspectual and prepositional changes in the Final Continuation of the Peterborough Chronicle may show that the text is affected by language contact. Demonstratives are replaced by articles, aspect marking on verbs is being lost and replaced by new adverbs, and prepositions become complementizers. This is all an indication of increasing analyticity and decreasing syntheticity, as we’ve also measured at the end of the last chapter. Concluding this section, it is very likely that grammaticalization and renewal can come about through language contact and social pressure. Examples that have been reviewed involve the following: negatives in Arabic, the perfect in the Charlemagne Sprachbund, the work on grammatical replication, and the fast changes in English in the twelfth century. In the remainder of this chapter, language internal reasons for these changes will be discussed. 3

Construction grammar

As mentioned in Chapter 1, Construction Grammar has always been interested in language change, for example, Barðdal (1999), Traugott (2008), Bergs and Diewald (2008), Traugott and Trousdale (2013), Barðdal and Gildea (2015), and Sommerer et al. (2015). This framework provides a network of form-meaning pairs that are called ‘constructions,’ and this knowledge arises in speakers based on usage, not from innate factors, as in the generative framework. In this section, I first provide some background into what Construction Grammar is good at, using Hilpert’s recent work, and then provide three approaches to how this

184

Explanations and mechanisms

framework handles grammaticalization. Finally, I speculate on how it might study cycles, because I haven’t been able to find much in the literature. The pairing of form and meaning in Construction Grammar is idiosyncratic and unpredictable. Hilpert (2021: 38–41) explains how Construction Grammar capitalizes on the notion of the unruliness and unpredictability of change and that renders it very different from, for instance, grammaticalization theory, which makes testable predictions. If change in this model is unpredictable, are there advantages to using Construction Grammar in historical linguistics? Hilpert answers affirmatively and states that it deals “with many changes that are in fact bidirectional” (2021: 43) so that a noun can be used as a verb and a verb as a noun. It is therefore comfortable with degrammaticalization, such as the use of ish as an independent word, derived from an affix. Construction Grammar is interested in how both the form and function of a construction change. Since it is usage-based, Construction Grammar makes copious use of corpora and examines the competition between constructions, for example, the verb dislike with a to-infinitival complement versus one with an -ing complement. Hilpert uses the COHA data in Figure 6.1 to argue that the “connection between dislike and the to-infinitive has all but disappeared from the grammar. An existing link has become much weaker, and you can see in the smooth decline of frequencies that this must have happened gradually” (2021: 59).

FIGURE 6.1

Changing complements to dislike

(from Hilpert 2021: 58, but with charts directly from COHA)

Explanations and mechanisms

185

Barðdal and Gildea (2015), as mentioned in Chapter 2, also emphasize that competitive aspect of this framework. Traugott (2008) emphasizes that grammaticalization involves the interaction of phonology, morphology, syntax, and pragmatics and that all these aspects appear in a Construction Grammar representation. “[S]emantics, morphosyntax, and phonology, and, in some models, pragmatics work together in a construction” (2008: 223) and it is the entire construction that changes in grammaticalization. The example she provides is the [NP1 of NP2] construction that changes from a construction where NP1 is the head, as in (34), to where NP2 is, as in (35), with accompanying changes in meaning. (34) He’s a sort of a prentice ‘He is a kind/type of apprentice’ (1632, Lithgow Travels, Traugott 2008: 228, my gloss). (35) I do think him but a sort of a, kind of a . . . sort of a Gentleman ‘I think of him only as a certain degree of gentleman’ (1720, Shadwell, Traugott 2008: 229, my gloss). These constructions involve “macro-constructions: meaning-form pairings that are defined by structure and function, e.g. Partitive, or Degree Modifier Constructions, etc.” (2008: 236), which bind the constructions of sort of, lot of, and shred of together, even though their semantics are slightly different. Sommerer (2015) argues that the Old English “demonstrative se developed into the definite article due to the emergence of an abstract, syntactic, and lexically underspecified macro-construction with a determination slot for marking definiteness in early Old English” (2015: 107). She shows that Old English texts already showed a high frequency of demonstratives and possessives in definite Old English NP constructions and that the learner abstracted the macro-construction, in (36). (36) Semantics Definite Reference Form Expressed in grammar by an NP construction with a determination slot (Sommerer 2015: 130) For Sommerer, the “driving force of lexically underspecified constructions” (2015: 107) is crucial to grammaticalization. The determination slot in (36) “becomes a functionally exploitable structural category itself, which leads to the recruitment of the demonstrative as a default slot filler” (2015: 107). Culicover (2021) presents a constructional approach to language change and variation, with a focus that differs from Hilpert’s. Culicover takes a construction to be “a correspondence between elements of four ‘tiers,’ PHON, SYN, GF (for

186

Explanations and mechanisms

‘grammatical functions’), and CS (for ‘conceptual structure’)” (2021: 19), where PHON is responsible for overt form and SYN for the hierarchical organization. Constructions are of the form in (37), for the double object construction, where each tier is connected to the others via an index, here ‘4.’ (37) Constructions: Double Object PHON [1–2–3]4 SYN [V1, NP2, NP3]4 VP GF [GF > GF2]4 CS [λz, λy, λx. transfer’ (SOURCE:X,GOAL:Y3,THEME:Z4, MEANS:1’)(2’)(3’)]4 (adapted from Culicover 2021: 25) The way the GF tier is formulated ensures that the indirect object is singled out as the relevant GF (Culicover 2021: 25). This means that only the indirect object is affected in a passive. Culicover argues that “there is no need to insist that language change occurs solely as the consequence of learner errors [sic!]” and “that change may occur spontaneously in speakers of any age, even when there is no explicit evidence for it” (2021: 75). A major driving force of change is economy, which reacts to three types of complexity, namely, representational complexity, computational complexity, and interpretive complexity (2021: 87). Representational complexity can be found in “constructions that are adjacent in PHON” but “not strictly local in SYN” (88), as in the case of the movement of a subject wh-element to the left periphery. Culicover’s account of this phenomenon is similar to generative accounts that prohibit such movement (e.g. Goto & Ishi 2019), and it does not seem particularly couched in a Construction Grammar explanation. A concrete example of representational complexity that relies on a Construction Grammar representation is generalization in the SYN CS interface, as demonstrated by “demonstratives [which] regularly undergo the diachronic sequence” known as Greenberg’s Cycle. The proximal demonstrative this has a representation, as in (38). (38) PHON SYN

this CATEGORY DEM LID this CATEGORY thing CS REFERENTIALITY referential DEFINITENESS definite LOCATION proximate (Culicover 2021: 90)

Explanations and mechanisms

187

We’ve encountered this “diachronic sequence” in Chapter 3. Culicover’s account of this change is “the loss of each of the conditions in turn, from LOCATION, to DEFINITENESS, to REFERENTIALITY.” The first loss resulting in a definite article, the second in a generic article, and the third in a nominal marker. Culicover is not concerned with the renewal of these features and perhaps uses “sequence” to avoid the term ‘cycle.’ I am sure there could be cognitive renewal in this schema, as in the case of Sommerer’s appearance of a determination slot in the history of Old English. The gradual loss of the features in (38) is key to its grammaticalization; it seems less the result of the status of the demonstrative this as a construction. The construction itself doesn’t provide a reason why one feature would disappear before another. Culicover also invokes computational complexity in cases of extraction from islands and interpretive complexity in cases where a grammatical form that has more than one meaning. The latter occurs when the simple present expresses both present tense and progressive aspect. Neither of these complexity avoidance strategies crucially rely on Construction Grammar; they can also be formulated using other frameworks. Concluding, Construction Grammar has from its start been interested in language change. Hilpert’s work stresses the usage-based nature of the framework by studying competition among constructions. He notes that Construction Grammar explains bidirectional change, which is not particularly suited to grammaticalization. In contrast to this, Traugott (2008) sees Construction Grammar as eminently suited because of its linking of the different linguistic levels. However, some years later, Traugott (2014: 87) writes that “unidirectionality has a less prominent theoretical status than is often assigned to it in non-constructionalist models of language change.” Sommerer (2015) and Culicover (2021) present an analysis of language in change, in particular the grammaticalization of the article. 4

Early Minimalism: structural and featural economy

Generative grammar has gone through many instantiations in its over 60 year history. As I mentioned in Chapter 1, only relatively recently have these models been employed to account for grammaticalization and linguistic cycles. The Minimalist Program, as in Chomsky (1995), makes explanations of these processes possible. In this section, I focus on two accounts given for grammaticalization and cyclical change within the early minimalist phase of generative grammar, namely, structural economy and featural economy, as developed in van Gelderen (2004a, 2011) and Roberts & Roussou (2003). Structurally, there are three main simplifications a child learning a language can make: phrase to head, reanalysis in a higher structural position, and incorporation of adjuncts as arguments. I start with the first of these.

188

Explanations and mechanisms

Jackendoff (1977) develops X-bar structure, where syntactic structures is seen as consisting of a head that projects upwards and results in derivations that follow X-bar principles, where the X in (39a) is replaceable by any category, as in (39b). The head is occupied by one word or morpheme, and the specifier and complement can contain full phrases, as shown in (39). (39) a.

XP

Specifier

b. X’

X

VP

DP They

Complement

V’ V read

NP books

As they grammaticalize, languages have been observed to change towards less structural complexity. For instance, van Gelderen (2004a, b) puts this as the head preference principle (40), in which a structure such as (41b), with just a negative head, is simpler than that in (41a), with a specifier and a head in the Neg(ative) P(hrase). Change is therefore from (41a) to (41b). (40) Head Preference Principle (HPP) Be a head (X) rather than a phrase (XP). (41) Indo-European (Sanskrit) Old English a. NegP > b. NegP AP ná

Neg’ Neg

VP

Neg ne

VP

Evidence for this change can be found in Indo European where the negative is flexible and can be analyzed as an adverb, situated in the specifier of the NegP. By Old English, in its function as sentential negative, it always precedes the verb, as in (42a), and is often contracted with it, as in (42b). (42) a. ne hyrde ic cymlicor ceol gegyrwan. not heard I more.beautiful ship prepare ‘I never heard before of a ship so well prepared’ (c1000, Beowulf 38–39, repeated from Chapter 1). b. N=is þæt heoru stow. NEG=is that pleasant place ‘That is not a happy place’ (c1000, Beowulf 1372). With the reanalysis of the phrase as a head in (41b), the verbal head cliticizes to it as it moves to the left, weakening it phonologically. At some point, the learner

Explanations and mechanisms

189

no longer perceives the negative features as salient and uses another negative in the sentence, as in (43). (43) ic ne=geseah nowiht buton eorðan. I NEG=saw nothing except earth ‘I saw nothing except earth’ (DOE, c1000, Lives of Saints, 310–311, repeated from Chapter 1). The next change is for the ne to disappear and for nowiht to shorten to not. The cycle then starts over again and the change to (44a) and is again one from phrase to head. Currently, the negative is again phonologically weak, and new negatives are used, as in (44b). (44) a. And to þis I cannot answere þee bot þus: ‘I wote neuer.’ ‘And to this I can’t answer you except thusly: I knew never’ (c1400, Cloud of Unknowing, 450–451, repeated from Chapter 1). b. ‘But I didn’t see no one there Who looked half as good as you’ (Lyrics, www.lyrics.com/lyric/3122769/Mary+Chapin+Carpenter/ How+Do). The same change as from (41a) to (41b) can be seen in the Pronoun and Determiner Cycles, as shown in (45) and (46), respectively, but also in many of the other cycles, for example, Copula and Aspect Cyles. (45) French a. TP DP Je (46) a.

b. T’

T j’

T DP

b. D’

that D

TP

NP

D the

VP DP NP

Grammaticalization not only involves the change of a phrase to a head but also the reanalysis of a head as a higher head. This can be accounted for in terms of structural economy as well, as in upwards reanalysis in Roberts and Roussou (2003) or Late Merge in van Gelderen (2004a), as in (47).

190

Explanations and mechanisms

(47) Late Merge Principle (LMP) Merge as late as possible: If words or phrase are ambiguous as to where they originate, assume the higher position and avoid movement. This mechanism involves the reanalysis of verbs as modals or aspectual auxiliaries, the reanalysis of negatives as question markers, and the reanalysis of PPs and adverbs as complementizers and discourse markers, respectively. For instance, if an adverbial is frequently preposed in a question because it links the discourse, as in (48a), it can be reanalyzed there, as in (48b). (48)

a.

PP After that

CP

> TP

...

VP ...

b.

CP

>

PP TP After (that) . . .

c. C after

CP TP ...

PP

The next step would be for the PP to reanalyze as P, which is a phrase to head reanalysis, as in (48c). Earlier in this section, I showed how negative phrases reanalyze as heads. It can also be shown that they start out as lower phrases and move to higher ones. Breitbarth et al. (2020: 45–49), in discussing the connection between the start of the Jespersen Cycle and information structure, note that narrow focus, as in (49b), historically precedes neutral, broad focus, as in (49a), in negative clauses. (49) a.

Edwin can’t/cannot climb trees, can he? NEG > MOD ‘Edwin is not permitted/able to climb trees.’ b. Edwin can NOT climb trees, can he? MOD > NEG ‘Edwin is permitted/able not to climb trees’ (after Cormack & Smith 2002: 137).

That means that the negative in (49a) is higher in the hierarchical structure than (49b) and has been reanalyzed upwards. In related work, Magistro et al. (2022) study the interaction between syntax, phonology, and information structure in varieties of Italian in various stages of the Jespersen Cycle. In the more advanced stages, where the emphatic negator is becoming a neutral negative, the new negator loses pitch and duration, as in (49a). According to Magistro et al., prosodic independence is key to a new negator, which can be lost when it becomes a new negator. The Case Cycles in Chapter 5 involve a verb or noun being reanalyzed as a case marker, and I present an analysis here for the change involving ba in (50),

Explanations and mechanisms

191

a sentence repeated from Chapter 5. Various structures have been argued for; I assume ba is an ASP(ect) head, as in (51), since affectedness is important. In this structure, ba also moves to v, and the object moves to the specifier of the ASPP (see Sun 2008), which accounts for ba and the nominal not forming a constituent. (50) wo ba shu mai le Chinese I PRT book buy PF ‘I bought the book’ (Li & Thompson 1981: 21). vP (51) v’

wo v ba

ASPP DP shu

ASP’ ASP

VP DP

V’ V mai

How did this structure arise? Looking at the origin, as in (52), if the verb is in a serial verb construction, ba can be reanalyzed from a full verb to an aspect marker in (50). This is, of course, compatible with structurally higher reanalysis. (52) . . . yin ba jian kan should hold sword see ‘I should take the sword and see it’ (Tang dynasty poem, Li & Thompson 1974: 202–203). An example of a noun reanalyzing as inherent case marker involves the French preposition chez ‘with.’ It is related to the Latin noun casa (cf. Vincent 1999; Lightfoot 1999). Longobardi (2001) examines locatives as in Italian (53) and suggests the reanalysis in (54). The noun first moves to D and then to P and is reanalyzed in a higher position. (53) va-go casa (mia) Veneto Italian go-1S home my ‘I am going home’ (Longobardi 2001: 289).

192

Explanations and mechanisms

(54) a.

PP (Latin)

P

b.

DP D

> NP casa

chez ‘with’

P (French) DP moi ‘me’

Longobardi (2001) argues that casa undergoes movement from N to D in Latin and early French (chies at this point) and that, once it is in D, it can be incorporated as a P. As we’ve seen in Chapter 5, case markers often derive from nouns via prepositions. The French chez could in principle become used as (comitative) case marker. The third structural factor, that is, incorporating adjuncts as arguments, addresses renewal. Renewal in the Pronoun Cycle, for example, in French, is through incorporation of topicalized, that is, adjunct, phrases, as in (55), for the ambiguous sentence in (56). The same is true for negatives: before the negative is a grammatical negative, it is an adjunct. Su

→

→

(55) TOP

V

SU Agr-V (56) Moi je dit toujours que . . . Me 1S say always that . . . ‘I always say that . . . ’

Spoken French

Later in this section, I come back to this renewal. Within early Minimalism, another possible explanation concentrates on features as the locus for language acquisition and change. If languages differ only in some of the grammatical features they use, we can see the linguistic cycle as a loss of features, followed by a renewal, a conclusion that is not surprising considering traditional work in grammaticalization. First, I’ll provide a little more background on features and then discuss how these can be used in cycles. Many types of features are introduced in Chomsky (1995) and shown in Table 6.2. The semantic ones are probably drawn for a pre-linguistic inventory and include cause, agent, event, state, thing, count/mass, and substance/ aggregate. The formal ones are relevant to feature checking in the syntax and are divided into intrinsic or optional. The intrinsic ones are inherently connected to the lexical item and include categorial features, the case assigning features of the verb, and the person and gender features of the noun. Optional features are predictable from linguistic principles, for example, nouns need case or some kind of licensing, and verbs need tense. These features include

Explanations and mechanisms

193

TABLE 6.2 Features of house and build

semantic: formal: intrinsic [nominal] [3 person] [non-human]

house e.g. [artifact]

build e.g. [action]

optional [number] [Case]

intrinsic [verbal] [assign accusative]

optional [phi] [tense]

(adapted from Chomsky 1995: 231)

the tense and agreement features of verbs and the number and case features of nouns. The person, number, and gender features are usually referred to as phi-features. Formal features can be interpretable or uninterpretable. The case features on nouns and the agreement features on verbs are uninterpretable (in English) because they are not relevant for the interpretation. In English, (57) is perfectly understandable but grammatically incorrect. That means the incorrect accusative case of me is not relevant to the interpretation, nor is the incorrect agreement on the verb. (57) Me sees him. The interpretable features are the tense of the sentence, the person and number features on nouns, and the case assigning features of the verb (e.g. a verb such as build can check/value the accusative of its object but arrive cannot). These are relevant for the interpretation of a sentence: the [i-3S] in (58a) stands for interpretable third person and singular number features on the noun and the [u-phi] for uninterpretable and, as yet unspecified, person and number features on the verb. The latter are crossed out in (58b) after having been checked because they are not relevant for the interpretation. In (58), I just show the phi-features, which are person and number in English but could be different in other languages. (58)

a. before checking b. after checking

Andy [i-3S] [i-3S]

reads [u-phi] [u-phi: 3S]

books [i-3P] [i-3P]

An uninterpretable feature is called a probe, and it needs an interpretable feature, that is, a goal. The semantic component doesn’t distinguish between an interpretable and semantic feature, and the latter may be the goal for uninterpretable features as well.

194

Explanations and mechanisms

After having gone over some features, I now return to two cyclical changes, negatives and pronouns, to show there is a change from semantic to interpretable to uninterpretable features. Other cycles can similarly be explained (see e.g. van Gelderen 2011). As for the Negative Cycle, there is a renewal of the Old English negative ne by means of nan wuht ‘no creature.’ At the early stages of the renewal, the renewing element can be any negative argument that reinforces the negative ne, for example, nan Þing ‘no thing,’ nan wuht, na(n)wuht, nan scild ‘no shield,’ and na mon ‘no man.’ There is not one designated form. The semantically negative features of these arguments aid in the renewal of the weakened negative. The next step is for these semantic features to be reanalyzed as interpretable negative ones, as has happened in Early Middle English, where nawiht is no longer an argument but an adverbial. The final step in the cycle is for the nawiht/noht to be reanalyzed as weak, that is, uninterpretable, and for renewal to start again. In tree form, using a Neg(ation)P(hrase), the changes are represented in (59). In (59a), the negative features are analyzed by the language learner as interpretable. Generations later, they are (re)analyzed as uninterpretable, as in (59b), because there is an argument that is negative in semantic features. The uninterpretable negative feature acts as a probe looking for a goal. In (59c), the nowuht becomes the designated negative with interpretable features, which it loses in (59d) and so on. (59) a.

NegP

ne/no [i-neg]

c.

> Neg’ • VP

NegP

nowuht [i-neg]

> Neg’

ne [u-neg] >

Neg’ • VP

b. NegP

d. NegP

VP • nænig mon [neg]

Neg’ -n’t VP [u-neg] • nothing [neg]

Cyclical changes in negatives can be explained by arguing that their (initially) semantic features are reanalyzed as interpretable and then as uninterpretable. So the pragmatically renewed negative has semantic features that end up being

Explanations and mechanisms

195

used grammatically, initially as a full phrase in the specifier position, as in (59a), and then in the head, as in (59b). Once the phrase is reanalyzed as a head (e.g. Old English ne ‘not’), another element is required. Van Gelderen (2008a, 2011) formulates these changes as an economy principle, as in (60). (60) Feature Economy Minimize the semantic and interpretable features in the derivation: DP in the VP Specifier of NegP Head Neg negative affix semantic negative > [iNEG] > [uNEG] > [uNEG] This means the child in acquiring their words will initially connect these with semantic features but later extrapolate the grammatical ones. In Figure 6.2, the changes can be represented as follows: semantic features are reanalyzed in the specifier position as interpretable and in the head as uninterpretable features. The change represented here occurs in the case of the Negative Cycle as well as the Subject Pronoun to Agreement Cycle: the interpretable person (and gender) features of a full pronoun are reanalyzed as uninterpretable when they become agreement, as represented in (61). (61) Subject Agreement Cycle emphatic pronouns > full pronoun > head pronoun > [phi] [i-phi] [u-1] [i-2/3]

agreement [u-phi]

Let’s take the Old French we’ve seen in the previous chapter, and repeated as (62). In (62), the emphatic pronoun is je; it is optional and need not immediately

FIGURE 6.2

The negative cycle

196

Explanations and mechanisms

precede the verb di. It has first person singular features, but these are semantic features, not necessarily involved in checking. If Old French is pro-drop (cf. Adams 1987; Vance 1997), it is the null subject that provides the interpretable features for the verb’s uninterpretable ones to check with. (62) Se je meïsme ne li If 1S myself NEG 3SM ‘If I don’t tell him myself.’

di tell

Old French

In the case of English and French pronouns, it is not so obvious that they have semantic person features. Southeast Asian languages show this better. Indonesian saya ‘I’ originates from ‘servant, slave’ but is now the regular first person. Thai is reported to have over 20 first and second person markers, all derived from nouns. In most languages, pro-drop is lost earlier for first person and later for second and third (van Gelderen 2000) and, hence, my differentiation between first and second/third person in (61). Once pro-drop is lost, the uninterpretable phifeatures on the verb need new semantic or interpretable features, which they find in the emphatic je pronoun of (62). This stage is then reached for all persons, and je ‘I,’ tu ‘you,’ and il/elle ‘he/she’ and others are reanalyzed as valuing the phifeatures of the verb. Currently, the last stage of the cycle is taking place because the first and second person pronouns are being reanalyzed as agreement on the verb, that is, as having uninterpretable phi-features in need of another semantic or interpretable phi-marker. This is the reason (63) is frequent. (63) Moi, je dit toujours que . . . me, 1S say always that . . . ‘I always say that . . . ’

Spoken French

Now we have come full cycle where a pronoun (je in (63)) becomes agreement and is replaced by a new pronoun (moi in (63)). Even though there is a comma between the new subject and the verb, the subject in (63) is a real subject and no longer a topic in many varieties of French because quantifiers and indefinites, as in (64), can appear. These represent new information and cannot be topics. (64) Un homme il a sorti a person 3S has l eft ‘Someone left’ (Nadasdi 1995: 82).

Ontario French

In conclusion to this section, I have translated the traditional ideas of comfort and clarity, and phonetic and pragmatic weakening, into minimalist structural

Explanations and mechanisms

197

and featural economy to account for the grammaticalization side of the various cycles. Adjunct incorporation and feature renewal ensure that grammatical information is not lost and therefore account for the renewal side of the cycle. The current Minimalist Program (e.g. Chomsky 2015; Chomsky et al. 2019) provides another way of looking at the changes, and I turn to that next. 5

Later Minimalism: minimal search

As we saw in the previous section, a possible explanation for the cyclical changes, where phrases reanalyze as heads and heads are lost, is that, during acquisition, principles of economy predispose the learner to use simpler structures and features. The urge of speakers to be innovative may introduce new, loosely adjoined elements into the structure. In this section, I show that it is also possible to see these changes as solutions to labeling problems: a phrase and another phrase are harder to label than a head and a phrase, which is behind the reanalysis. As I argue in van Gelderen (2022), providing less ambiguity of labeling is an alternative more in line with current thinking on attributing less to Universal Grammar and more to more general factors. Labeling a tree node as NP or VP used to be an automatic part of the syntax. Early phrase structure rules start with a label (VP → V NP), and later models have a V and NP project a VP. Chomsky’s current focus (2013, 2015) on labeling as a third factor effect necessitated by the interfaces and unconnected to merge makes it possible to see syntactic change in another way, namely, as a resolution to labeling problems. The labeling algorithm (LA) operates after parts of the sentence have been formed, at the time that a syntactic derivation is transferred to the interfaces. When a head and a phrase merge, as in (65), the LA determines that the D head the determines the label. (65) Merge the and red apple:

{the, {red, apple}}

Where two phrases merge, the LA cannot find the head, and one of the phrases either has to move or share features with the other. In certain cases, the heads of the phrases share phi-features (or Q-features), and the set is successfully labeled, in the case of a subject in (66). (66) DP Tom [i-3S]

TP T VP [u-phi: 3S] . . .

198

Explanations and mechanisms

The interpretable features of the DP ([i-]) in (66) value the uninterpretable ones on the T ([u-]), but that is not crucial here. In addition to Chomsky’s resolutions (movement or sharing), van Gelderen (2019) argues that reanalyzing a phrase as a head also resolves the labeling problem. In fact, the labeling problem doesn’t even occur, and many of the cycles involve phrase to head reanalysis. The reanalyzes of phrases as heads are varied: (a) subject and object pronouns to agreement, that is, DPs to T and v heads, respectively; (b) demonstratives to C, D, and T heads; (c) wh-elements to C heads; and (d) adverb phrases to ASP heads, PPs to C heads, and negative adverbs to Neg heads. All these changes provide insight into some of the labeling mechanisms: minimal search into the phrase of (67a) renders a TP in (67b). The frequent reanalysis of (66) as (67) shows that minimal search is preferred over the sharing of features. (67) a. T

?

b. VP

T

TP VP

An example of a phrase being reanalyzed as a head occurs when an independent pronoun changes to an agreement head. (After the pronoun is reanalyzed as agreement, there is optional renewal in many languages.) As known from Chapter 5, in Old French (68), the pronoun je need not be adjacent to the finite verb but, in Modern Spoken French (69ab), it does, indicating it is part of the verb. (68) Se je meïsme ne li di Old French If I myself not him tell ‘If I don’t tell him myself’ (Franzén 1939: 20, Cligès 993). (69) a. b.

Je lis et j’-écris 1S read and 1S-write *Je lis et écris 1S read and write ‘I read and write.’

Spoken French

This and a number of other differences show that the erstwhile pronouns, such as those in (68), have been reanalyzed as agreement markers in (69a) and are now obligatory. A scenario for the change from (68) to (69a) due to labeling pressure is that the labeling in (68) of the phrasal pronoun and the clause is not as straightforward, shown by the question mark in (70a), as the labeling through a head in (70b). In (70a), the DP and TP are both phrases, so no head can be detected by

Explanations and mechanisms

199

the LA. That’s why they have to share features. In (70b), labeling is through minimal search for a head, which is simpler. (70) a. DP je phi

?

> TP

T phi

vP

b. TP T je

vP

The change discussed in this section, from DP pronoun to T agreement, shows that, even though both {XP, YP} and {X, YP} can be labeled, the latter is preferred because that labeling is just through minimal search. As I show in van Gelderen (2020), the sharing of features in the CP does not result in a reanalysis because those features are semantic, unlike the morphological ones in (66). Section 5 has discussed some recent developments in generative grammar that account for grammaticalization. I have not discussed renewal in this framework because that account is similar to the one in Section 4, namely, through adjunct incorporation (see also van Gelderen 2022). The difference between the approaches in Sections 4 and 5 is due to a shift of emphasis, away from language-specific rules to more general principles. However, the specifier to head reanalysis of Section 4 is readily translated into labeling determinacy in Section 5 and Late Merge can actually be translated as a workspace determinacy. Late Merge in (47) avoids movement by reanalyzing a word or phrase in a higher position if there is evidence in the data; workspace determinacy serves a similar function (see van Gelderen 2022). Table 6.3 sums up the start to the changes discussed in this book with an explanation in terms of specifier to head reanalysis (HPP, i.e. labeling determinacy) or

TABLE 6.3 What sets the cycles in motion, using generative grammar

Start of cycle Determiner Cycles Copula Cycles Imperfective Cycles Perfective Cycles Mood Cycles Voice Cycles Negative Cycles

Spec to Head = labeling determinacy Spec to Head = labeling determinacy and Late Merge = workspace determinacy Late Merge = workspace determinacy Spec to Head = labeling determinacy and Late Merge = workspace determinacy Late Merge = workspace determinacy and Spec to Head = labeling determinacy Spec to Head = labeling determinacy and Late Merge = workspace determinacy Late Merge = workspace determinacy and Spec to Head = labeling determinacy

200

Explanations and mechanisms

TABLE 6.3 (Continued)

Start of cycle Interrogative Cycles

Late Merge = workspace determinacy and Spec to Head = labeling determinacy Complementizer Cycles Late Merge = workspace determinacy and Spec to Head = labeling determinacy Pragmatic Cycles Late Merge = workspace determinacy and Spec to Head = labeling determinacy Subject Agreement Cycle Spec to Head = labeling determinacy Object Agreement Cycle Spec to Head = labeling determinacy Case Cycles Late Merge = workspace determinacy

in terms of Late Merge (i.e. workspace determinacy). Determiner and Agreement Cycles have only been shown to start from full phrases, but there is in principle nothing that prevents a start using heads. In Imperfective and Case Cycles, the start has just been from lexical to grammatical heads, for example, a full verb reanalyzing as epistemic modal. Again, in principle, a phrase could start these cycles too, for example, a temporal adverb could start to be used as imperfective auxiliary. As mentioned, the linguistic cycle also involves renewal and one of the main mechanisms accounting for it is adjunct incorporation. 6

Attractor states and linguistic cycles

Dynamical systems are an area of mathematics, and in it, “an attractor is a set of states toward which a system tends to evolve” (wiki). This term has also been used in typological and historical linguistics, for example, in Nichols (2018: 1), to “include sounds, forms, and grammatical structures that are prone to be selected when sociolinguistics and language contact make it possible for speakers to choose between competing forms.” Her examples include (a) the first and second person pronoun system in a language often include a pronoun with an /m/ and with a /t/-like consonant, respectively, for example, Finnish minä ‘I’ and sinä ‘you’; English me and older English thou, and (b) alternations between noncausatives and causatives are often causativizing, for example, the Dutch causative zetten ‘set’ being derived from the non-causative zitten ‘sit.’ Haig (2018a, b) also appeals to attractor states, and I will use his specific ideas on the Object Cycle and Seržant and Moroz’ (2022) work on Subject Cycles. In this book, we have seen cycles fixed in certain stages. For instance, as Tables 5.4 and 5.5 show, (subject) agreement is more common than having a pronoun or zero agreement, so this means loss of agreement and replacement by a pronoun is less common. This section will first explore what is responsible for attractor states in linguistics and provide a list with examples of cyclical inertia. I then explore if attractor states provide any insight into such cyclical stasis.

Explanations and mechanisms

201

Nichols (2018: 1) writes that “attractor states . . . accumulate more members than non-attractors, other things being equal.” “The reasons why an element is an attractor are linguistic (auditory salience, ease of processing, paradigm structure, etc.), but the factors that make selection possible and propagate selected items through the speech community are non-linguistic” (Nichols 2018: 1). They involve the right sociolinguistic and demographic conditions, for example, dense and distant social networks and sufficient population mobility. Nichols doesn’t mention cycles or grammaticalization, but the work by Haig (2018a) and Seržant and Moroz (2022) is immediately relevant for cycles, that is, for the stages in which certain cycles seem to be in balance, and these will be discussed later. What follows first is my observations on stasis in cycles, and some of this may be debatable. I’ll consider the cycles in order of appearance in the book. As for Determiner Cycles, only half of world’s languages have developed articles (see Chapter 3), so the first step of the cycle is not taken by many languages. The last stage of the cycle, the loss of the article, is not that frequent either, but it does occur in varieties of English and Aramaic. Stassen (2013) provides data on the presence of copulas with predicate nominals. His data show that 45% (175/386) of the world’s languages do not need a copula with predicate nominals so, like articles, the first step of the Copula Cycle hasn’t started in those languages. I don’t know of any data on the presence or absence of copulas with adjectival or locational predicates, but these would be predicted to be less deletable because of being stage-level predicates. TMA Cycles are very varied, and loss and gain of auxiliaries is frequent, so perhaps stasis is less relevant. For instance, Future and Progressive Cycles seem to me the most common, but some Aspect Cycles don’t seem to get off the ground, for example, begin and start continue to behave like real verbs, as Sims (2008) shows, and the Inceptive Cycle is therefore more like the Object Cycle. Negative Cycles are the most pervasive of all, and looking at Dryer’s (2013c) data, we get an impression on how the languages are patterned for the Jespersen Cycle (I leave out the data on negative auxiliaries). The numbers in (71) show that the single negative word (or particle) is the most frequent, followed by the affix. (71) grammaticalization negative word 502/1016 (=49%)

affix 395/1016 (=39%)

renewal double 119/1016 (=12%)

Interrogative and Complementizer Cycles involve movement to the left with subsequent reanalysis. I don’t see a stasis there. Meillet (1915/1916: 165; 169), in his ‘[l]e renouvellement des conjonctions,’ notes that conjunctions are of infinite diversity as the vast diversity across Indo European shows. The same lack of stasis appears with Pragmatic Cycles. Turning to the macro cycles, Table 5.2 provides numbers by Nichols and Bickel on head, dependent, double, and no marking, and the numbers are

202

Explanations and mechanisms

repeated in (72). They are relevant to the Analytic/Synthetic Cycle in that the data show languages hanging on to their inflection, either as agreement or case or both. (72) Head Dependent Double No Marking 71/234 (=30%) 63/234 (=27%) 58/234 (=25%) 42/234 (=18%) Subject and Object Cycles involve the loss and gain of head marking. As we also know from Chapter 5, subject pronouns are a lot rarer than verbal agreement, which argues for the stability of the inflectional stage in Subject Cycles. This preference is in accordance with the conclusions Seržant and Moroz (2022) reached, namely, that verbal agreement in languages tends to be relatively short, and person and number tend to combine into one morpheme. They describe these two attractor states in the following way: (a) the “attractor is characterized by different lengths for each person and number combination, structured along Zipf’s predictions,” that is, the more frequent the morpheme is, the shorter in length; and (b) “compositionality tends to be avoided,” that is, person and number are not separately coded. As Haig (2018a, b) shows, the Object Cycle is often disrupted before the pronoun becomes agreement. The object pronoun mostly remains a differential object marker (DOM). The DOM is an attractor state for Haig. His reason (p. 808), as is well-known, objects are more often expressed by full NPs than pronouns due to objects expressing new information more frequently. But this doesn’t explain it fully, because Object Cycles do start, but the categories involved are different than those of Subject Cycles. Case represents dependent marking. Baerman and Brown (2013) examine syncretism in case marking, but their data also provide insight into the presence and absence of case, as in (73). Case is therefore different than agreement in favoring the analytic stage of no marking. (73) case 75/198 (=38%)

no case 123/198 (=62%)

In this section, I have discussed ‘bottleneck’ stages in cycles. Attractor states might provide some insight in that speakers “strive towards less processing and articulatory effort while [they] strive towards lower lexicon complexity and lower memory costs” (Seržant & Moroz 2022) and therefore prefer synthetic stages. The explanations for grammaticalization presented in Sections 4 and 5 do not rely on usage and predictability but on structural economy. There is a similarity between this kind of economy and Seržant and Moroz’ (2022) “efficiency pressures.” The latter “strive towards less processing and articulatory effort” and “lower lexicon complexity.”

Explanations and mechanisms

203

Concluding this section, attractor states focus on efficiency and not on renewal, that is, cyclical change. The reason include them in this book is that they draw attention to cyclical stasis. 7

Conclusions

In this chapter, I discuss mechanisms and explanations for cyclical change. In Section 1, I first provide a little more background on when the terms ‘clarity and comfort’ start to be used. Sections 2 to 5 discuss answers from various approaches. Construction Grammar is concerned with cyclical change and grammaticalization but less so with cycles per se. Generative grammar provides a number of answers, especially to what starts cycles. That is also the case with a usage-based approach, such as attractor states. Suggestions for further reading External influences: Heine and Kuteva (2005). Other frameworks: Hilpert (2021), Traugott (2022), and Mithun (2016). Generative approaches: Roberts and Roussou (2003), Bouzouita et al. (2020). Competition: Rudanko (2006), Mair (2001).

Review questions and exercises

1. Explain comfort and clarity in your own words. 2. Look at the description of standard average European (SAE), for example, on https://en.wikipedia.org/wiki/Standard_Average_European. Apart from the perfect, as considered in Section 2, are there other constructions that European languages have in common? If yes, do any of these involve the cycles discussed in this book? 3. Construction Grammar examines constructions in competition. Take a look at -ing complements versus the to-infinitive in COHA with a verb of your choice. 4. In this chapter, there have been accounts for the Negative, Determiner, Pronoun, and Complementizer Cycles in terms of phrases reanalyzing as heads (using the HPP). I have not provided an account for the Interrogative Cycle. How would you go about doing that? Advanced exercise

For the language you have selected in Chapter 1, summarize what evidence there is for cycles based on the answers you gave for the questions in Chapters 3 to 5. Then, if there is a theoretical framework you use/are familiar with, how would that model account for the changes?

7 CONCLUSIONS AND FUTURE DIRECTIONS

This book has provided some background on the linguistic cycle and given numerous examples of cycles. It has distinguished between micro cycles, which affect only a subpart of a language’s grammar, and macro cycles, which affect the typological character of a language. Cycles emphasize loss and renewal, and the mechanisms they use provide insights into what the faculty of language looks like and what the universal grammatical and semantic features are. Table 7.1 provides a list of all the cycles discussed in detail. In this last chapter, I will provide some final thoughts. In Section 1, I review some insights that the study of linguistic cycles can bring. In Section 2, I provide a list of criticisms of cyclical change and a reply to those. In Section 3, I catalogue the cycles not discussed here and provide suggestions for future directions. 1

Insights from cycles

Cycles involve all grammatical and lexical categories. The grammatical ones considered in this book include determiners, complementizers, copulas, negatives, agreement, case, tense, mood, and aspect, voice, and adverbs. They give insight in the typological inventory. Some categories, for example, demonstratives, are resilient. Although they reanalyze as articles and copulas, they can continue to be used as demonstratives. The same is true with pronouns that may look similar over hundreds of years, showing that they renew using the old forms. I have considered accounts of what sets the cycle in motion. The traditional idea of clarity corresponds to semantic or pragmatic strengthening and that of comfort to phonetic weakening. These two forces are translatable into various frameworks. Both renewal and weakening can set cyclical change into motion DOI: 10.4324/9781003272564-7

Conclusions and future directions

205

TABLE 7.1 The linguistic cycles in this book

Micro Determiner Cycles

Copula Cycles

Imperfective Cycles

Perfective Cycles

Mood Cycles

Voice Cycles

Negative Cycles

Interrogative Cycles

Complementizer Cycles

Pragmatic Cycles

Macro Analytic to synthetic to analytic Subject Agreement Cycles

Object Agreement Cycles

Case Cycles

(see Table 6.3). For instance, Negative Cycles show both starts, Adverb Cycles start with renewal, and Determiner and Copula Cycles show weakening and loss first. When grammaticalization occurs first, certain features can disappear from a language, for example, the perfective in Middle English, and they can also be introduced, for example, the progressive in relatively recent English. These two different forces have been treated as pull and push chains: if grammaticalization is first, it is a pull-chain, and if renewal is first, it is a push-chain (see also Blaxter & Willis 2018)! Cycles show shortcomings in various theoretical frameworks. If cyclical change indeed occurs, as I have argued, we need to account for the reasons behind the two forces that set them in motion. For me, working in a generative framework, it is easy to explain the grammaticalization side of the cycle in terms of structural economy or third factor principles. However, the renewal side is not as easy since languages do not need to renew their grammatical categories and can function with lexical expressions of tense, mood, and aspect, for instance. Therefore, one of the remaining puzzles is why certain features are expressed grammatically. One answer could be that lexical categories are slower to retrieve and articulate and therefore preferably used for emphasis; grammatical means are preferred from an economy point of view. In Chapter 6, I have formulated this as phrase to head reanalysis. This puzzle could also be thought of as entropy. As Mansfield (2021: 1428) writes, “predictability plays a major role in human language structure and use.” The more unpredictable the message, the more informative it is; the more predictable, the more compressed it can be. Mansfield connects this to entropy, which is a way to measure the predictability of a word or phrase (as in Shannon 1948, 1951), and he hypothesizes that words will be more predictable than phrases. The argument is that the former is more predictable, and the latter has “non-compositional semantics and morphological

206

Conclusions and future directions

irregularity.” As with other accounts, entropy doesn’t predict which categories are expendable. 2

Criticisms of the cycle

The following criticisms have been leveled against the notion of a linguistic cycle: (a) although micro cycles exist, macro cycles are not well-attested; (b) languages cannot be without certain markers, for example, a negative, so they cannot disappear as the cycle might imply; (c) stages in a cycle constitute separate, independent changes; and (d) cycles are not unidirectional. I’ll discuss each of these. As for (a), I mentioned in Chapter 1 that cycles such as the Negative, Determiner, and Copula Cycles are generally accepted by typological and historical linguists. The original cycle that von der Gabelentz had in mind is about major typological change and is more controversial. In Chapter 5, I have provided the few cases of Analytic to Synthetic to Analytic Cycles that are known. One of the problems with this cycle is that it is hard to define analytic and synthetic, and I have therefore used Nichols’ characterization of a language in terms of head and dependent marking. Regarding (b), Reinöhl and Himmelmann (2017: 386; 399) ask “[i]f the old construction was so important that it had to be replaced, why was it lost or allowed to weaken in the first place” (384)? In fact, they argue that positing a separate process of renewal is not needed because it overlaps with grammaticalization. Their view “by contrast, [is] that a language and its elements and structures will always in principle be functionally adequate, being shaped by speaker-hearers’ needs” (384). That, however, begs the question why speakers and hearers of different languages would have different needs: why would not all languages need the same? My own view on this point is that certain features cannot be lost, for example, negatives, and that this provides typological and cognitive insight. It may be the case such truth conditional categories are different. However, they need not be expressed grammatically: negation can be expressed lexically, as in (1). Of course, these lexical mens can become grammaticalized but need not. (1)

a. This language lacks negation. b. This language functions without negation.

As for (c), how quickly do the developments in a cycle have to follow each other to be considered as a cycle (Hansen p.c.)? In Chapter 3, I have considered the Determiner Cycle, which shows a gap of at least 200 years between the creation of an article and the strengthening of the demonstrative. I think cycles can

Conclusions and future directions

207

remain in a stage of stasis for a long time and, as we saw with the discussion of Attactor States in Chapter 6, do not have to keep going. Finally, regarding (d), degrammaticalization has been recognized for some time and that goes against the unidirectionality of the cycle. See, for example, Newmeyer (1998) and Norde (2009). Traugott (2001) provides an assessment of degrammaticalization and writes that some cases “are not well understood yet,” for example, the change of the English genitive clitic ‘s, as in (2), which supposedly derives from a (more grammatical) case affix. (2)

in that king’s days < on þæs cyninges On that.GEN king.GEN.S

dagum day.DAT.P

Traugott also suggests that some of the older forms may have been retained in a language so that what looks degrammaticalization is the retention, side by side, of the older (less grammatical) form and the younger (more grammatical) one. There are genuine cases of degrammaticalization, but these seem isolated and occur under special circumstances, for example. a hypercorrection or euphemism. For instance, the case of Pennsylvanian German wotte ‘would’ that Traugott mentions changes from a subjunctive modal into a lexical verb, but none of the other modals do. So those cases of degrammaticalization seem to be lexicalizations. Van der Auwera and Plungian (1998: 116) call attention to a case of demodalization of må ‘may’ in Swedish and mogen ‘may’ Dutch, but the meanings of the new lexical verbs seem unpredictable. Note the semantic differences between them in (3) and (4). (3) (4)

Jag mår bra 1S feel well ‘I feel well.’ Ik mag hem 1S like him ‘I don’t like him.’

Swedish niet not

Dutch

Kiparsky (2011) provides ten examples of what look like degrammaticalization – many the same as Traugott – and attributes them to analogical extension. Igartua (2015) presents counterexamples to a cycle I discuss in Chapter 5 (Section 2), namely, to the cycle of attachment types in which isolating languages turn to agglutinative stages and then to fusional ones. Igartua chronicles some cases where fusional languages develop agglutinative morphology. For instance, in Old Armenian (5a), case and number are marked by one morpheme, and in Modern Armenian (5b), they are marked as two separate morphemes.

208

Conclusions and future directions

(5)

a. azg-ac‘ > people-DAT.P =fusional

b. azg-er-i people-DAT-P =agglutinative

Old Armenian

Similar changes in the nominal paradigm occur in Ossetic, Georgian, Tocharian, Lithuanian, Marathi, and Bengali. The verbal paradigm is similarly reorganized in Cappadocian and Bythinian Greek. Igartua identifies three mechanisms of change: grammaticalizing adpositions for case, insertion of special morphemes, and affix secretion. The first two are additive and could be seen as renewal of one of the two markers and not as a counterexample to the cycle. However, even if a language-internal reason cannot be found, an external one exists in these case. As Igartua notes, many of these cases “have evolved in conditions of close contact with languages characterized by inflectional systems of the agglutinative type” (2015: 688) and “language contact . . . seems to be a highly favoring, if not determinant, factor in fusionalto-agglutinative changes” (2015: 697). The regions that show close contact are the Caucasus (Armenian, Ossetic, and Georgian), Turkey (Cappadocian Greek dialects), and South Asia (Bengali and Marathi). Igartua’s conclusions are that there are two opposing forces, “the predominantly language-internal tendency toward fusion (through reductive phonological processes), and the predominantly language-external tendency toward agglutination” (712). Concluding, genuine questions can be raised as to the status of cycles: do all exist, and how fast does change have to occur for it to be seen as one set of related changes? Degrammaticalization is not a widespread, systematic phenomenon, and features can renew themselves before they become obsolete. 3

Future directions

I’ll mention two future directions, one more philosophical and one practical. A comment that has come up occasionally is why certain languages do or don’t participate in certain (cyclical) changes. As mentioned, East Asian languages do not undergo phonological erosion in their cycles of change, as Post (2007) maintains for Thai and Bisang (2008) for languages in East and mainland Southeast Asia. Ritter and Wiltschko (2019) and McDonald (2021) have argued that the pronouns in languages such as Japanese and Korean are paranouns situated in the discourse area and do not agree with the verb. Such a position outside of the grammatical layer would stop grammaticalization of the paranoun as clitic and agreement marker. Another possible reason behind the blocking of a cycle could be the loss of V to T to C, for example, in English. The movement of the verb to the left of the subject pronoun is necessary for an agreement clitic and suffix to arise. Linking the typology of a language to cyclical change will provide much needed insight.

Conclusions and future directions

209

TABLE 7.2 Other cycles, not discussed or not in detail

Nominal domain Diminuative: Degree: Dual Cycle: Plural Cycle: Partitive: Clausal Domain Quantifier Cycle: Comparative Cycle: Comitative Cycle Distributive Cycle: Future Cycle: Reflexive Cycle: Relative Cycle:

Rose (2018) Wood (2016) Slobodchikoff (2013, 2019); Guðmundsson (1972) Lehmann (2015); Heine and Reh (1984) Seržant (2021b) Willis (2011) Jäger (2021); Hubers and de Hoop (2013) Haspelmath (2007) Mithun (2016) Meillet (1912); Bybee et al. (1994); van Gelderen (2011); Detges (2020) Kiss and Mus (2022) Bacskai-Atkari (2018); van de Velde (2021); van Gelderen (2009b); Rutten and van der Wal (2014)

A more practical direction is discovering new cycles. This book has examined ten types of micro cycles in detail and, in passing, has mentioned a few others, for example, the Dual Cycle in Chapter 1 and the Future Cycle in Chapter 3. More cycles have been identified, and these are listed in Table 7.2, together with references to them. The cited studies have not always asked the same questions, for example, what starts the cycle, what the mechanisms are, and whether or not there is a stage of stasis. These are also possible avenues for future work. This book has covered cycles in one modality of language, namely, spoken language and that is because of my own background. Sign languages, however, provide a lot of evidence for grammaticalization and cyclical change, with similar pathways as spoken languages. See Wilcox and Wilcox (1995) for an early formulation and Pfau and Steinbach (2011) for a review. Pointing gestures grammaticalize into demonstrative and personal pronouns and then into auxiliaries and agreement markers (e.g. Fischer 1996; Pfau et al. 2018). Aspect and tense markers develop from lexical verbs. Thus, verbs of completion grammaticalize to perfective markers (Zucchi 2003) and motion verbs to future markers (Janzen & Shaffer 2002). Negatives undergo what Pfau (2015) calls a Jespersen Cycle, and Sampson and Mayberry (2022) have identified a Copula Cycle. This too is extremely promising work.

210

4

Conclusions and future directions

Conclusion

In conclusion, languages change in systematic ways. The only plausible reason for this is that the learners have principles guiding these changes. The urge of speakers to be innovative may introduce new, loosely adjoined elements into the structure, and that may provide the impetus to the language learner to reanalyze the older form as grammatical. Review questions and exercises

1. What insights have linguistic cycles given you? 2. Which aspect of the cycle do you see as the most problematic? 3. Table 7.2 mentions a Plural Cycle. Dahl (2004: 122) has argued that all the could be becoming a plural definite article. How might you go about arguing that this is the case or not? Final advanced exercise

Turn the answers you gave (on the advanced exercises) in Chapters 1 and 3 to 6 into a paper. You could write about one cycle or, more generally, ponder if there is evidence of cyclical change in the language (or family). You could try to explore a cycle using a few different methodologies: what does the comparative method point to; what do you find in the historical documents; how about the internal method, and does the cycle involve borrowing? What does the typology point to? Does the cycle appear to affect any other part of the grammar, or did any other change in the grammar appear to set this change into motion?

POSSIBLE ANSWERS TO REVIEW QUESTIONS AND EXERCISES

Chapter 1

1. A cycle involves a change whereby a lexical word may be reanalyzed as grammatical (e.g. from motion verb to future auxiliary) and is then renewed by a new lexical source. In a micro cycle, a small area of the grammar is reorganized and, in a macro cycle, the typological character of a language is. Analytic languages express grammatical relations, definiteness, tense, aspect, etc., by means of single words where synthetic languages use affixes. A spiral is an alternative to cycle expressing that the renewal can be a very different origin from the original lexical item. 2. When you look up may in the OED, you get to a list of different uses, for example, the noun used for the month of May and three other noun meanings. There are also three verb meanings, and since you are looking at the auxiliary, look here. The first in that list of verbs is the one used for future predictions, so that is the relevant sense. Its first listed use is as “a full verb, or as an auxiliary with a verb in the infinitive understood,” meaning, for example, “to be strong; to have power or influence,” as in (33). (33) aris dryhten ne meg mon rise.up lord NEG may man ‘Rise up, let no man have the power’ (OED, Vespasian Psalter (1965) ix. 20). Then look at a later use, and you see a change to “an auxiliary verb with a following bare infinitive,” which is evidence of a Mood Cycle, as we will see in Chapter 3. 3. That would depend on which cycle you find interesting and which language descriptions you can find. Let’s say you find the Negative Cycle interesting.

212

Possible answers to review questions and exercises

You could compare some languages in a family and see if there is an independent negative word in one language and a clitic-like one in another. 4. In Chapter 4, I will look at the verb fail as a possible replacement for the weakening negative, and you could try the same in anticipation of that discussion. You could also check in COHA if the use of never has gone up. Figure 1.1 shows the changes, and not much has in fact changed; it may even have decreased in use. 5. In (31), there is in fact a renewal because bam ‘both’ is used while the old dual inc ‘you two’ is still there. Checking the first 20 instances in (the Spoken) COCA, most are predeterminers, as in (34), but four are used to express dual pronouns, as in (35). (34) ‘a whole lot of people are dissatisfied with both the Democratic and Republican parties’ (CBS 2019). (35) ‘Both of you, thank you so much for being here tonight and being an example’ (Fox 2019). 6. Chapter 6 shows that the use of -ing in complements to verbs like dislike is increasing, and that is the case here too. As for how lexical start is, there are quite a few lexical uses in the first 20 instances of it in COCA, so start is not only used with other verbs to be an inceptive auxiliary but also functions as a verb on its own. The verb begin likewise continues to be lexical and has an almost identical trajectory as start in terms of increasing the use of -ing complements, as shown in Figure 1.2. However, the numbers are much smaller, possibly an indication of a replacement of begin by start.

FIGURE 1.2

The use of never in COHA

FIGURE 1.3

Begin followed by an -ing verb

Possible answers to review questions and exercises

213

Chapter 2

1. As far as we know, cyclical change was of interest in the eighteenth century, although the term wasn’t used. It would be the macro cycle. 2. I’ll just give an answer from my perspective: because the generative framework is interested in how the internal system that generates language is acquired, it has influenced the way I look at change. Thus, I look at biases that the cognitive system may put on the process, for example, economy. 3. As mentioned in Chapter 1, the OED doesn’t list a linguistic use under ‘cycle,’ so I did an advanced search (instead of a quick search) to see if it appeared elsewhere, but it didn’t. Chapter 3

1. A possible formulation: through the Copula Cycle, a copula is lost because it may be reanalyzed as another element, for example, a negative. It can be renewed through new locative/deictic features. As to why it is optional, it links a subject and predicate and adds shades of meaning of that relationship, for example, identity or location. That may not be as crucial. This impacts Copula Cycles because these may not be as frequent as other cycles. 2. A possible question could be if there currently is an increase in the reinforced demonstrative, and the data to answer this question could come from COCA. These data show that although the increase in that there is not very strong overall, it is strong in spoken. 3. This is evidence of a Determiner Cycle. In Old Norse, the demonstrative hinn can also be placed after the noun so it has a bridging context. 4. There is evidence of grammaticalization in Russian in which a full reflexive pronoun in (203) and a clitic -sja in (204) are related. There is also a middle voice use in (205). The direction of independent pronoun to clitic to voice marker would make sense from what we know about other languages as well. 5. In Section 4.1, we have seen that the -ing may be generalizing from progressive to continuous aspect. If it were to generalize further, it would include the habitual to become a general imperfective. That is precisely what is happening in (206).

FIGURE 3.8

That there + noun (downloaded 18 April 2022)

214

Possible answers to review questions and exercises

FIGURE 3.9

That + noun + there (downloaded 18 April 2022)

Chapter 4

1. The NEC starts out with a negative marker that is used for both existential and other verbs. At some point, the existential verb and the negative contract and the two types of verbs have a different negative shape. In the last stage, the special negative used in the existential is generalized to other verbs as well. 2. This is a case where a preposed wh-word is reanalyzed as a yes/no marker. 3. Look in 2011 for the solution: unexpected. 4. Emilian Italian has a weakened negative and an innovated new minimizer, as compared to Standard Italian. 5. I searched for instances of ‘how’ at the beginning of a sentence followed by a modal auxiliary but only found examples with a manner adverbial ‘how,’ such as (120), not with a yes/no marker. (120) ‘How would he find peace to move on with his life?’ (COCA, NBC 2013). 6. In the early stages of the language, the oblique plural u initially spreads to the oblique singular before spreading to all second person polite forms in the current stage. The nominative plural ghi (gij) similarly spreads to the nominative singular in the early stages and changes into jij, which is the familiar second person. Renewal of the singular oblique comes in the form of jou and of the plural in the form of jullie, which takes gij, adds lieden ‘people,’ and contracts those. Something similar happened in English where oblique plural you is initially used for the polite singular and plural and then spreads to all forms. 7. In (118), the standard English -n’t is strengthened by a stronger negative never. In (119), it is replaced by it. Chapter 5

1. That depends. If we just count inflectional morphology, English will only have one category, the plural on nouns. If we counted derivational morphology, it would be more, for example, anti-dis-establish-ment-arian-ism.

Possible answers to review questions and exercises

215

The first person singular pronoun separated from and adjacent to the auxiliary/verb was

FIGURE 5.3

2. Szmrecsanyi examines if a word has more than one morpheme, and if it does, it counts as synthetic. He doesn’t measure how synthetic it is, unlike Bickel and Nichols. Szmrecsanyi also measures both syntheticity and analyticity, and a word can be both, for example, the auxiliary were. 3/4. Jamaican is very analytic. There is no evidence of any synthetic endings. It would therefore neither be head nor dependent marking. Tawala is head marking (of both subject and object) and is therefore synthetic; it is not dependent marking. 5. You could try to find a way to measure if pronouns are always/often adjacent to verbs. I looked in the Spoken COCA and found some evidence of (nominative) pronouns needing to be more adjacent to the verb. Figure 1.5 shows that the first person pronoun I is more often adjacent to the auxiliary/verb was. This might be evidence of change. 6. Most of the agreement markers are reduced versions of the pronouns. The exception is the first person (and third person feminine). There could have been a sound or other change in Arabic (see Dimmendaal 2011: 96, if you are interested, which is based on Hetzron 1976). 7. The main tendency is the same: suffixing (529/969 = 55%) is more prevalent than prefixing (152/969 = 16%). Table 5.11 also distinguishes the different word orders. Table 5.18 is helpful in connection with the discussion around (27). Dryer’s and Siewierska’s numbers of zero agreement are 9% and 22%. Table 5.18 is somewhere in between at 15% (141/969). Chapter 6

1. Another word for comfort is ‘ease,’ and this describes changes due to economy of speaking. Clarity is relevant to the hearer but also to the speaker: the speaker has to decide if the message is clear but can also decide to be creative! 2. SAE languages typically have articles, and these are relevant to the Determiner Cycle. However, even non-Indo European SAE languages such as

216

Possible answers to review questions and exercises

Finnish are reanalyzing demonstratives as articles (Laury 1991). Analytic comparatives (greater than) may be part of macro cycles, but most of the other characteristics are not relevant to cyclical change. 3. In Chapter 1, we looked at possible changes involving begin and start. If you look at other verbs, you will find that the -ing complement is generally encroaching. 4. In the Interrogative Cycle, a verb or auxiliary with the negative marker attached moves to the left periphery of the sentence to check the Q-features and can then be reanalyzed as ‘just’ a question-marker. Chapter 7

1. The delight in finding patterns! (My personal answer) 2. The varying speed of the changes. (My personal answer) 3. The quantifier all is used for the second person plural, as in (5). (5) ‘It is an absolute privilege to join you all tonight’ (COCA Spoken 2018). Arguing that all is used to mark the plural in the definite article is trickier. It would have to be used in cases where it is the plurality that matters, not the totality, as in perhaps (6). (6) I cooked all the rice.

REFERENCES

Abraham, Werner 1991. The grammaticalization of German modal particles. In Elizabeth Traugott & Bernd Heine (eds.), Approaches to Grammaticalization, 331–380. Amsterdam: Benjamins. Abraham, Werner & Jac Conradie 2001. Präteritumschwund und Diskursgrammatik. Amsterdam: John Benjamins. Adamović, Milan 1985. Konjugationsgeschichte der türkischen Sprache. Leiden: Brill. Adams, Mariane 1987. From Old French to the theory of pro-drop. Natural Language & Linguistic Theory 5.1: 1–32. Ahn, Hyug 2006. The Semantics of SJA in Russian: Focus on the Action. UNC dissertation. Al Aloula, Mashael 2021. The Emergence of a Progressive Aspect in Najdi Arabic. George Mason University PhD. Alasmari, Abdullah 2022. The Periphrastic Passive Grammaticalization in Modern Standard Arabic. ASU MA thesis. Albuhayri, Salem 2013 The Pronominal System in Standard Arabic: Strong, Clitic and Affixal Pronouns. ASU MA thesis. Aldridge, Edith 2011. Neg-to-Q: The historical origin and development of question particles in Chinese. The Linguistic Review 28: 411–447. Allen, Cynthia 1980. Movement and deletion in Old English. Linguistic Inquiry 11.2: 261–323. Allen, Cynthia 1995. Case Marking and Reanalysis. Oxford: Clarendon Press. Allen, Cynthia 1997. Middle English case loss and the ‘creolization’ hypothesis. English Language and Linguistics 1.1: 63–89. Allen, Cynthia 2007. The case of the genitive in the Peterborough Continuations. In Alexander Bergs & Janne Skaffari (eds.), The Language of the Peterborough Chronicle, 77–92. Frankfurt: Peter Lang Verlag. Alsaeedi, Mekhlid 2015. The Rise of New Copulas in Arabic. ASU MA thesis. Alsaeedi, Mekhlid 2019. Existential and Negative Existential Constructions in Arabic. ASU PhD.

218

References

Altamimi, Mansour 2021. The Grammaticalization of Lexical Verbs into Progressive and Future Markers in Saudi Najdi Arabic. ASU PhD. Andersen, Henning 1973. Abductive and deductive change. Language 48: 11–50. Andersen, Henning 2008. Grammaticalization in a speaker-oriented theory of change. In Thórhallur Eythórsson (ed.), Grammatical Change and Linguistic Theory: The Rosendal Papers, 11–44. Amsterdam: John Benjamins. Anderson, Stephen 1992. A-Morphous Morphology. Cambridge: Cambridge University Press. Ansaldo, Umberto & Lisa Lim 2004. Phonetic absence as syntactic prominence. Grammaticalization in isolating tonal languages. In Olga Fischer, Muriel Norde & Harry Perridon (eds.), Up and Down the Cline – the Nature of Grammaticalization, 345–362. Amsterdam: Benjamins. Arends, Jacques 1986. Genesis and development of the equative copula in Sranan. In Pieter Muysken & Norval Smith (eds.), Substrata Versus Universals in Creole Genesis, 103–127. Amsterdam: John Benjamins. Ashby, William 1974. Il parle or Iparle? Prefixed inflection in French. Semasia 1: 83–93. Ashby, William 1976. The loss of the negative morpheme, ne, in Parisian French. Lingua 39: 119–137. Ashby, William 1977. Clitic Inflection: An Historical Perspective. Amsterdam: Rodopi. Auderset, Sandra 2021. The Antipassive and Its Relationship to Person Markers. https:// sauderset.github.io/publication/auderset-2021-the-antipassive/auderset-2021-theantipassive.pdf Bácskai-Atkári, Júlia 2014a. Cyclical change in Hungarian comparatives. Diachronica 31.4: 465–505. Bacskai-Atkari, Julia 2014b. The Syntax of Comparative Constructions: Operators, Ellipsis Phenomena and Functional Left Peripheries. Potsdam: Universitätsverlag Potsdam. Bacskai-Atkari, Julia 2018. The relative cycle in Hungarian declaratives. In Ana Maria Martins & Adriana Cardoso (eds.), Word Order Change, 68–87. Oxford: Oxford University Press. Bácskai-Atkári, Júlia & Éva Dékány 2014. From non-finite to finite subordination. The history of embedded clauses. In Katalin Kiss (ed.), The Evolution of Functional Left Peripheries in Hungarian Syntax, 148–223. Oxford: Oxford University Press. Baerman, Matthew & Dunstan Brown 2013. Case syncretism. In Matthew Dryer & Martin Haspelmath (eds.), The World Atlas of Language Structures Online. Leipzig: Max Planck Institute for Evolutionary Anthropology. Bahrt, Nicklas 2021. Voice Syncretism. Berlin: Language Science Press. Bahtchevanova, Mariana & Elly van Gelderen 2016. The interaction between the French subject and object cycles. In Elly van Gelderen (ed.), Cyclical Change Continued, 113–136. Amsterdam: Benjamins. Bailey, Charles & K. Maroldt 1977. The French lineage of English. In Jürgen Meisel (ed.), Langues en Contact, 21–53. Tübingen: Narr. Bailey, Laura 2013. Question particles: Thai, Japanese and English. Linguistica Atlantica 32: 34–51. Baker, Robin 1985. The Development of the Komi Case System. Helsinki: suomaleisUgrilainen seura.

References

219

Bakir, M.J. 2010. Notes on the verbal system of Gulf Pidgin Arabic. Journal of Pidgin and Creole Languages 25.2: 201–228. https://doi.org/10.1075/jpcl.25.2.01bak Bally, Charles. 1965[1913]. Le langage et la vie. Genève: Droz. Barðdal, Johanna 1999. Case in Icelandic – A construction grammar approach. Tijdschrift voor Skandinavistiek 20.2: 65–100. Barðdal, Johanna & Spike Gildea 2015. Diachronic Construction Grammar: Epistemological Context, Basic Assumptions and Historical Implications. Amsterdam: John Benjamins. Bashir, Elena 2006. Change in progress: Negation in Hindi and Urdu. In Rajendra Singh (ed.), The Yearbook of SA Languages and Linguistics, 3–30. Berlin: Mouton. Bately, Janet (ed.) 1980. The Old English Orosius. Oxford: Oxford University Press. Bencini, Giulia 2003. Toward a diachronic typology of yes/no question constructions with particles. Chicago Linguistic Society 39: 604–621. Benítez-Burraco, Antonio 2017. Grammaticalization and language evolution: Focusing the debate. Language Sciences 63: 60–68. Benson, Larry 1987. The Riverside Chaucer. Boston: Houghton Mifflin Company. Benveniste, Emile 1966 [1971]. Problèmes de linguistique générale. Paris: Gallimard. Benveniste, Emile 1968. Mutations of linguistic categories. In Y. Malkiel & W.P. Lehmann (eds.), Directions for Historical Linguistics, 83–94. Austin: University of Austin Press. Bergs, Alexander & Gabrielle Diewald (eds.) 2008. Constructions and Language Change. Berlin: Mouton de Gruyter. Bergsland, Knut 1994. Sydsamisk grammatik. Karasjok: Davvi Girji o.s. Bermúdez-Otero, Ricardo & Graeme Trousdale 2012. Cycles and continua: On unidirectionality and gradualness in language change. In Terttu Nevalainen & Elizabeth Closs Traugott (eds.), The Oxford Handbook on the History of English, 691–720. Oxford: Oxford University Press. Bernstein, Judy 1997. Demonstratives and reinforcers in Romance and Germanic languages. Lingua 102: 87–113. Bernstein, Judy, Francisco Ordóñez & Francesc Roca 2021. Emphatic elements and the development of definite articles. Journal of Historical Syntax 5.22: 1–32. Biber, Douglas & Susan Conrad 2019. Register, Genre, and Style. Cambridge: Cambridge University Press. Bickel, Balthasar & Martin Gaenszle 2015. First person objects, antipassives, and the political history of the Southern Kirant. Journal of South Asian Languages and Linguistics 2.1: 63–86. Bickel, Balthasar & Johanna Nichols 2005. Inflectional morphology. In Timothy Shopen (ed.), Language Typology and Syntactic Description. Cambridge: Cambridge University Press. Bickel, Balthasar & Johanna Nichols 2013. Inflectional synthesis of the verb. In Matthew Dryer & Martin Haspelmath (eds.), The World Atlas of Language Structures Online. Leipzig: Max Planck Institute for Evolutionary Anthropology. Available online at http://wals.info/chapter/22, accessed on 2021–06–08. Bickerton, Derek 1990. Language and Species. Chicago: University of Chicago Press. Binyam Sisay Mendisu 2008. Aspects of Koorete Verb Morphology. Oslo PhD. www.yumpu. com/en/document/read/16449958/aspects-of-koorete-verb-morphology-bitbucket Bisang, Walter 2008. Grammaticalization and the areal factor: The perspective of East and mainland Southeast Asian languages. In María José López-Couso & Elena Seoane

220

References

(eds.), Rethinking Grammaticalization: New Perspectives, 15–35. Amsterdam: John Benjamins. Blake, Barry 2001. Case. Cambridge: Cambridge University Press. Blaxter, Tam & David Willis 2018. Pragmatic differentiation of negative markers in the early stages of Jespersen’s cycle in North Germanic. Diachronica 35.4: 451–486. Bogomolova, Natalia 2018. The rise of person agreement in East Lezgic: Assessing the role of frequency. Linguistics 56.4: 819–844. Bolinger, Dwight 1980. Language – The Loaded Weapon. London and New York: Longman. Bont, A.P. de 1962. Het Dialect van Kempenland I. Assen. Bopp, Franz 1816. Über das Conjugationssystem der Sanskritsprache in Vergleichung mit jenem der griechischen, lateinischen, persischen und germanischen Sprachen. Frankfurt-am-Main: Andreäischen Buchhandlung. Bopp, Franz 1833. Vergl eichende grammatik des sanskrit, zend, griechischen, lateinischen, litthauischen, gothischen und deutschen. Berlin: F. Dümmler. Bopp, Franz 1885. A Comparative Grammar of the Sanscrit, Zend, Greek, Latin, Lithuanian, Gothic, German, and Slavonic Languages. London: Williams. Bouzouita, Miriam & Anne Breitbarth 2019. ‘U is’ or ‘U bent’? Polite forms of address and verbal agreement in historical Dutch and Ibero Romance. In Metin Bağrıaçık, et al. (eds.), Mapping Linguistic Data. www.haegeman.ugent.be/wp-content/ uploads/2019/06/BouzouitaBreitbarth.pdf Bouzouita, Miriam, Anne Breitbarth, Lieven Danckaert & Elisabeth Witzenhausen (eds.) 2020. Cycles in Language Change. Oxford: Oxford University Press. Bray, Dorothy & The White Mountain Apache Tribe 1998. Western Apache-English Dictionary. Tempe: Bilingual Press. Bréal, Michel 1924. Essai de Semantique. Paris: Hachette. Breitbarth, Anne 2014. The History of Low German Negation. Oxford: Oxford University Press. Breitbarth, Anne, Christopher Lucas & David Willis 2020. The History of Negation in the Languages of Europe and the Mediterranean II: Patterns and Processes. Oxford: Oxford University Press. Breu, Walter 1990. Sprache und Sprachverhalten in den slavischen Dörfern des Molise (Süditalien). In Walter Breu (ed.), Slavistische Linguistik 1989, 35–65. Munich: Otto Sagner. Brook, G. & R. Leslie 1963. Layamon: Brut. Oxford: Oxford University Press [EETS 250]. Broz, Vlatko 2013. Aspectual properties of the verbal prefix – a in Old English with reference to Gothic. In Gabriele Diewald, et al. (eds.), Comparative Studies in Early Germanic Languages, 235–262. Amsterdam: John Benjamins. Brunner, Karl 1962. An Outline of Middle English Grammar. Oxford: Blackwell [1970 translation]. Butt, Miriam & Aditi Lahiri 2002. Historical Stability vs Historical Change. Ms. Bybee, Joan 2002. Word frequency and context of use in the lexical diffusion of phonetically-conditioned sound change. Language Variation and Change 14: 261–290. Bybee, Joan 2010. Language, Usage and Cognition. Cambridge: Cambridge University Press. Bybee, Joan & Östen Dahl 1989. The creation of tense and aspect systems in the languages of the world. Studies in Language 13.1: 51–103.

References

221

Bybee, Joan, Revere Perkins & William Pagliuca. 1994. The Evolution of Grammar: Tense, Aspect and Modality in the Languages of the World. Chicago: University of Chicago Press. Carey, Kathleen 1995. Subjectification and the development of the English perfect. In Dieter Stein & Susan Wright (eds.), Subjectivity and Subjectification, 83–102. Cambridge: Cambridge University Press. Chatzopoulou, Aikaterini 2019. Negation and Nonveridicality in the History of Greek. Oxford: Oxford University Press. Chatzopoulou, Aikaterini 2021. The history of Greek conditionals & one modal cycle. Journal of Historical Syntax 5: 16–25. https://ojs.ub.uni-konstanz.de/hs/index.php/hs/ article/view/71 Cheng, Lisa, James Huang & Jane Tang 1996. Negative particle questions. In James Black et al. (eds.), Microparametric Syntax and Dialect variation, 41–78. Amsterdam: John Benjamins. Chomsky, Noam 1965. Aspects of a Theory of Syntax. The Hague: Mouton. Chomsky, Noam 1966. Cartesian Linguistics. New York: Harper & Row. Chomsky, Noam 1995. The Minimalist Program. Cambridge, MA: MIT Press. Chomsky, Noam 2005. Three factors in language design. Linguistic Inquiry 36.1: 1–22. Chomsky, Noam 2013. Problems of projection. Lingua 130: 33–49. Chomsky, Noam 2015. Problems of projection: Extensions. In Elisa Di Domenico et al. (eds.), Structures, Strategies, and Beyond, 3–16. Amsterdam: John Benjamins. Chomsky, Noam, Ángel J. Gallego, and Dennis Ott 2019. Generative grammar and the faculty of language: Insights, questions, and challenges. Catalan Journal of Linguistics, Special issue, 229–261. https://doi.org/10.5565/rev/catjl.288 Closs, Elizabeth 1965. Diachronic syntax and generative grammar. Language 41.3: 402–415. Comrie, Bernard 1976. Aspect. Cambridge: Cambridge University Press. Comrie, Bernard 1980. Language Universals and Linguistic Typology. Chicago: Chicago University Press. Comrie, Bernard 1985. Tense. Cambridge: Cambridge University Press. Condillac, Etienne Bonnot de 1746. Essai sur l’origine des connaissances humaines. Paris. http://classiques.uqac.ca/classiques/condillac_etienne_bonnot_de/essai_origine_des_connaissances/condillac_origine_des_connaissances.pdf Conrady, A. 1915. Der altchinesische Fragesatz und der steigende Ton. Mitteilungen des Seminars für Orientalische Sprachen 18: 261–297. Cook, Eung-Do 2004. Dëne Sųłiné (Chipewyan). Algonquian and Iroquoian Linguistics, Memoir 17, Special Athabaskan Number. Cormack, Annabel & Neil Smith 2002. Modals and negation in English. In Sjef Barbiers, Frits Beukema & Wim van der Wurff (eds.), Modality and Its Interaction with the Verbal System, 133–163. Amsterdam: John Benjamins. Corver, Norbert & Henk van Riemsdijk 2001. Semi-lexical Categories. Berlin: Mouton de Gruyter. Coseriu, Eugen 1972. Die Geschichte der Sprachphilosophie von der Antike bis zur Gegenwart. 2 vols. Tübingen: Eine Übersicht. Cournane, Ailís 2014. In search of L1 evidence for diachronic reanalysis: mapping modals. Language Acquisition 21.1: 103–117.

222

References

Cournane, Ailís 2017. In defense of the child innovator. In Éric Mathieu & Robert Truswell (eds.), Micro Change and Macro Change in Diachronic Syntax, 10–24. Oxford: Oxford University Press. Croft, William 1991. The evolution of negation. Journal of Linguistics 27: 1–27. Croft, William. 2001. Radical Construction Grammar: Syntactic Theory in Typological Perspective. Oxford: Oxford University Press. Crowley, Terry 1983. Uradhi. In R.M.W. Dixon & Barry Blake (eds.), Handbook of Australian Languages 3, 306–428. Amsterdam: John Benjamins. Crowley, Terry 1992. An Introduction to Historical Linguistics. Auckland: Oxford University Press. Culicover, Peter 2021. Language Change, Variation, and Universals. Oxford: Oxford University Press. Curtius, Georg 1870. Bermerkungen über die Tragweite der Lautgesetze. In Berichte der philologisch-historischen Classe der Königlich-Sächsischen Gesellschaft der Wissenschafter, vol. 22, 1–39. Leipzig: Hirzel. Curtius, Georg 1885. Zur Kritik der neuesten Sprachforschung. Leipzig: Hirzel. Dahl, Östen 1979. Typology of sentence negation. Linguistics 17: 79–106. Dahl, Östen 1985. Tense and Aspect Systems. Oxford: Blackwell. Dahl, Östen 1999. Grammaticalization and the Life-Cycles of Constructions. Unpublished manuscript, University of Stockholm. Dahl, Östen 2001. Inflationary effects in language and elsewhere. In Joan Bybee & Paul Hopper (eds.), Frequency and the Emergence of Linguistic Structure, 471–480. Philadelphia: John Benjamins. Dahl, Östen 2004. The Growth and Maintenance of Linguistic Complexity. Amsterdam: John Benjamins. d’Ardenne, S. 1977. The Katherine Group. Paris: Les Belles Lettres. Davison, Alice 2006. Correlative Clause Features in Sanskrit and Hindi/Urdu. DIGS Paper. www.uiowa.edu/%7Elinguist/faculty/davison/DIGSTXT.pdf. Dayley, Jon P. 1985. Tzutujil Grammar. Berkeley: University of California Press. Degand, Liesbeth and Anne-Marie Simon-Vandenbergen 2011. Introduction: Grammaticalization and (inter)subjectification of discourse markers. Linguistics 49.2: 287–294. Delbrück, Berthold 1919. Einleitung in das Studium der Indogermanischen Sprachen, 6th ed. Leipzig: Breitkopf & Härtel. Denison, David 1993. English Historical Syntax. London: Longman. Deo, Ashwini 2015. Diachronic semantics. Annual Review of Linguistics 1.1: 179–197. Detges, Ulrich 2000. Time and truth: The grammaticalization of resultatives and perfects within a theory of subjectification. Studies in Language 24.2: 345–377. Detges, Ulrich 2018. Strong pronouns in Modern Spoken French. Linguistics 56.5: 1059–1098. Detges, Ulrich 2020. Future markers in Western Romance. Journal of Historical Pragmatics 21.2: 290–315. Devos, Maud, Michael Kasombo Tshibanda, & Johan van der Auwera 2010. Jespersen cycles in Kanincin: Double, triple and maybe even quadruple negation. Africana Linguistica 1: 155–181. De Wit, Astrid 2017. The Present Perfective Paradox Across Languages. Oxford: Oxford University Press.

References

223

De Wit, Astrid, Adeline Patard & Frank Brisard 2013. A contrastive analysis of the present progressive in French and English. Studies in Language 37.4: 846–879. Diem, Werner 2014. Negation in Arabic. Wiesbaden: Harrassowitz. Diessel, Holger 1999a. Demonstratives: Form, Function, and Grammaticalization. Amsterdam: John Benjamins. Diessel, Holger 1999b. The morphosyntax of demonstratives in synchrony and diachrony. Linguistic Typology 3: 1–49. Diessel, Holger 2019. The Grammar Network. How Linguistic Structure is Shaped by Language Use. Cambridge: Cambridge University Press. Dimmendaal, Gerrit 2005. Nilo-saharan. In Bedilu Wakjira, Ronny Meyer, Yvonne Treis & Zelealem Leyew (eds.), The Oxford Handbook of Ethiopian Languages. Oxford: Oxford University Press. Dimmendaal, Gerrit 2011. Historical Linguistics and the Comparative Study of African Languages. Amsterdam: John Benjamins. Domingue, Nicole 1977. Middle English: Another creole? Journal of Creole Studies 1: 89–100. Donohue, Mark 1999. A Grammar of Tukang Besi. Berlin: Mouton de Gruyter. Drinka, Bridget 2017. Language Contact in Europe: The Periphrastic Perfect through History. Cambridge: Cambridge University Press. Dryer, Matthew 2013a. Definite articles. In Matthew Dryer & Martin Haspelmath (eds.), The World Atlas of Language Structures Online. Leipzig: Max Planck Institute for Evolutionary Anthropology. http://wals.info/chapter/37. Dryer, Matthew 2013b. Expression of pronominal subjects. In Matthew Dryer & Martin Haspelmath (eds.), The World Atlas of Language Structures Online. Leipzig: Max Planck Institute for Evolutionary Anthropology. http://wals.info/chapter/101. Dryer, Matthew 2013c. Position of polar question particles. In Matthew Dryer & Martin Haspelmath (eds.), The World Atlas of Language Structures Online. Leipzig: Max Planck Institute for Evolutionary Anthropology. http://wals.info/chapter/92. Dryer, Matthew 2013d. Prefixing vs. suffixing in inflectional morphology. In Matthew Dryer & Martin Haspelmath (eds.), The World Atlas of Language Structures Online. Leipzig: Max Planck Institute for Evolutionary Anthropology. http://wals.info/chapter/26. Du Ponceau, Peter 1819. Report of the corresponding secretary to the committee, of his progress in the investigation committed to him of the general character and forms of the languages of the American Indians: Read, 12th Jan. 1819. Transactions of the Historical & Literary Committee of the American Philosophical Society, Held at Philadelphia, for Promoting Useful Knowledge 1: xvii–xlvi. Easterday, Shelece, Matthew Stave, Marc Allassonnière-Tang & Frank Seifart 2021. Syllable complexity and morphological synthesis: A well-motivated positive complexity correlation across subdomains. Frontiers in Psycholology 12: 1–18. Eckardt, Regine & George Walkden. 2022. A particle-like use of hwæþer: Wisdom’s questions in Boethius. In Xabier Artiagoitia, Arantzazu Elordieta & Sergio Monforte (eds.), Discourse Particles, 41–64. Amsterdam: John Benjamins. Eco, Umberto 1995. The Search for the Perfect Language. Oxford: Blackwell. Edwards, Malcolm 2006. Pronouns, agreement and focus in Egyptian Arabic. SOAS Working Papers in Linguistics 14: 51–62. Egedi, Barbara 2014. The DP cycle in Hungarian. In Katalin É. Kiss (ed.), The Evolution of Functional Left Peripheries in Hungarian Syntax, 56–82. Oxford: Oxford University Press.

224

References

Elsness, Johan 1997. The Perfect and the Preterite in Contemporary and Earlier English. Berlin: DeGruyter Mouton. Emonds, Joseph & Jan Terje Faarlund 2014. English: The Language of the Vikings. Olomouc: Palacky University. Epstein, Richard 1995. The later stages in the development of the definite article: Evidence from French. In Henning Andersen (ed.), Historical Linguistics 1993, 159–175. Amsterdam: John Benjamins. Etter, Annemarie 1985. Die Fragesätze im Rigveda. Berlin: Mouton de Gruyter. EyÞórrson, Tolli 1995. Verbal Syntax in the Early Germanic Languages. Cornell PhD. Ezard, Bryan. 1997. A Grammar of Tawala, an Austronesian Language of the Milne Bay Area, Papua New Guinea. Canberra: Australian National University. Faarlund, Jan Terje 2004. The Syntax of Old Norse. Oxford: Oxford University Press. Fairbanks, Gordon 1977. Case inflections in Indo-European. Journal of Indo-European Studies 5: 101–132. Farquharson, Joseph 2013. Jamaican. In Susanne Maria Michaelis, Philippe Maurer, Martin Haspelmath & Magnus Huber (eds.), English-based and Dutch-based Languages, 81–91. Oxford: Oxford University Press. Fedriani, Chiara & Piera Molinelli 2020. Functional expansions of temporal adverbs and discursive connectives: From Latin tum, tunc, dumque to Old Italian dunque. Journal of Historical Pragmatics 21.2: 182–207. Ferrell, Raleigh 1982. Paiwan Dictionary. Canberra: Pacific Linguistics. Filppula, Markku, Juhani Klemola & Heli Paulasto 2008. English and Celtic in Contact. London: Routledge. Filppula, Markku, Juhani Klemola & Heli Pitkänen (eds.) 2002. The Celtic Roots of English. Joensuu: University of Joensuu. Finegan, Ed & Douglas Biber 2001. Register variation and social dialect variation. In P. Eckert & J.R. Rickford (eds.), Style and Sociolinguistic Variation, 235–267. Cambridge: Cambridge University Press. Fischer, Olga 2003. The development of the modals in English. In Maurizio Gotti (ed.), English Modality in Context, 17–32. Bern: Peter Lang. Fischer, Susan 1996. The role of agreement and auxiliaries in sign language. Lingua 98.1–3: 103–119. Flathe, Philipp Jacob 1803. Neue italienische Sprachlehre für die Deutschen. Leipzig. Fonseca-Greber, Bonnibeth 2000. The Change from Pronoun to Clitic to Prefix and the Rise of Null Subjects in Spoken Swiss French. University of Arizona dissertation. Fortescue, Michael. 1984. West Greenlandic (Croom Helm Descriptive Grammars). London: Croom Helm. Franzén, Torsten 1939. Etude sur la syntaxe des pronoms personnels sujets en ancien français. Uppsala: Almqvist & Wiksells Boktryckeri. Fromkin, Victoria, Robert Rodman & Nina Hyams 2019. An Introduction to Language, 11th ed. Boston: Cengage. Fryd, M. 1998. Present perfect et datation: une dérive aoristique. Cahiers Chronos 2: 29–50. Fuß, Eric 2005. The Rise of Agreement. Amsterdam: John Benjamins. Fulk, R.D., Robert Bjork & John Niles (eds.) 2008. Klaeber’s Beowulf. Toronto: University of Toronto Press. Gabelentz, Georg von der 2016[1891; 1901]. Die Sprachwissenschaft: ihre Aufgaben, Methoden und bisherigen Ergebnisse. Herausgegeben von Manfred Ringmacher and

References

225

James McElvenny. Berlin: Language Science Press. http://langsci-press.org/catalog/ book/97 Gardiner, Alan H. 1904. The word . Zeitschrift für Ägyptische Sprache und Altertumskunde 41: 130–135. Gardiner, Alan H. 1957. Egyptian Grammar. Oxford: Oxford University Press. Gelderen, Elly van 2000. A History of English Reflexive Pronouns: Person, Self, and Interpretability. Amsterdam: John Benjamins. Gelderen, Elly van 2004a. Grammaticalization as Economy. Amsterdam: John Benjamins. Gelderen, Elly van 2004b. Economy, innovation, and prescriptivism: From spec to head and head to head. The Journal of Comparative Germanic Linguistics 7.1: 59–98. Gelderen, Elly van 2007a. The definiteness cycle in Germanic. Journal of Germanic Linguistics 19.4: 275–305. Gelderen, Elly van 2007b. Accelerated grammaticalization in the Peterborough Chronicle. In Alex Bergs & Janne Skaffari (eds.), The Language of the Peterborough Chronicle, 93–110. Frankfurt am Main: Peter Lang. Gelderen, Elly van 2008a. The negative cycle. Linguistic Typology 12.2: 195–243. Gelderen, Elly van 2008b. Cycles of negation in Athabaskan. Working Papers in Athabaskan Languages 7: 49–64. Gelderen, Elly van 2009a. Cyclical Change. Amsterdam: John Benjamins. Gelderen, Elly van 2009b. Renewal in the left periphery: Economy and the complementizer layer. Transactions of the Philological Society 107.2 (2009): 131–195. Gelderen, Elly van 2011. The Linguistic Cycle. Oxford: Oxford University Press. Gelderen, Elly van 2013. The linguistic cycle and the language faculty. Language and Linguistics Compass 7: 233–250. Gelderen, Elly van 2016. Cyclical Change Continued. Amsterdam: John Benjamins. Gelderen, Elly van 2018. The Diachrony of Verb Meaning: Aspect and Argument Structure. New York: Routledge. Gelderen, Elly van 2019. Cyclical change and problems of projection. In Anne Breitbarth, Miriam Bouzouita, Lieven Danckaert & Elisabeth Witzenhausen (eds.), Cycles in Language Change, 13–32. Oxford: Oxford University Press. Gelderen, Elly van 2020. A new modal cycle. In Andreas Trotzke & George Walkden (eds.), Nordlyd 44.1: 3–17, CGSW 34: Proceedings of the 34th Comparative Germanic Syntax Workshop Held in Konstanz in 2019. University of Konstanz. Published at the University of Tromsø – The Arctic University of Norway https://septentrio.uit. no/index.php/nordlyd/article/view/5209 Gelderen, Elly van 2022. Third Factors in Syntactic Variation and Change. Cambridge: Cambridge University Press. Gergel, Remus 2009. Rather – On a modal cycle. In Elly van Gelderen (ed.), Cyclical Change, 243–264. Amsterdam: John Benjamins. Gergel, Remus 2016. Modality and gradation: Comparing the sequel of developments in ‘rather’ and ‘eher’. In Elly van Gelderen (ed.), The Linguistic Cycle Continued, 319–350. Amsterdam: John Benjamins. Geurts, Bart 2000. Explaining grammaticalization (the standard way). Linguistics 38.4: 781–788. Ghezzi, Chiara & Piera Molinelli 2020. Connectives and cyclicity. Journal of Historical Pragmatics 21.2: 208–235.

226

References

Gianollo, Chiara 2016. The Latin system of negation at the syntax-semantics interface. Rivista di Grammatica Generativa 38: 115–135. Gil, David 2002. The prefixes di- and N- in Malay/Indonesian dialects. In Faye Wouk & Malcolm Ross (eds.), The History and Typology of Western Austronesian Voice Systems, 241–283. Canberra: Pacific Linguistics. Giusti, Giuliana 2001. The birth of a functional category: From latin ILLE to the romance article and personal pronoun. In Current Studies in Italian Syntax, 157–171. Leiden: Brill. Givón, Talmy 1971. Historical syntax and synchronic morphology. Chicago Linguistic Society Proceedings 7: 394–415. Givón, Talmy 1973. The time-axis phenomenon. Language 49.4: 890–925. Givón, Talmy 1976. Topic, pronoun, and grammatical agreement. In Charles Li (ed.), Subject and Topic, 151–188. New York: Academic Press. Givón, Talmy 1978. Negation in language. In Peter Cole (ed.), Syntax & Semantics 9, 69–112. New York: Academic Press. Givón, Talmy 1979. On Understanding Grammar. New York: Academic Press. Givón, Talmy 1984. Syntax: A Functional-typological Introduction. Amsterdam: John Benjamins. Givón, Talmy 2000. Internal reconstruction; as method, as theory. In Spike Gildea (ed.), Reconstructing Grammar, 107–159. Amsterdam: John Benjamins. Givón, Talmy 2016a. The diachrony of pronominal agreement. In Elly van Gelderen (ed.), Cyclical Change Continued, 251–286. Amsterdam: John Benjamins. Givón, Talmy 2016b. Nominalization and re-finitization. In Claudine Chamoreau & Zarina Estrada-Fernández (eds.), Finiteness and Nominalization, 271–296. Amsterdam: John Benjamins. Givón, Talmy 2018. On Understanding Grammar, revised ed. Amsterdam: John Benjamins. Givón, Talmy 2021. The Life Cycle of Adpositions. Amsterdam: John Benjamins. Goldberg, Adele 1995. Constructions: A Construction Grammar Approach to Argument Structure. Chicago: University of Chicago Press. Gómez Vázquez, Diana & Jaume Mateu 2022. How to obtain a copula from a process verb: Insights from the event structure of posture verbs. Glossa 7.1: 1–40. Gordon, E.V. 1956. Introduction to Old Norse. Oxford: Clarendon Press. Gordon, Lynn 1980. Maricopa Morphology and Syntax. UCLA dissertation. Gordon, Lynn 1986. Maricopa Morphology and Syntax. Berkeley: University of California Press. Goto, Nobu & Toru Ishii 2019. The principle of determinacy and its implications for MERGE. In Proceedings of the 12th GLOW in Asia & 21st SICOGG, 91–110. http:// sicogg.or.kr/GLOW-Asia-12-2019/proceedings/ Greenberg, Joseph 1960. A quantitative approach to the morphological typology of language. International Journal of American Linguistics 26.3: 178–194. Greenberg, Joseph 1963. The Languages of Africa. Bloomington/The Hague: Indiana University Press, Research Center in Anthropology, Folklore and Linguistics/ Mouton. Greenberg, Joseph 1966. Some universals of grammar with particular reference to the order of meaningful elements. In Joseph Greenberg (ed.), Universals of Language, 73–113. Cambridge, MA: MIT Press.

References

227

Greenberg, Joseph 1978. How does a language acquire gender markers? In Greenberg et al. (eds.), Universals of Human Language 3, 47–82. Stanford: Stanford University Press. Greenberg, Marc. 2006. A Short Reference Grammar of Standard Slovene. SEELRC Reference Grammar Network. Duke University/University of North Carolina, Chapel Hill: SEELRC. Greenough, J.B., et al. (eds.) 1931. Allen and Greenough’s New Latin Grammar. Boston: Ginn. Grestenberger, Laura 2020. The diachrony of participles in the (pre)history of Greek and Hittite. Diachronica 37.2: 215–263. Grüter, Therese 2006a. Object Clitics and Null Objects in the Acquisition of French. McGill PhD. Grüter, Therese 2006b. Object (Clitic) omission in L2 French. In Mary Grantham O’Brien, Christine Shea & John Archibald (eds.), GASLA Proceedings, 63–71. Somerville, MA: Cascadilla. Guilfoyle, Eithne, Henrietta Hung & Lisa Travis 1992. Spec of IP and spec of VP: Two subjects in Austronesian languages. Natural Language and Linguistic Theory 10: 375–414. Gul, Ahmed 2009. Grammaticalization of thoRi. Ms. Guðmundsson, Helgi 1972. The Pronominal Dual in Icelandic. Reykjavík: University of Iceland. Gzella, Holger 2015. A Cultural History of Aramaic. Leiden: Brill. Haase, Martin 1994. Tense and aspect in Basque. In Rolf Thieroff & Joachim Ballweg (eds.), Tense Systems in European Languages, 279–292. Tübingen: Niemeyer. Haeberli, Eric & Tabea Ihsane 2016. Revisiting the loss of verb movement in the history of English. Natural Language & Linguistic Theory 34.2: 497–542. Hagège, Claude 1993. The Language Builder: An Essay on the Human Signature in Linguistic Morphogenesis. Amsterdam: John Benjamins. Haig, Geoffrey 2018a. The grammaticalization of object pronouns. Linguistics 56.4: 781–818. Haig, Geoffrey 2018b. Grammaticalization and inflectionalization in Iranian. In Heiko Narrog & Bernd Heine (eds.), Grammaticalization from a Typological Perspective, 57–78. Oxford: Oxford University Press. Halm, Tamás 2020. Grammaticalization without feature economy: Evidence from the voice cycle in Hungarian. Diachronica 37.1: 1–42. Handel, Zev 2012. Valence-changing prefixes and voicing alternation in Old Chinese and Proto-Sino-Tibetan: Reconstructing *s- and *N- Prefixes. Language and Linguistics 13.1: 61–82. Hanham, Alison (ed.) 1975. The Cely Letters 1472–1488. Oxford: Oxford University Press. Hansen, Björn 2005. Modality in Slavonic languages. In Björn Hansen & P. Karlík (eds.), Modality in Slavonic Languages. New Perspectives. München: Sagner. Hansen, Maj-Britt Mosegaard 2014. Cyclicity in semantic/pragmatic change: Old French ja between Latin IAM and Modern French déjà. In Chiara Ghezzi & Piera Molinelli (eds.), Discourse and Pragmatic Markers from Latin to Romance, 139–165. Oxford: Oxford University Press. Hansen, Maj-Britt Mosegaard 2020. Introduction: The role of pragmatics in cyclic language change. Journal of Historical Pragmatics 21.2: 165–181.

228

References

Hansen, Maj-Britt Mosegaard & Jacqueline Visconti 2009. On the diachrony of ‘reinforced’ negation in French and Italian. In Corinne Rossari, Claudia Ricci & Adriana Spiridon (eds.), Grammaticalization and Pragmatics, 137–171. Leiden: Brill. Hargus, Sharon 2002. Negation in Kwadacha (Ft. Ware Sekani). Alaska Native Language Center Working Papers 2: 106–120. Harris, Martin 1977. ‘Demonstratives’, ‘articles’ and ‘third person pronouns’ in French: Changes in progress. Zeitschrift für romanische Philologie 93: 249–261. Harris, Randy 2021. The Linguistic Wars, 2nd ed. Oxford: Oxford University Press. Harrison, Sheldon 1978. Transitive marking in Micronesian languages. In S.A. Wurm & Lois Carrington (eds.), Second International Conference of Austronesian Linguistics Proceedings, 1067–1127. Canberra, Australia: Pacific Linguistics. Haspelmath, Martin 1990. The grammaticization of passive morphology. Studies in Language 14.1: 25–71. Haspelmath, Martin 1998. Does grammaticalization need reanalysis? Studies in Language 22.2: 315–351. Haspelmath, Martin 1999. Why is grammaticalization irreversible? Linguistics 37.6: 1043–1068. Haspelmath, Martin 2000. The relevance of extravagance: A reply to Bart Geurts. Linguistics 38.4: 789–798. Haspelmath, Martin 2007. Pre-established categories don’t exist: Consequences for language description and typology. Linguistic Typology 11.1: 119–132. Haspelmath, Martin 2011a. The indeterminacy of word segmentation and the nature of morphology and syntax. Folia Linguistica 45.1: 31–80. Haspelmath, Martin 2011b. The gradual coalescence into “words” in grammaticalization. In Heiko Narrog & Bernd Heine (eds.), The Oxford Handbook of Grammaticalization, 342–355. Oxford: Oxford University Press. Haspelmath, Martin 2013. Argument indexing: A conceptual framework for the syntactic status of bound person forms. In Dik Bakker & Martin Haspelmath (eds.), Language Across Boundaries: Studies in Memory of Anna Siewierska, 197–226. Berlin: Mouton. Haspelmath, Martin 2018. Revisiting the anasynthetic spiral. In Bernd Heine & Heiko Narrog (eds.), Grammaticalization from a Typological Perspective, 97–115. Oxford: Oxford University Press. Haspelmath, Martin 2022. Types of Clitics in the World’s Languages. https://ling.auf.net/ lingbuzz/007071 Havers, Wilhelm. 1931. Handbuch der erklärenden Syntax. Heidelberg: Winter. Heaney, Seamus. 1999. Beowulf. New York: Norton. Heath, Jeffrey 1978. Linguistic Diffusion in Arnhem Land. Canberra: Australian Institute of Aboriginal Studies. Heath, Jeffrey 1998. Hermit crabs: Formal renewal of morphology by phonologically mediated affix substitution. Language 74.4: 728–759. Hegedüs, Veronika 2014. The cyclical development of Ps in Hungarian. In Katalin É. Kiss (ed), The Evolution of Functional Left Peripheries in Hungarian Syntax, 122–147. Oxford: Oxford University Press. Heine, Bernd, Gunther Kaltenböck, Tania Kuteva & Haiping Long 2021. The Rise of Discourse Markers. Cambridge: Cambridge University Press. Heine, Bernd, Ulrike Claudi & Friederike Hünnemeyer 1991. Grammaticalization: A Conceptual Framework. Chicago: University of Chicago Press.

References

229

Heine, Bernd & Tania Kuteva 2002. World Lexicon of Grammaticalization. Cambridge: Cambridge University Press. Heine, Bernd & Tania Kuteva 2005. Language Contact and Grammatical Change. Cambridge: Cambridge University Press. Heine, Bernd & Tania Kuteva 2006. The Changing Languages of Europe. Oxford: Oxford University Press. Heine, Bernd & Mechthild Reh 1984. Grammaticalization and Reanalysis in African Languages. Hamburg: Buske Verlag. Heine, Bernd & Kyung-An Song 2010. On the genesis of personal pronouns: Some conceptual sources. Language and Cognition 2.1: 117–147. Heine, Bernd & Kyung-An Song 2011. On the grammaticalization of personal pronouns. Journal of Linguistics 47: 587–630. Hengeveld, Kees 1992. Non-verbal Predication. Berlin: Mouton de Gruyter. Henry, Victor 1883. Etude sur analogie en général et sur les formations analogiques de la langue grecque. Paris Doctoral thesis. Hetzron, Robert. 1976. The Agaw languages. Afroasiatic Linguistics 3: 1–45. Hilpert, Martin 2021. Ten Lectures on Diachronic Construction Grammar. https://library. oapen.org/handle/20.500.12657/50782 Hintze, Fritz 1947. Die Haupttendenzen der ägyptischen Sprachentwicklung. Zeitschrift für Phonetik und allgemeine Sprachwissenschaft 1.3: 85–108. Hock, Hans 1996. Who’s on first? Toward a prosodic account of PS clitics. In Aaron Halpern & Arnold Zwicky (eds.), Approaching Second. Second Position Clitics and Related Phenomena, 199–270. Stanford: CSLI. Hodge, Carleton 1970. The linguistic cycle. Linguistic Sciences 13: 1–7. Hoeksema, Jack 2009. Jespersen recycled. In Elly van Gelderen (ed.), Cyclical Change, 15–34. Amsterdam: John Benjamins. Holes, C. 1990. Gulf Arabic. London: Routledge. Honti, Lásló 1997. Die Negation im Uralischen I-III. Linguistica Uralica 2: 81–96; 161–176; 241–252. Hopper, Mike & Elizabeth Traugott 2003. Grammaticalization. Cambridge: Cambridge University Press. Horne Tooke, John 1786–1805. The Diversions of Purley. London: William Tegg and Co. Hróarsdóttir, Thorbjörg 2000. Word Order Change in Icelandic. From OV to VO. Amsterdam: John Benjamins. Huang, James 2014. On syntactic analyticity and parametric theory. In Audrey Li, Andrew Simpson & Wei-Tian Dylan Tsai (eds.), Chinese Syntax in a Cross-Linguistic Perspective, 1–48. Oxford: Oxford University Press. Hubers, Frank & Helen de Hoop 2013. The effect of prescriptivism on comparative markers in spoken Dutch. Linguistics in The Netherlands 30: 89–101. Humboldt, Wilhelm von 1822. Über die Entstehung der grammatischen Formen und ihren Einfluss auf die Ideenentwicklung. Abhandlungen der Akademie der Wissenschaften zu Berlin. Reprinted in 1972. Darmstadt: Wissenschaftliche Buchgesellschaft. Humboldt, Wilhelm von 1827–9. Über die Verschiedenheiten des menschlichen Sprachbaues. Reprinted in 1963 as Schriften zur Sprachphilosophie 3 (pp. 368–756). Darmstadt: Wissenschaftliche Buchgesellschaft. Humboldt, Wilhelm von 1836. Über die Verschiedenheit des menschlichen Sprachbaus und seinen Einfluss auf die geistige Entwicklung des Menschengeschlechts. Berlin: Königlich-Preussische Akademie der Wissenschafte.

230

References

Hundt, Marianne 2004. Animacy, agentivity, and the spread of the progressive in Modern English. English Language and Linguistics 8.1: 47–69. Hyslop, Catriona 2001. The Lolovoli Dialect of the North-East Ambae Language, Vanuatu. Canberra: Research School of Pacific and Asian Studies, Australian National University. Igartua, Iván 2015. From cumulative to seperative exponence in inflection. Language 91: 676–722. Ingham, Richard. 2006. Negative concord and the loss of the negative particle ne in late Middle English. Studia Anglica Posnaniensia 42: 77–96. Ingham, Richard. 2013. Negation in the history of English. In David Willis, Christopher Lucas & Anne Breitbarth (eds.), The History of Negation in the Languages of Europe and the Mediterranean 1: Case Studies, 119–150. Oxford: Oxford University Press. Inglese, Guglielmo to appear. The Rise of Middle Voice Systems. Diachronica. Ishiyama, Osamu 2019. Diachrony of Personal Pronouns in Japanese. Amsterdam: John Benjamins. Jackendoff, Ray 1977. X[bar] Syntax: A Study of Phrase Structure. Cambridge, MA: MIT Press. Jacques, Guillaume 2015. The origin of the causative prefix in Rgyalrong languages and its implication for proto-Sino-Tibetan reconstruction. Folia Linguistica 36.1: 165–198. Jäger, Agnes 2010. Der Komparativzyklus und die Position der Vergleichspartikeln. Linguistische Berichte 224: 467–493. Jäger, Agnes 2021. The comparative cycle in crosslinguistic perspective. Linguistic Typology at the Crossroads 1.1: 124–178. Jakubowicz, Celia, Natascha Müller, Beate Riemer & Catherine Rigaut 1997. The case of subject and object omissions in French and German. In Proceedings of the 21st BUCLD, 331–342. Somerville: Cascadilla. Janzen, Terry & Barbara Shaffer 2002. Gesture as the substrate in the process of ASL grammaticization. In R.P. Meier, K. Cormier & D. Quinto-Pozos (eds.), Modality and Structure in Signed and Spoken Languages, 199–223. Cambridge: Cambridge University Press. Jarad, N. I. 2015. From bodily posture to progressive aspect marker. Lingua Posnaniensis 1. https://doi.org/10.1515/linpo-2015-0005. Jędrzejowski, Lukasz 2016. All you need is “need”. On the verbal NPI cycle in the history of German. In Elly van Gelderen (ed.), Cyclical Change Continued, 351–394. Amsterdam: John Benjamins. Jelinek, Eloise 1984. Empty categories, case, and configurationality. Natural Language and Linguistic Theory 2: 39–76. Jendraschek, Gerd 2003. When old paradigms die, new ones are formed. Journal of Universal Language 4: 27–59. Jespersen, Otto 1917 [1966]. Negation in English and Other Languages, 2nd ed. Copenhagen: A.F. Høst. Jespersen, Otto 1922. Language. London: Allen & Unwin. Jespersen, Otto 1940. Modern English Grammar V. London: Allen & Unwin. Jespersen, Otto 1941. Efficiency in Language Change. Copenhagen: Munksgaard. Jetté, Jules, Eliza Jones & James Kari 2000. Koyukon Athabaskan Dictionary. Fairbanks: ANLC.

References

231

Ji, Ming Ping 2007. The Left-Periphery in Chinese. ASU dissertation. Johannessen, Janne 2006. Just any pronoun anywhere? Pronouns and ‘new demonstratives’ in Norwegian. In Torgrim Solstad, Alte Grønn & Dag Haug (eds.), A Festschrift for Kjell Johan Sæbø, 91–106. Oslo: University of Oslo. Jones, William 1807. Works of Sir William Jones III. London. Kahr, Joan 1975. Adpositions and locationals: Typology and diachronic development. Working Papers on Language Universals 19: 21–54. Kahr, Joan 1976. The reneweal of case morphology: Sources and constraints. Working Papers on Language Universals 20: 107–151. Kari, James 1990. Ahtna Dictionary. Fairbanks: Alaska Native Language Center. Kari, James 1993. Diversity in morpheme order in several Alaskan Athabaskan languages. Berkeley Linguistic Society Proceedings: 50–56. Karlgren, Bernhard 1920. Le Proto-Chinois, langue flexionnelle. Journal Asiatique (avril–juin): 205–232. Katz, Aya 1996. Cyclical Grammaticalization and the Cognitive Link between Pronoun and Copula. Rice dissertation. Katz, Jerrold & Jerry Fodor 1963. The structure of a semantic theory. Language 39.2: 170–210. Kellogg, S.H. 1938 [1893]. A Grammar of the Hindi Language. London: Kegan Paul. Kendall, Martha 1972. Selected Problems in Yavapai Syntax. Indiana University PhD. Kerke, Simon van der 1996. Affix Order and Interpretation in Bolivian Quechua. Meppel: Krips Repro. Kiparsky, Paul 2011. Grammaticalization as optimization. In Diane Jonas, John Whitman & Andrew Garrett (eds.), Grammatical Change: Origins, Nature, Outcomes, 14–51. Oxford: Oxford University Press. https://web.stanford.edu/~kiparsky/Papers/yalegrammaticalization.pdf Kiparsky, Paul & Cleo Condoravdi 2006. Tracking Jespersen’s cycle. In Mark Janse, Brian Joseph & A. Ralli (eds.), Proceedings of the 2nd International Conference of Modern Greek Dialects and Linguistic Theory, 172–197. Mytilene: Doukas. Kirby, Ian 2022. Wrong-side Affixation: Univerbation of Agreement Morphemes. Ms. Kiss, Katalin É. 2017. The rise and fall of Hungarian complex tenses. In Eric Mathieu & Robert Truswell (eds.), Micro-change and Macro-change in Diachronic Syntax, 61–74. Oxford: Oxford University Press. Kiss, Katalin É. & Nikolett Mus 2022. The reflexive cycle: From reflexive to personal pronoun in Uralic. Journal of Uralic Linguistics 1.1: 43–66. Klamer, Marian 1997. Spelling out clitics in Kambera. Linguistics 35: 895–927. Klamer, Marian 1998. A Grammar of Kambera. Berlin: Mouton de Gruyter. Klima, Edward 1965. Studies in Diachronic Transformational Syntax. Harvard PhD. Klippi, Carita 2004. La Notion d’analogie dans les thèses de doctorat de Victor Henry. In Christian Puech (ed.), Charnière Des XIXe Et XXe Siecles Victor Henry (1850–1907). Leuven: Peeters. Koh, Ann 1990. Topics in Colloquial Malay. University of Melbourne PhD. Kóna-Tas, András 1998. The reconstruction of Proto-Turkic and the genetic question. In Lars Johanson & Éva Àgnes Csató (eds.), The Turkic Languages, 67–80. London: Routledge. König, Christa 2008. Case in Africa. Oxford: Oxford University Press. Koopman, Hilda 1984. The Syntax of Verbs: From Verb Movement Rules in the Kru Languages to Universal Grammar. Dordrecht: Foris Publications.

232

References

Korostovcev, M.A. 1963. Vvedenie v egipetskuju filologiju. Moscow: Izdateljstvo Vostočnoj Literatury. Kortlandt, Frederik 1983. Demonstrative pronouns in Balto-Slavic, Armenian, and Tocharian. Studies in Slavic and General Linguistics 3: 311–322. Kozinsky, Isaac, Vladimir Nedjalkov & Maria Polinskaja. 1988. Antipassive in chukchee. In Masayoshi Shibatani (ed.), Passive and Voice, 651–706. Amsterdam: John Benjamins. Kranich, Svenja 2010. The Progressive in Modern English: A Corpus-Based Study of Grammaticalization and Related Changes. Leiden: Brill. Kranich, Svenja 2021. Decline and loss in the modal domain in recent English. In Svenja Kranich & Tine Breban (eds.), Lost in Change, 261–290. Amsterdam: John Benjamins. Krauss, Michael 1969. On the classification in the athabascan, eyak, and tlingit verb. Supplement to IJAL 35.4. Kulikov, Leonid 2006. Case systems in a diachronic perspective. In Leonid Kulikov et al. (eds.), Case, Valency and Transitivity, 23–47. Amsterdam: John Benjamins. Kulikov, Leonid 2009. Evolution of case systems. In Andrej Malchukov & Andrew Spencer (eds.), Handbook of Case, 439–457. Oxford: Oxford University Press. Künnap, Ago. 1999. Kamass. München: Lincom. Kuryłowicz, Jerzy 1964. The Inflectional Categories of Indo-European. Wiesbaden: Harrassowitz. Kuteva, Tanya, Bernd Heine, Bo Hong, Haiping Long, Heiko Narrog & Seongha Rhee (eds.) 2019. World Lexicon of Grammaticalization. Cambridge: Cambridge University Press. Kuular Klara B. 2007 Reciprocals, sociatives, comitatives, and assistives in Tuvan. In Vladimir P. Nedjalkov (ed.), Reciprocal Constructions, 1163–1229. Amsterdam: John Benjamins Labarge, Robert 2016. The grammaticalization of 要 Yào: The future cycle from Archaic Chinese to modern mandarin. In Elly van Gelderen (ed.), Cyclical Change Continued, 395–418. Amsterdam: John Benjamins. Lagercrantz, Eliel 1929. Sprachlehre des Nordlappischen. Oslo: Oslo Etnografiske Museum. Lahiri, Aditi 2000. Hierarchical restructuring in the creation of verbal morphology in Bengali and Germanic: Evidence from phonology. In Aditi Lahiri (ed.), Analogy, Levelling, Markedness, 71–123. Berlin: Mouton. Lambrecht, Knud 1981. Topic, Antitopic, and Verb Agreement in Non-Standard French. Amsterdam: John Benjamins. Lambrecht, Knud & Kevin Lemoine 1996. Vers une grammaire des compléments zéro en francais parlé. In Jean Chuquet et Marc Fryd (eds.), Travaux linguistiques de CERLICO 9, 279–309. Rennes: Presses universitaires de Rennes. Lam Chit Yu, Cherry 2022. Croft’s cycle in mandarin and cantonese throughout history and across varieties. In Ljuba Veselinova & Arja Hamari (eds.), The Negative Existential Cycle, 357–401. Berlin: Language Science Press. Lampp, Claire 2006. Negation in Modern Hindi-Urdu. UNC MA thesis. Langacker, Ronald 1977. Syntactic reanalysis. In Charles Li (ed.), Mechanics of Syntactic Change, 59–139. Austin: University of Texas Press. Langdon, Margaret 1968. The Proto-Yuman demonstrative system. Folia Linguistica 2: 61–81.

References

233

Langdon, Margaret 1970. A Grammar of Diegueño: The Mesa Grande Dialect. Berkeley: University of California Press. Langdon, Margaret 1977. Syntactic change and SOV structure: The Yuman case. In Charles N. Li (ed.), Mechanisms of Syntactic Change, 255–290. Austin: University of Texas Press. Larjavaara, Meri 2000. Présence ou absence de l’object. http://ethesis.helsinki.fi/julkaisut/ hum/romaa/vk/larjavaara/presence.pdf Larrivée, Pierre 2010. The pragmatic motics of the Jespersen cycle: Default, activation, and the history of negation in French. Lingua 120.9: 2240–2258. Larrivée, Pierre & Richard Ingham 2011. The Evolution of Negation: Beyond the Jespersen Cycle. Berlin: Mouton de Gruyter. Laury, Ritva 1991. Demonstratives in Interaction. Amsterdam: John Benjamins. Ledgeway, Adam 2012. From latin to romance: Configurationality, functional categories and head-marking. Transactions of the Philological Society 110.3: 422–442. Leech, Geoffrey 2003. Modality on the move: The English modal auxiliaries 1961–1992. In Roberta Facchinetti, Manfred Krug & Frank Palmer (eds.), Modality in Contemporary English, 233–240. Berlin: Mouton de Gruyter. Leech, Geoffrey 2011. The modals ARE declining: Reply to Neil Millar’s ‘Modal verbs in TIME: Frequency changes 1923–2006, International Journal of Corpus Linguistics 14:2, 191–220. International Journal of Corpus Linguistics 16(4): 547–564. Leech, Geoffrey, Marianne Hundt, Christian Mair & Nicholas Smith 2009. Change in Contemporary English. Cambridge: Cambridge University Press. Leer, Jeff 2000. The negative/irrealis category in athabaskan-eyak-tlingit. In Ted Fernald & Paul Platero (eds.), The Athabaskan Languages, 101–138. Oxford: Oxford University Press. Lefèvre, André 1877. Etudes de Linguistique et de Philologie. Paris: Ernest Leroux. www.google.com/books/edition/%C3%89tudes_de_linguistique_et_de_philologie/ N1xMAQAAMAAJ?hl=en&gbpv=1&dq=%22cycle+linguistique%22&pg=PA151& printsec=frontcover Lefèvre, André 1893. Les Races et les langues. Paris: Félix Alcan. Lefèvre, André 1894. Race and Language. London: Kegan, Paul, Trench, Trübner, & Co. Lehmann, Christian 1982 [2015]. Thoughts on Grammaticalization. 3rd ed. 2015. Berlin: Language Science Press. Lehmann, Winfred 1958. On the earlier stages of the Indo-European nominal inflection. Language 34: 179–202. Lehmann, Winfred 1967. A Reader in Nineteenth Century Historical Indo-European Linguistics. Indiana University Press. https://liberalarts.utexas.edu/lrc/resources/books/ reader/index.php Lehmann, Winfred 1976. From topic to subject in Indo-European. In Charles Li (ed.), Subject and Topic, 445–456. New York: Academic Press. Lehmann, Winfred 1993. Theoretical Bases of Indo-European Linguistics. New York: Routledge. Leibniz, Gottfried Wilhelm von 1768 Opera Omnia. Geneva: de Tournes. Leiss, Elisabeth 2000. Artikel und Aspekt. Berlin: Walter de Gruyter. Leuvensteijn, A. van 2002. Epistolaire aanspreekvormen in de correspondentie van Maria van Reigersberch. Tijdschrift voor Nederlandse Taal- en Letterkunde 118: 288–298.

234

References

Levshina, Natalia & David Lorenz 2022. Communicative efficiency and the principle of no synonymy. Language and Cognition 14: 1–26. Lewis, Diana 2000. Some Emergent Discourse Connectives in English. University of Oxford PhD. Lewis, G.L. 1967. Turkish Grammar. Oxford: Oxford University Press. Li, Charles & Sandra Thompson 1974. Co-verbs in mandarin Chinese: Verbs or prepositions? Journal of Chinese Linguistics 2.3: 257–278. Li, Charles & Sandra Thompson 1977. A mechanism for the development of copula morphemes. In Charles N. Li (ed.), Mechanisms of Syntactic Change, 419–444. Austin: University of Texas Press. Li, Charles & Sandra Thompson 1981. Mandarin Chinese: A Functional Reference Grammar. Berkeley: University of California Press. Li, Fang Kuei 1967. Chipewyan. Viking Fund Publications in Anthropology 6. New York: Viking Fund. Li, Yen-Hui Audrey 1990. Order and Constituency in Mandarin Chinese. Dordrecht: Kluwer. Lightfoot, David 1979. Principles of Diachronic Syntax. Cambridge: Cambridge University Press. Lightfoot, David 1999. The Development of Language. Oxford: Blackwell. Lightfoot, David 2006. How New Languages Emerge. Cambridge: Cambridge University Press. Lillehaugen, Danielle 2006. Expressing Location in Tlacolula Valley Zapotec. UCLA PhD. Lima, José Pinto de 2006. Zur Grammatikalisierung von dt. drohen und pg. ameaçar. In Jürgen Schmidt-Radefeldt (ed.), Portugiesisch kontrastiv gesehen und Anglizismen weltweit, 205–220. Frankfurt: Peter Lang. Lin, Nina Yuhsun 2002. A corpus-based study on the development of mandarin negative marker mei. In International Conference on Sino-Tibetan Languages and Linguistics 35 (ICSTLL). Arizona State University. Lloyd, Albert 1979. Anatomy of the Verb: The Gothic Verb as a Model for a Unified Theory of Aspect, Actional Types, and Verbal Velocity. Amsterdam: John Benjamins. Longobardi, Giuseppe 2001. Formal syntax, diachronic minimalism, and etymology: The history of French Chez. Linguistic Inquiry 32.2: 275–302. Loprieno, Antonio 1995. Ancient Egyptian. Cambridge: Cambridge University Press. Lorenz, David 2020. Converging variations and the emergence of horizontal link. In Lotte Dommerer & Elena Smirnova (eds.), Nodes and Networks in Diachronic Construction Grammar, 243–274. Amsterdam: John Benjamins. Los, Bettelou 2009. The consequences of the loss of verb-second in English. English Language and Linguistics 13.1: 97–125. Love, Robbie & Niall Curry 2021. Recent change in modality in informal spoken British English: 1990s–2010s. English Language and Linguistics 25.3: 537–562. Lucas, Christopher 2007. Jespersen’s cycle in Arabic qnd Berber. Transactions of the Philological Society 105.3: 398–431. Lucas, Christopher 2010. Negative-š in Palestinian (and Cairene) Arabic: Present and possible past. Brill’s Journal of Afroasiatic Languages and Linguistics 2.1: 165–201.

References

235

Lucas, Christopher 2020. Contact and the expression of negation. In Christopher Lucas & Stefano Manfredi (eds.), Arabic and Contact-Induced Change, 643–667. Berlin: Language Science Press. Lucas, Christopher & Elliott Lash 2010. Contact as catalyst: The case for Coptic influence in the development of Arabic negation. Journal of Linguistics 46.2: 379–413. Lyons, John 1999. Introduction to Theoretical Linguistics. Cambridge: Cambridge University Press. MacLeod, Morgan 2012. The Perfect in Old English and Old Saxon: The Synchronic and Diachronic Correspondence of form and Meaning. Cambridge PhD. www.repository. cam.ac.uk/bitstream/handle/1810/242374/I.pdf?sequence=1 Magistro, Giuseppe, Claudia Crocco & Anne Breitbarth 2022. Information structure and Jespersen’s cycle. In Nicholas Catasso, Marco Coniglio & Chiara De Bastiani (eds.), Language Change at the Interfaces: Intrasentential and Intersentential Phenomena, 35–60. Amsterdam: John Benjamins. Mailhammer, Robert & Elena Smirnova 2013. Incipient grammaticalisation: Sources of passive constructions in Old High German and Old English. In Gabriele Diewald, Leena Kahlas-Tarkka & Ilse Wischer (eds.), Comparative Studies in Early Germanic Languages, 41–70. Amsterdam: John Benjamins Mair, Christian 2001. Gerundial complements after begin and start. In Günter Rohdenburg & Britta Mondorf (eds.), Determinants of Grammatical Variation in English, 329–345. De Gruyter. Mansfield, John 2021 The word as a unit of internal predictability. Linguistics 59.6: 1427–1472. Margetts, Anna 1999. Valence and Transitivity in Saliba: An Oceanic Language of Papua New Guinea. Radboud University PhD dissertation. Martinet, André 1955. Economie des changements phonétiques. Bern: A. Francke. Martínez-Areta, Mikel 2013. Demonstratives and personal pronouns. In Mikel MartínezAreta (ed.), Basque and Proto-Basque, 283–321. Frankfurt am Main: Peter Lang. Marušič, Franc & Rok Žaucer 2016. The modal cycle vs. negation in Slovenian. In Franc Marušič & Rok Žaucer (eds.), Formal Studies in Slovenian Syntax, 167–191. Amsterdam: John Benjamins. McCawley, James 1971. Interpretive semantics meets frankenstein. Foundations of Language 7: 285–296. Reprinted in Grammar and Meaning, by James McCawley (1976), pp. 333–342. New York: Academic Press. McDonald, Brittany 2021. The Interactional Structure of Nominals: An Investigation of Paranouns. MA University of Calgary. https://prism.ucalgary.ca/bitstream/handle/ 1880/113709/ucalgary_2021_mcdonald_brittany.pdf?sequence=2&isAllowed=y McElvenny, James 2016. The Secret History of Grammaticalization. https://hiphilangsci. net/2016/04/28/the-secret-history-of-grammaticalization/ McFadden, Thomas 2015. Preverbal ge- in Old and Middle English. Ms. McGregor, W. 2008. Indexicals as sources of case markers in Australian languages. In F. Josephson & I. Söhrman (eds.), Interdependence of Diachronic and Synchronic Analyses, 299–321. Amsterdam: Benjamins. McWhorter, John 2007. Language Interrupted. Oxford: Oxford University Press. McWhorter, John 2016. Is radical analyticity normal: Implications of Niger-Congo and Southeast Asia for typology and diachronic theory. In Elly van Gelderen (ed.), Cyclical Change Continued, 49–92. Amsterdam: John Benjamins.

236

References

Mei, Tsu-lin. 1989. The causative and denominative functions of the *s- prefix in Old Chinese. In Proceedings of the 2nd International Conference on Sinology, 33–51. Taipei: Academia Sinica. Mei, Tsu-lin. 2012. The causative *s- and nominalizing *-s in Old Chinese and related matters in Proto-Sino-Tibetan. Language and Linguistics 13: 1–28. Meillet, Antoine 1910. Compte-rendu: Ch. Bally – Traité de stylistique française. Bulletin de la Société de linguistique de Paris 16: cxviii–cxxii. Meillet, Antoine 1912. L’évolution des formes grammaticales. In Édouard Champion (ed.), Linguistique historique et linguistique générale, 130–148. Paris: Librairie Ancienne Honoré Champion [reprint 1958]. Meillet, Antoine 1913. Compte-rendu: Ch. Bally – Le langage et la vie. Bulletin de la Société de linguistique de Paris 18: clxxix–clxxxii. Meillet, Antoine 1915/1916. Le renouvellement des conjonctions. In Antoine Meillet (ed.), Linguistique historique et linguistique générale, 159–174. Paris: Librairie Ancienne Honoré Champion [reprint 1958]. Meillet, Antoine 1926. Compte-rendu: Ch. Bally – Le langage et la vie, seconde édition. Bulletin de la Société de linguistique de Paris 27.82: 14–16 Meisner, Charlette & Natascha Pomino 2014. Synchronic variation in the expression of French negation: A Distributed morphology approach. Journal of French Language Studies 24.1: 9–28. Melchert, Craig 1975. ‘Exceptions’ to exceptionless sound laws. Lingua 35: 135–153. Michaelis, Susanne 1997. Antikausativ als Brücke zum Passiv. In Otto Winkelmann et al. (eds.), Neuere Beschreibungsmethoden der romanischen Syntax. Tübingen: Narr. Miestamo, Matti & Eino Koponen 2015. Negation in Skolt Saami. In Miestamo et al. (eds.), Negation in Uralic Languages, 353–375. Amsterdam: John Benjamins. Millar, Robert McColl 2000. System Collapse, System Rebirth: The Demonstrative Pronouns of English 900–1350 and the Birth of the Definite Article. Bern: Peter Lang. Mithun, Marianne 1998. Fluid aspects of negation in Central Pomo. In Leanne Hinton & Pamela Munro (eds.), Studies in American Indian Languages: Description and Theory, 77–86. Berkeley: University of California Press. Mithun, Marianne 2008a. The extension of dependency beyond the sentence. Language 84.1: 69–119. Mithun, Marianne 2008b. Borrowed rhetorical constructions as starting points for grammaticalization. In Alexander Bergs & Gabriele Diewald (eds.), Constructions and Language Change, 195–230. Berlin: Mouton de Gruyter. Mithun, Marianne 2011. Grammaticalization and explanation. In Bernd Heine & Heiko Narrog (eds.), The Oxford Handbook of Grammaticalization, 177–192. Oxford: Oxford University Press. Mithun, Marianne 2013. Prosody and independence: Free and bound person marking. In Dik Bakker & Martin Haspelmath (eds.), Language Across Boundaries: Studies in Memory of Anna Siewierska, 291–312. Berlin: Mouton de Gruyter. Mithun, Marianne 2016. What cycles when and why? In Elly van Gelderen (ed.), Cyclical Change Continued, 19–45. Amsterdam: John Benjamins. Mithun, Marianne 2021. Stories behind post-verbal negation clustering. Studies in Language 45.1: 684–706. Morin, Yves-Charles 1979. La morphophonologie des pronoms clitiques en français populaire. Cahier de linguistique 9: 1–36.

References

237

Morris, Richard 1874–1893. Cursor Mundi, 7 Parts. London: Trübner & Co. Mosel, Ulrike 1994. Saliba. München: Lincom Europa. Mossé, Fernand 1938. Histoire de la Forme Périphrastique Participe Présent. Paris: Klincksieck. Mounole, Céline 2011. Le Verbe Basque Ancient. University of the Basque Country PhD. Mukherjee, Sayantan 2019. Syntheticity and analyticity: Indian English and the need for an analytic index. Jadavpur Journal of Languages and Linguistics 3.2: 124–137. Munaro, Nicola & Hans-Georg Obenauer 1999. On underspecified wh-elements in pseudo-interrogatives. University of Venice Working Papers in Linguistics 9: 181–253. Munro, Pamela 1974. Topics in Mojave Syntax. San Diego: University of California. Munro, Pamela 1976. Mojave Syntax. New York: Garland. Munro, Pamela 1977. From existential to copula: The history of Yuman BE. In Charles Li (ed.), Mechanisms of Syntactic Change, 445–490. Austin: University of Texas Press. Mustanoja, Tauno 1960 [2016]. A Middle English Syntax: Parts of Speech. Amsterdam: John Benjamins. Myhill, John 1995. Change and continuity in the functions of the American English modals. Linguistics 33: 157–211. Nadasdi, Terry 1995. Subject NP doubling, matching and minority French. Language Variation and Change 7: 1–14. Narrog, Heiko 2017. Typology and grammaticalization. In Alexandra Aikhenvald & R.M.W. Dixon (eds.), The Cambridge Handbook of Linguistic Typology, 151–177. Cambridge: Cambridge University Press. Narrog, Heiko & Bernd Heine 2021. Grammaticalization. Oxford: Oxford University Press. Naumann, Bernd 1992. The history of the earth and the origin of language: The case for catastrophism. In Bernd Naumann, Frans Plank & Gottfried Hofbauer (eds.), Language and Earth: Elective Affinities Between the Emerging Sciences of Linguistics and Geology, 29–56. Amsterdam: John Benjamins. Nerlich, Brigitte 1990. Change in Language: Whitney, Bréal, and Wegener. London: Routledge. Nesselhauf, Nadja 2012. Mechanisms of language change in a functional system: The recent semantic evolution of English future time expressions. Journal of Historical Linguistics 2.1: 83–132. Newmeyer, Frederick 1998. Language Form and Language Function. Cambridge: MIT Press. Nicholas, Nick 1996. Copulas from Pronouns. Linguist List 7–1776. Nichols, Johanna 1986. Head-marking and dependent-marking grammar. Language 66.1: 56–119. Nichols, Johanna 1992. Linguistic Diversity in Space and Time. Chichago: Chacago University Press. Nichols, Johanna 2018. Non-linguistic conditions for causativization as a linguistic attractor. Frontiers in Psychology 8: 2366. https://doi.org/10.3389/fpsyg.2017.02356 Nichols, Johanna & Balthasar Bickel 2013a. Locus of marking in the clause. In Matthew Dryer & Martin Haspelmath (eds.), The World Atlas of Language Structures Online. Leipzig: Max Planck Institute for Evolutionary Anthropology. http://wals. info/chapter/23.

238

References

Nichols, Johanna & Balthasar Bickel. 2013b. Locus of marking: Whole-language typology. In Matthew Dryer & Martin Haspelmath (eds.), The World Atlas of Language Structures Online. Leipzig: Max Planck Institute for Evolutionary Anthropology. http://wals.info/chapter/25. Noailly, M. 1997. Les mystères de la transitivité invisible. Languages 127: 96–109. Norde, Muriel 2009. Degrammaticalization. Oxford: Oxford University Press. O’Grady, William, John Archibald, Mark Aronoff & Janie Rees-Miller 2017. Contemporary Linguistics, 7th ed. Boston: Bedford/St. Martins. Öhl, Peter 2009. Sprachwandel und kognitive Ökonomie. Linguistische Berichte 220: 393–438. Oinas, Felix 1961. The Development of Some Postpositional Cases in Balto-Finnic Languages. Helsinki: Suomalais-Ugrilainen Seura. Operstein, Natalie 2014. Origin of the zapotec causative marker *k-: A diachronictypological perspective. International Journal of American Linguistics 80.1: 99–114. Ottosson, Kjartan 2008. The old nordic middle voice in the pre-literary period. In Folke Josephson & Ingmar Söhrman (eds.), Interdependence of Diachronic and Synchronic Analyses, 185–219. Amsterdam: John Benjamins. Owens, Jonathan 2018. Dialects (speech communities), the apparent past and grammaticalization: Towards an understanding of the history of Arabic. In H. Clive (ed.), Arabic Historical Dialectology: Linguistic and Sociolinguistic Approaches, 206–253. Oxford: Oxford University Press. Padilla-Moyano, Manuel 2013. Non-finite verbal morphology. In Mikel Martínez-Areta (ed.), Basque and Proto-Basque, 323–357. Frankfurt am Main: Peter Lang. Parodi, Claudia 1995. Object shift in Old Spanish. In Henning Andersen (ed.), Historical Linguistics 1993, 371–378. Amsterdam: John Benjamins. Parry, Mair 2013. Negation in the history of Italo-Romance. In David Willis et al. (eds.), The History of Negation in the Languages of Europe and the Mediterranean, 77–118. Oxford: Oxford University Press. Pauwels, J.L. 1959. Afrikaans hierdie daardie. Leuvense Bijdragen 48.1–2: 1–3. Payne, Thomas 1997. Describing Morphosyntax: A Guide for Field Linguists. Cambridge: Cambridge University Press. Pedersen, Holger 1907. Neues und nachträgliches. Zeitschrift für vergleichende Sprachforschung 40: 129–217. Pedersen, Holger 1931. The Discovery of Language: Linguistic Science in the Nineteenth Century. Bloomington: Indiana University Press. Persson, Maria 2008. The role of the “b”-prefix in Gulf Arabic dialects as a marker of future, intent and/or irrealis. Journal of Arabic and Islamic Studies 8: 26–52. Petersen, Holger. 1931[1962]. The Discovery of Language. Bloomington: Indiana University Press. Pfau, Roland 2015. The grammaticalization of headshakes: From head movement to negative head. In Andrew Smith et al. (eds.), New Directions in Grammaticalization Research, 9–50. Amsterdam: John Benjamins. Pfau, Roland, Martin Salzmann & Markus Steinbach 2018. The syntax of sign language agreement. Glossa 3.1: 107, 1–46. Pfau, Roland & Markus Steinbach 2011. Grammaticalization in Sign languages. In Bernd Heine & Heiko Narrog (eds.), The Oxford Handbook of Grammaticalization, 681–693. Oxford: Oxford University Press.

References

239

Pinkster, Harm 1990. Latin Syntax and Semantics. London: Routledge. Plag, Ingo 1998. The syntax of some locative expressions in Sranan: Preposition, postposition, or noun? Journal of Pidgin and Creole Languages 13:2: 335–354. Plank, Frans 1999. Split Morphology: How agglutination and flexion mix. Linguistic Typology 3: 279–340. Platts, John 1884. A Dictionary of Urdu, Classical Hindi, and English. London: Oxford University Press [1930 reprint]. Polinsky, Maria 2013. Antipassive constructions. In Matthew Dryer & Martin Haspelmath (eds.), The World Atlas of Language Structures Online. Leipzig: Max Planck Institute for Evolutionary Anthropology. http://wals.info/chapter/108. Polotsky, H.J. 1960. The Coptic conjugation system. Orientalia 29.4: 392–422 Poser, William 2009. The Carrier Language: A Brief Introduction. Ms. Post, Mark 2007. Grammaticalization and compounding in Thai and Chinese. Studies in Language 31.1: 117–175. Poussa, Patricia 1982. The evolution of early standard English: The creolization hypothesis. Studia Anglica Posnaniensia 14: 69–85. Powell, John 2021. Deictic Clitic Cycles in Pashto. Ms. Pustet, Regina. 2003. Copulas: Universals in the Categorization of the Lexicon. Oxford: Oxford University Press. Putnam, Michael 2006. Those there demonstratives. Leuvense Bijdragen 95: 159–177. Pye, Clifton 2016. Mayan negation cycles. In Elly van Gelderen (ed.), Cyclical Change Continued, 219–248. Amsterdam: John Benjamins. Pysz, Agnieszka 2007. The usage of demonstratives in the Peterborough Chronicle against the background of the Old English paradigm. In Alexander Bergs & Janne Skaffari (eds.), The Language of the Peterborough Chronicle, 57–75. Frankfurt: Peter Lang Verlag. Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech & Jan Svartvik 1985. A Comprehensive Grammar of the English Language. London: Longman. Raidt, Esther 1993. Linguistic variants and language change: Deictic variants in some German and Dutch dialects vis-à-vis Afrikaans. In Jaap van Marle (ed.), Historical Linguistics 1991, 281–294. Amsterdam: John Benjamins. Raith, Esther 1993. Linguistic variants and language change. In Jaap van Marle (ed.), Historical Linguistics, 281–294. Amsterdam: John Benjamins. Rauth, Philipp & Augustin Speyer 2021. Adverbial reinforcement of demonstratives in dialectal German. Glossa 6.1(4): 1–24. Rautionaho, Paula & Robert Fuchs 2021. Recent change in stative progressives: A collostructional investigation of British English in 1994 and 2014. English Language and Linguistics 25.1: 35–60. Raynouard, M. 1816. Elements de la grammaire de la langue Romane. Paris: Didot. https://books.google.com/books?id=PtVLAAAAcAAJ&printsec=frontcover&source =gbs_ge_summary_r&cad=0#v=onepage&q&f=false Reinöhl, Uta & Nikolaus Himmelmann 2017. Renewal: A figure of speech or a process sui generis? Language 93.2: 381–413. Rice, Keren 1989. A Grammar of Slave. Berlin: Mouton de Gruyter. Rijk, Rudolf de 1995. Standard Basque: A Progressive Grammar. Cambridge, MA: MIT Press. Rissanen, Matti 2007. From oþ to till: Early loss of an adverbial coordinator. In Ursula Lenker et al. (eds.), Connectives in the History of English, 61–75. Amsterdam: Benjamins.

240

References

Ritter, Elizabeth & Martina Wiltschko 2019. The syntax of formality: Universals and variation. In Proceedings of the 2019 Annual Conference of the Canadian Linguistic Association, 1–14. http://martinawiltschko.com/__static/2200ee473bafcfb88993e970 26dfe6fb/ritter-and-wiltschko_cla-2019-_proceedings_au2019.pdf?dl=1 Roberts, Ian 1993. Verbs and Diachronic Syntax. Dordrecht: Kluwer. Roberts, Ian & Anna Roussou 2003. Syntactic Change. Cambridge: Cambridge University Press. Robins, Robert H. 1967. A Short History of Linguistics. London: Longman. Roehrs, Dorian 2010. Demonstrative reinforcer constructions. Journal of Comparative Germanic Linguistics 13.3: 225–268. Romero, Sergio G. 2012. A Maya version of the Jespersen’s cycle: The diachronic evolution of negative markers in K’iche’ Maya. International Journal of American Linguistics 78: 77–96. Rose, Françoise 2018. The rise and fall of Mojeño diminutives through the centuries. Studies in Language 42.1: 146–181. Rosés Labrada, Jorge Emilio 2018. The Piaroa subject marking system and its diachrony. Journal of Historical Linguistics 8.1: 31–58. Ross, Malcolm D. 2002. The history and transitivity of western Austronesian voice and voice-marking. In Fay Wouk & Malcolm David Ross (eds.), The History and Typology of Western Austronesian Voice Systems, 17–62. Canberra: Pacific Linguistics. Ross, Malcolm D. 2004. The morphosyntactic typology of Oceanic languages. Language and Linguistics 5.2: 491–541. Royen, Gerlach 1929. Die nominalen Klassifikations-Systeme in den Sprachen der Erde. Wien: Anthropos. Rudanko, Juhani 2006. Emergent alternation in complement selection: The spread of the transitive into -ing construction in British and American English. Journal of English Linguistics 34.4: 312–331. Rupp, Laura 2007. The (socio-)linguistic cycle of definite article reduction. Folia Linguistic Historica 28: 215–249. Rupp, Laura & Sally Tagliamonte 2017. This here town: Evidence for the development of the English determiner system from a vernacular demonstrative construction in York English. English Language and Linguistics 23.1: 81–103. Rutten, Gijsbert & Marijke van der Wal 2014. Social and constructional diffusion: Relative clauses in seventeenth- and eighteenth-century Dutch. In R. Boogaart et al. (eds.), Extending the Scope of Construction Grammar. Berlin: de Gruyter. Saddour, Inès 2009. The expression of progressivity in Tunisian Arabic: A study of progressive markers in oral retellings of simultaneous situations. Revue de Sémantique et Pragmatique 25: 265–280. Sagart, Laurent 1999. The Roots of Old Chinese. Amsterdam: John Benjamins. Said, Edward 1978. Orientalism. Pantheon Books. Sampson, Tory & Rachel Mayberry 2022. An emerging SELF: The copula cycle in American sign language. Language 98.2: 327–358. Sansò, Andrea 2017. Where do antipassive constructions come from? Diachronica 34.2: 175–218. Sapir, Edward 1911. The problem of noun incorporation in American languages. American Anthropologist 13: 250–282. Sapir, Edward 1921. Language. New York: Harcourt, Brace and Company.

References

241

Sasse, Hans-Jürgen 1984. Case in cushitic, semitic and berber. In James Bynon (ed.), Current Progress in Afro-Asiatic Linguistics, 111–126. Amsterdam: John Benjamins. Savijärvi, Ilkka 1977. Itämerensuomalaisten kielten kieltoverbi. I. Helsinki: Suomalaisen Kirjallisuuden Seura. Scheffer, Johannes 1975. The Progressive in English. Amsterdam: North-Holland Publishing. Schiering, René 2010. Reconsidering erosion in grammaticalization. In Katherina Stathi, Elke Gehweiler & Ekkehard König (eds.), Grammaticalization, 73–100. Amsterdam: John Benjamins. Schiering, René, Balthasar Bickel & Kristine Hildebrandt 2010. The prosodic word is not universal, but emergent. Journal of Linguistics 46: 657–709. Schlegel, August Wilhem von 1808. Über die Sprache und Weisheit der Indier. Heidelberg. Schlegel, August Wilhem von 1818. Observations sur la langue et la literature provençales. Paris. Schourup, Lawrence 1999. Discourse markers. Lingua 107: 227–265. Schwegler, Armin 1990. Analyticity and Syntheticity. Berlin: Mouton de Gruyter. Scivoletto, Giulio 2020. Semasiological cyclicity in the evolution of discourse markers. Journal of Historical Pragmatics 21.2: 236–262. Seržant, Ilja 2021a. Cyclic changes in verbal person-number indexes are unlikely. Folia Linguistica Historica 42.1:49–86. Seržant, Ilja 2021b. Diachronic typology of partitives. In Petra Sleeman & Giuliana Giusti (eds.), Partitive Determiners, Partitive Pronouns and Partitive Case, 111–168. Berlin: De Gruyter. Seržant, Ilja & George Moroz 2022. Universal Attractors in Language Evolution Provide Evidence for the Kinds of Efficiency Pressures Involved. www.nature.com/articles/ s41599-022-01072-0 Shannon, Claude E. 1948. A mathematical theory of communication. Bell System Technical Journal 27.3: 379–423. Shannon, Claude E. 1951. Prediction and entropy of printed English. Bell System Technical Journal 30: 50–64. Shaul, Dave 1986. Topics in Nevome Syntax. Berkeley: UC Berkeley Press. Shi, Yuzhi & Charles N. Li 2004. A History of Grammaticalization in Chinese. Beijing, China: Peking University Press. Shibasaki, Reijirou 2014. More thoughts on the grammaticalization of personal pronouns. In Sylvie Hancil & Ekkehard König (eds.), Grammaticalization – Theory and Data, 129–156. Amsterdam: John Benjamins. Shields, Kenneth 1978. Speculations concerning the I.E. root *es. Archivum Linguisticum 9: 73–78. Shields, Kenneth 2004. On the origin of the *es-/*wes- suppletion in the Germanic copula: A “new image” perspective. Nowele 45.1: 25–32. Siewierska, Anna 2004. Person. Cambridge: Cambridge University Press. Siewierska, Anna 2013a. Passive constructions. In Matthew Dryer & Martin Haspelmath (eds.), The World Atlas of Language Structures Online. Leipzig: Max Planck Institute for Evolutionary Anthropology. http://wals.info/chapter/107. Siewierska, Anna 2013b. Verbal person marking. In Matthew Dryer & Martin Haspelmath (eds.), The World Atlas of Language Structures Online. Leipzig: Max Planck Institute for Evolutionary Anthropology. http://wals.info/chapter/102.

242

References

Simoncsics, Péter 1998. Kamassian. In Daniel Abondolo (ed.), The Uralic Languages, 580–601. London: Routledge. Simonenko, Alexandra, Benoit Crabbé & Sophie Prévost 2019. Agreement syncretization and the loss of null subjects: Quantificational models for Medieval French. Language Variation and Change 31.3: 275–301. Sims, Lynn 2008. Ingressive Aspect in the History of English. ASU. PhD. Skeat, Walter (ed.) 1881–7. The Gospel According to St. Matthew, St. Mark, St. Luke and St. John. Cambridge. Reprint: Darmstadt: Wissenschaftliche Buchgesellschaft. Slade, Benjamin 2013. The diachrony of light and auxiliary verbs in Indo-Aryan. Diachronica 30.4: 531–578. Slobodchikoff, C.N. 2010. Alarm calls in birds and mammals. In M. Breed & J. Moore (eds.), Encyclopedia of Animal Behavior 1, 40–43. Oxford: Academic Press. Slobodchikoff, Tatyana 2013. The Slavic Dual: Morphosyntactic Feature Economy as a Factor in Language Change. University of Arizona PhD. Slobodchikoff, Tatyana 2019. The Evolution of the Slavic Dual: A Biolinguistic Perspective. Washington, DC: Rowman & Littlefield. Smith, Aaron 2006. The universal tendency for renewal among grammatical expressions for anterior and related aspect. Journal of Universal Language 7: 139–160. Smith, Adam 1759. The Theory of Moral Sentiments. To Which is Added a Dissertation on the Origin of Languages. London [1869, 3rd ed]. Smitterberg, Erik 2005. The Progressive in Nineteenth-Century English. A Process of Integration. Amsterdam: Rodopi. Sneddon, James 1996. Indonesian. London: Routledge. Sohn, Ho-Min 1999. The Korean Language. Cambridge: Cambridge University Press. Sollid, Hilde 2002. Features of Finnish in a Northern Norwegian Dialect. www.joensuu. fi/fld/methodsxi/abstracts/sollid.html Sommerer, Lotte 2015. The influence of constructions in grammaticalization. In Lotte Sommerer, Spike Gildea, Jóhanna Barðdal & Elena Smirnova (eds.), Diachronic Construction Grammar, 107–137. Amsterdam: John Benjamins. Sommerer, Lotte et al. 2015. Diachronic Construction Grammar. Amsterdam: John Benjamins. Song, Jae Jung 1990. On the rise of causative affixes: A universal-typological perspective. Lingua 82.2–3: 151–200. Specht, Franz 1947. Der Ursprung der Indogermanishen Deklination. Göttingen: Vandenhoeck & Ruprecht. Starosta, Stanley, Andrew Pawley & Lawrence Reid 1981. The evolution of focus in Austronesian. In Third International Conference on Austronesian Linguistics, January, 24–29, 1981. Denpasar, Bali, Indonesia. https://vdocuments.mx/starosta-pawley-reid.html?page=1 Stassen, Leon 1997. Intransitive Predication. Oxford: Clarendon Press. Stassen, Leon 2013. Zero Copula for Predicate Nominals. In Matthew Dryer & Martin Haspelmath (eds.), The World Atlas of Language Structures Online. Leipzig: Max Planck Institute for Evolutionary Anthropology. http://wals.info/chapter/120 Stiernhielm, Georg 1671. Evangelia ab Ulfila Gothorum in Moesia Episcopo ex Graeco Gothica translata. Stockholm: Wankif. Strang, Barbara 1982. Some aspects of the history of the Be+ing construction. In John Anderson (ed.), Language Form and Linguistic Variation: Papers Dedicated to Angus McIntosh, 427–474. Amsterdam: John Benjamins.

References

243

Streitberg, Wilhelm. 1891. Perfective und imperfective Actionsart im Germanischen. Beiträge zur Geschichte der deutschen Sprache und Literatur 15: 70–177. Sullens, Idelle (ed.) 1983. Robert Mannyng’s Handlyng Synne. Binghampton: Medieval & Renaissance Texts & Studies. Sun, Chaofeng 1996. Word-Order Change and Grammaticalization in the History of Chinese. Stanford: Stanford University Press. Sun, Chen Chen 2008. Variations in the BA Construction and its Relevance to DP: A Minimalist Perspective. ASU PhD. Sundquist, John 2006. Syntactic variation in the history of Norwegian and the decline of XV word order. Diachronica 23.1: 105–141. Sweet, Henry 1871 [1934]. King Alfred’s West-Saxon Version of Gregory’s Pastoral Care. London: Oxford University Press. Sweetser, Eve 1988. Grammaticalization and semantic bleaching. In Proceedings of the Fourteenth Annual Meeting of the Berkeley Linguistics Society, 389–405. Szmrecsanyi, Benedikt 2012. Analyticity and syntheticity in the history of English. In Terttu Nevalainen & Elizabeth Traugott (eds.), Oxford Handbook on Historical English Linguistics, 654–665. Oxford: Oxford University Press. Szmrecsanyi, Benedikt 2016. An analytic-synthetic spiral in the history of English. In Elly van Gelderen (ed.), Cyclical Change Continued, 93–112. Amsterdam: John Benjamins. Tan, Tamisha 2020. Two Types of Pronoun ‘Doubling’ in Amarasi. AFLA27. Tan, Tamisha 2022. Inclined to Agree: From Pronominal Copula to Predicative Agreement. DIGS Talk.7 Tauli, Valter 1956. The origin of affixes. Finnisch-Ugrische Forschungen 32.1/2: 170–225. Tauli, Valter 1958. The Structural Tendencies of Languages. Helsinki: Suomalaisen Tiedeakatemian Toimituksia. Tauli, Valter 1966. Structural Tendencies in Uralic Languages. Den Haag: Mouton. Tesnière, Lucien 1932. Synthétisme et Analytisme. Charisteria Gvilelmo Mathesio Qvinqvagenario. Prague Linguistic Circle. Tesnière, Lucien de 1959. Eléments de Syntaxe Structural. Paris: Klincksieck. Thomason, Sarah & Terrence Kaufman 1988. Language Contact, Creolization and Genetic Linguistics. Berkeley: University of California Press. Thorpe, Benjamin 1861. Anglo-Saxon Chronicle I and II. London: Longman. Tiedemann, Dieterich 1772. Versuch einer Erklärung des Ursprunges der Sprache. Riga: Hartknoch. Tops, Guy 1974. The Origin of the Germanic Dental Preterit. Leiden: Brill. Traugott, Elizabeth 1972. A History of English Syntax. New York: Holt, Rinehart, and Winston. Traugott, Elizabeth 1998. Historical pragmatics and the development of scalar focus particles. In Paper presented at the lOth International Conference on English Historical Linguistics, Manchester. Traugott, Elizabeth 2001. Legitimate Counterexamples to Unidirectionality. Talk. https:// web.stanford.edu/~traugott/papers/Freiburg.Unidirect.pdf Traugott, Elizabeth 2008. Grammaticalization, constructions, and the incremental development of language. In Regine Eckardt et al. (eds.), Variation, Selection, Development, 219–250. Berlin: Mouton de Gruyter.

244

References

Traugott, Elizabeth 2014. Toward a constructional framework for research on language change. In Sylvie Hancil & Ekkehard Kõnig (eds.), Grammaticalization – Form and Data, 87–105. Amsterdam: John Benjamins. Traugott, Elizabeth 2018. Modeling language change with constructional networks. In Salvador Pons Bordería & Óscar Loureda Lamas (eds.), Beyond Grammaticalization and Discourse Markers: New Issues in the Study of Language Change, 17–50. Leiden: Brill. Traugott, Elizabeth 2022. Ten Lectures on a Diachronic Constructionalist Approach to Discourse Structuring Marker. Leiden: Brill. Traugott, Elizabeth & Bernd Heine 1991. Approaches to Grammaticalization. Amsterdam: John Benjamins. Traugott, Elizabeth & Ekkehard König 1991. The semantics-pragmatics of grammaticalization revisited. In E.C. Traugott & B. Heine (eds.), Approaches to Grammaticalization, vol. 1, 189–218. Amsterdam: John Benjamins. Traugott, Elizabeth & Graeme Trousdale 2013. Constructionalization and Constructional Changes. Oxford: Oxford University Press. Trips, Carola 2001. From OV to VO in Early Middle English. Amsterdam: John Benjamins. Trudgill, Peter 2011. Sociolinguistic Typology. Oxford: Oxford University Press. Truppi, Chiara 2021. The copula cycle in Kriol. Journal of Historical Syntax 5: 1–27. Turner, R.L. 1966. A Comparative Dictionary of the Indo-Aryan Languages. London: Oxford University Press. Uhlenbeck, C.C. 1901 Agens und Patiens im Kasussystem der indogermanischen Sprachen. Indogermanische Forschungen 12: 170–171. Vance, Barbara. 1997. Syntactic Change in Medieval French: Verb-Second and Null Subjects. Dordrecht: Kluwer. van der Auwera, Johan (ed.) 1998. Adverbial Constructions in the Languages of Europe. Berlin: Mouton de Gruyter. van der Auwera, Johan 2009. The Jespersen cycles. In: Elly van Gelderen (ed.), CyclicalChange, 35–71. Amsterdam: John Benjamins. van der Auwera, Johan & Vladimir Plungian 1998. Modality’s semantic map. Linguistic Typology 2: 79–124. Van der Auwera, Johan & Frens Vossen 2016. Jespersen cycles in the Mayan, Quechuan and Maipurean languages. In Elly van Gelderen (ed.), Cyclical Change Continued, 189–218. Amsterdam: John Benjamins. Van de Velde, Mark 2021. The Bantu relative agreement cycle. Linguistics 59.4: 981–1015. Veenstra, Tonjes 2014. The Copular System in Surinamese Creoles. Workshop on Copulas, Bologna. Vennemann, Theo 1973. Topics, subjects, and word order: From SXV to SVX via TVX. In John Anderson & Charles Jones (eds.), Historical Linguistics: Proceedings of the First International Congress of Historical Linguistics, vol. II, 339–376. Amsterdam: North-Holland. Verhagen, Arie 1992. Praxis of linguistics: Passives in Dutch. Cognitive Linguistics 3.3: 301–342. Berlin, New York: Walter de Gruyter. Vermaas, J.A.M. 2005. Veranderingen in de Nederlandse aanspreekvormen van de dertiende t/m de twintigste eeuw. Utrecht: LOT.

References

245

Veselinova, Ljuba 2013. Negative existentials: A cross-linguistic study. Rivista di Linguistica 25.1: 107–145. Veselinova, Ljuba 2016. The negative existential cycle viewed through the lens of comparative data. In Elly van Gelderen (ed.), Cyclical Change Continued, 139–187. Amsterdam: John Benjamins. Veselinova, Ljuba & Arja Hamari 2022. Introducing the negative existential cycle. In Ljuba Veselinova & Arja Hamari (eds.), The Negative Existential Cycle, 1–56. Berlin: Language Science Press. Viberg, Åke 2013. Posture verbs. Languages in Contrast 13.2: 139–169. Vincent, Nigel 1999. The evolution of c-structure: Prepositions and PPs from IndoEuropean to romance. Linguistics 37.6: 1111–1153. Vindenes, Urd 2018. Cyclic renewal of demonstratives. Studies in Language 42.3: 641–668. Visser, F. Th. 1946–1952. A Syntax of the English Language of St Thomas More. Leuven: Librairie Universitaire. Visser, F. 1963–1973. An Historical Syntax of the English Grammar, Vol I-IIIb. Leiden: Brill. Vossen, Frens & Johan van der Auwera 2014. The Jespersen cycles seen from Austronesian. In Maj-Britt Mosegaard Hansen & Jacqueline Visconti (eds.), The Diachrony of Negation, 47–82. Amsterdam: John Benjamins. Walkden, George 2022. Sources for Question Particles. DIGS talk. Walker, Jim 2011. The emergence of the narrative present perfect in British English: Reality or illusion? GAGL: Groninger Arbeiten zur Germanistischen Linguistik 53.2: 71–87. Walker, Jim 2017. The present perfect has only gone and changed, hasn’t it? In Aorists and Perfects: Synchronic and Diachronic Perspectives. https://hal.univ-lyon2.fr/hal02501520v1/preview/article%206%20aorists.pdf Waltereit, Richard 2020. Parallels between the negative cycle and the rise of interrogative marking in French. Journal of Historical Pragmatics 21.2: 263–288. Whaley, Lindsay 1997. Introduction to Typology. Los Angeles: Sage. Whitman, John 1999. Personal pronoun shift in Japanese: A case study in lexical change and point of view. In Akio Kamio & Ken-ichi Takami (eds.), Function and Structure: In Honor of Susumu Kuno, 357–386. Amsterdam: John Benjamins. Whitney, William Dwight 1875. The Life and Growth of Language. London: King & Co. Whitney, William Dwight 1889. Sanskrit Grammar, 2nd ed. Boston: Harvard University Press. Wilcox, Phyllis & Sherman Wilcox 1995. The gestural expression of modality in American Sign Language. In Joan Bybee & Suzanne Fleischman (eds.), Modality in Grammar and Discourse, 135–162. Amsterdam: John Benjamins. Willis, David 2011. Negative polarity and the quantifier cycle: Comparative diachronic perspectives from European languages. In Pierre Larrivée & Richard Ingham (eds.), The Evolution of Negation: Beyond the Jespersen Cycle, 285–324. Berlin: De Gruyter Mouton. Willis, David 2016. Incipient Jespersen’s cycle in Old English negation. In Sten Vikner et al. (eds.), Let Us Have Articles Betwixt Us – Papers in Historical and Comparative Linguistics in Honour of Johanna L. Wood, 465–491. Aarhus: Aarhus University Library.

246

References

Willis, David, Christopher Lucas & Anne Breitbarth (eds.) 2013. The History of Negation in the Language of Europe and the Mediterranean. Oxford: Oxford University Press. Willson, Heather 2008. Subject Positions in Marshallese. UCLA PhD dissertation. Wischer, Ilse & Mechthild Habermann 2004. Der Gebrauch von Präfixverben zum Ausdruck vons Aspekt/Aktionsart im Altenglischen und Althochdeutschen. Zeitschrift für Germanistische Linguistik 32: 262–285. Woidich, Manfred 1968. Negation und negative Sätze im Agyptisch-Ambischen. Universität München. Dissertation. Woidich, Manfred 2006. Das Kairenisch-Arabische. Wiesbaden: Harrasowitz. Wolff, John 1973. Verbal inflection in Proto-Austronesian. In Andrew Gonzalez (ed.), Parangal Kay Cecilio Lopez, 71–91. Manila: Linguistic Society of the Philippines. Wood, Johanna 2003. Definiteness and Number: Determiner Phrase and Number Phrase in the History of English. ASU PhD. Wood, Johanna 2016. The degree cycle. In Elly van Gelderen (ed.), Cyclical Change Continued, 287–318. Amsterdam: John Benjamins. Yang, Hui-Ling 2012. The Grammaticalization of Hakka, Mandarin and Southern Min. ASU PhD. Yang, Ning 2008. The Indefinite Object in Mandarin Chinese. Leiden: HIL Publications. Zepeda, Ofelia 1983. A Papago Grammar. Tucson: University of Arizona Press. Ziegeler, Debra 1999. Agentivity and the history of the English progressive. Transactions of the Philological Society 97.1: 51–101. Zieglschmid, Friedrich 1930. The disappearance of Werdan in English. Philological Quarterly 9: 11–116. Zribi-Hertz, Anne. 1994. La syntaxe des clitiques nominatifs. Travaux de Linguistique et Litterature: 131–147. Zucchi, Sandro 2003. The semantics of FATTO. Paper Presented at the 14th Amsterdam Colloquium, Amsterdam. Zúñiga, F. & S. Kittilä 2019. Voice. Cambridge: Cambridge University Press.

INDEX

Note: Page locators in bold indicate a table. Page locators in italics indicate a figure. ABL see ablative ablative 162 ABS see absolutive absolutive 82–83, 130, 152, 160 ACC see accusative accusative 68, 107, 137–138, 143–144, 159–160, 180–181, 193 Adamović, Milan 146 Adams, Mariane 144 adverb: functions 51–53, 56; locative 44, 50–51, 53–54, 56; modality 77–78; negative 4, 9, 95, 98, 106, 198; positive 8; renewal 19, 24; temporal 11, 18, 113, 117, 119–121, 200; tense markers 58, 70 affixation 6 affixes 1, 3, 14, 20–21, 29; subject 141, 148; subjunctive 73; synthetic 130, 210 AFOC see assertive focus agglutination 3, 14, 28, 208 agglutinative 3, 14, 30, 135–136, 207–208 Al Aloula, Mashael 61 Alasmari, Abdullah 86 Aldridge, Edith 107 Allen, Cynthia 157, 179, 181 Altamimi, Mansour 61, 77

analytic 1, 3, 26; language 41, 129–132, 135, 163, 210; modern 30, 32; to synthetic 8, 10, 19, 32, 35, 128–131, 140, 167 Analytic to Synthetic Cycle 16, 130, 138 Andersen, Henning 5, 37 Anderson, Stephen 160 Anglo-Saxon Chronicles 23 Ansaldo, Umberto 6 ANTIP see antipassive antipassive 82–84, 89–90 Apache 12 Arabic 19, 53–55; standard 54, 60, 73, 86, 177 ART see article article 21, 43, 54, 73; definite 44, 46, 51, 185, 187, 210, 216; generic/nongeneric 44, 187; introduction 180; loss of 44, 201 Ashby, William 136, 142, 163 aspect: anterior 71; auxiliary 58; continuous 59, 62, 64, 65, 213; habitual 59, 66, 92; markers 58, 83, 108, 124; perfect 59; perfective 58–59, 104–105, 176, 180; progressive 59–61, 72, 187; voice head 90, 191, 198 Aspect Cycle 177, 201

248

Index

assertive focus 100 Athabaskan 104, 119, 142 Auderset, Sandra 89 Austronesian 133, 151 auxiliaries 2, 7, 59, 71, 93, 100–103; modal 75–76, 78, 82, 91; voice 138–139, 155 auxiliary: future marker 7, 14, 210; modal 24, 42, 58, 73, 75, 77–78, 82, 214 Bácskai-Atkári, Júlia 209 Baerman, Matthew 202 Bahrt, Nicklas 90 Bahtchevanova, Mariana 136, 163, 166 Bailey, Charles 179 Bailey, Laura 106 Baker, Robin 160 Bakir, Murtadha J. 60 balance 15, 32, 173, 201 Bally, Charles 26, 33 Bantu 2, 12, 96, 147 Barðdal, Johanna 39, 183, 185 Bashir, Elena 102 Bately, Janet 23 Bencini, Giulia 106 Benson, Larry 23 Benveniste, Emile 26, 34–35, 54 Bergs, Alexander 183 Bernstein, Judy 45 Bickel, Balthasar 89, 132–135, 140, 156, 168, 214 Bickerton, Derek 16 Binyam, Sisay Mendisu 99 Bisang, Walter 6, 176, 208 Blake, Barry 157, 161 bleaching 6, 18; see also semantic bleaching BNC see British National Corpus Bogomolova, Natalia 147–148 Bolinger, Dwight 76 Bopp, Franz 26, 29, 159–160 Bouzouita, Miriam 2, 15, 26, 36, 121, 126, 203 Bréal, Michel 162 Breitbarth, Anne 4, 26, 36, 97, 121, 126, 190 British English 76, 103 British National Corpus 22, 99, 103 Brook, G. 23 Brown, Dunstan 202 Butt, Miriam 139 Bybee, Joan 16, 58–60, 68, 91, 209

C see complementizer Carey, Kathleen 70 Case Cycle 8, 10, 37, 46, 128, 156, 167, 190, 200 case loss 157, 162 CdES see Corpus d’entretiens spontanés Cely Letters 23 Chatzopoulou, Aikaterini 4, 26, 91 Chaucer, Geoffrey 23, 64 Chipewyan 23 Chomsky, Noam 16, 27, 33, 36, 38, 187, 192, 197–198 Chron. see Chronicle Chronicle 69, 112, 182–183 circa 23 CL see classifier clarity 3, 7; see also comfort and clarity classifier 132 cline 5–6, 15, 21, 71 cliticization 4, 6, 47, 100, 150 CM see common gender COCA see Corpus of Contemporary American English COHA see Corpus of Historical American English comfort 14–15, 30, 34; see also comfort and clarity comfort and clarity 14–15, 26, 32, 172–175, 196, 203 common gender 43 comparative method 12, 210 complementizer 2, 13, 93, 109–110, 113–116, 120, 179, 182–183 Complementizer Cycle 93, 111, 116, 124, 179, 200, 201, 203, 205 Comrie, Bernard 58, 155 COND see conditional Condillac, Étienne Bonnot de 26–28 conditional 88, 118, 139, 206 Condoravdi, Cleo 6, 97 Conrady, A. 138 Construction Grammar 16, 38–39, 173, 183, 203 COP see copula copula 19, 41; negative 55, 93, 103–105, 123 Copula Cycle 8, 9, 41, 53–55, 90, 98, 103, 123, 199, 201, 205–206, 209, 212 corpus 13; analysis 76; data 23, 49, 165 Corpus d’entretiens spontanés 22 Corpus of Contemporary American English 22

Index

Corpus of Historical American English 25 Corver, Norbert 20 Coseriu, Eugen 27 Cournane, Ailís 38, 75 Croft, William 16, 18–19, 94, 100 Croft’s Cycle 18–19, 55, 94, 98, 100–101; see also Negative Existential Cycle Crowley, Terry 135 Culicover, Peter 16, 39, 185–187 Cura Pastoralis (Pope Gregory) 23 Cursor Mundi 23 Curtius, Georg 26, 30 cycle: meaning 1; study of 8, 17–18 Cycles of pragmaticalization 11 cyclical change: history 26; study of 2, 11 d’Ardenne, Simone 23 Dahl, Östen 2, 58–60, 97, 175 DAT see dative dative 143, 180 DEC see declarative declarative 75 DEF see definite definite: affix 43; article 44, 46, 51, 185, 187; marker 45–46, 160 Degand, Liesbeth 120 Delbrück, Berthold 159 DEM see demonstrative Demonstative cycles 25, 45 demonstrative 13, 23, 44; original 159–160; use of 46, 112, 118 Dëne Sųłiné 12, 23, 104 Deo, Ashwini 60, 62 DET see determiner determiner 13, 21, 46, 50–51, 90, 137, 161 Determiner Cycle 12, 41, 43–44, 46, 53, 90, 176, 179, 206 Detges, Ulrich 69, 145, 209 diachronical study 12–13, 97 Dictionary of Old English 22 Diessel, Holger 44, 91, 114 Diewald, Gabrielle 183 Differential Object Marker 149, 161, 202 Dimmendaal, Gerrit 157 DOE see Dictionary of Old English DOM see Differential Object Marker Domingue, Nicole 179 Donohue, Mark 150 double negative 95 Drinka, Bridget 178 Dryer, Matthew 141–142, 168, 201, 215 Du Ponceau, Peter 131

249

Easterday, Shelece 21, 132–134 Eckardt, Regine 125 Elseness, Johan 71 Emonds, Joseph 179 EMPH see emphatic emphatic 8, 106, 119–121; negative 97, 100, 190; particles 102; pronoun 121, 143, 145, 152, 195 English 11, 13, 18–19, 24; history of 1–2, 4; linguistic cycle 31, 49, 71, 75; see also British English Epstein, Richard 45 ERG see ergative ergative 82–83, 160 erosion 3, 6–7, 14, 35, 120, 145, 174–175, 208 EX see existential Existential; see also Negative Existential Cycle: negative 98–99, 104; verbs 100, 102 expression 34–35, 42, 58; linguistic 33, 117; new 3, 6, 14, 117; periphrastic 3, 14, 34, 174 EyÞórrson, Tolli 68 F see feminine Faarlund, Jan Terje 179 Fairbanks, Gordon 162 Fedriana, Chiara 119–120 feminine 47, 62, 77, 123, 139 Ferrell, Raleigh 87 first person 13, 21, 89; singular 196, 215 Fischer, Olga 42, 74–76 Flathe, Philipp Jacob 173 Fodor, Jerry 15 Fryd, Marc 72 Fuchs, Robert 67 Fulk, R.D. 23 Fuß, Eric 147, 155, 168 FUT see future future: direction 204, 208; inferred 6; marker 7, 9–10, 14, 77, 136 Future Cycle 7, 139, 209 Gabelentz, Gerg von der 1–3, 7–8, 10, 14–15, 30, 34, 39, 117–, 172–174, 206 Gaenszle, Martin 89 Gardiner, Alan 2 Gelderen, Elly van 8, 15, 17, 26, 37, 41, 93, 106, 128, 136, 163, 166, 172, 176, 187–189, 195, 197–199, 209

250

Index

GEN see genitive genitive 129, 180–181, 207 Gergel, Remus 9, 18, 42, 75 Geurts, Bart 174–176 Ghezzi, Chiara 119 Gil, David 84 Gilbertese 149, 152–153 Gildea, Spike 39, 183, 185 Givón, Talmy 2, 7, 9, 12, 27, 34–35, 82, 94, 140, 149, 155, 168 Givón’s Cycle 94, 98, 100, 101–107, 123, 140 glosses 23, 25, 114 Goldberg, Adele 16 grammatical: category 5, 8, 15, 17, 34, 41, 54, 178, 205; word 5, 10, 20, 42 grammaticalization: cyclical change 30, 32–37, 39, 172–173, 177–178, 187; framework 15–17, 111, 185, 187; linguistic cycle 3, 5, 10, 26–28, 174; occurrence of 176; perspective 6, 9–10; pragmaticalization 119, 121–122; renewal 91, 100, 106, 116, 129, 142, 201; stages of 45–46, 75, 82, 86, 91 grammation/regrammation 6 Greek 26, 30, 90, 97, 162, 208 Greenberg, Joseph 44–46, 132–133, 160 Greenberg’s Definiteness Cycle 159, 162, 186 Grestenberger, Laura 90 Grüter, Therese 163 Guðmundsson, Helgi 209 Guilfoyle, Eithne 84 Gul, Ahmed 102 Haase, Martin 72–73 Haeberli, Eric 74 Hagège, Claude 7, 174 Haig, Geoffrey 139–140, 149–150, 200–202 Halm, Tamá 87–88 Handel, Zev 138 Handlyng Synne (Mannyng) 23 Hanham, Alison 23 Hansen, Maj-Britt Mosegaard 2, 6, 11, 39, 116–118, 124–125 Harris, Martin 45, 144 Harris, Randy 40 Harrison, Sheldon 151–153

Haspelmath, Martin 10, 16, 20–21, 82–83, 85, 136–137, 140, 142, 174–174, 176, 209 Havers, Wilhelm 26, 33 head marking: cycles 140, 167, 202; language 10, 128, 130, 133 Head Preference Principle 37, 188 Heaney, Seamus 129 Heine, Bernd 5, 9–11, 16, 19, 34–35, 39, 109, 120–121, 158, 178, 180, 209 Hilpert, Martin 39, 183–185, 187 Himmelmann, Nikolaus 206 Hintze, Fritz 137 Hodge, Carleton 2–4, 8, 10, 26–27, 34–36, 129, 137 Holes, Clive 60 Hoop, Helen de 209 Hopper, Mike 6–7, 27, 35, 39 Horne Tooke, John 26–27, 29 HPP see Head Preference Principle Hróarsdóttir, Thorbjörg 157 Huang, James 138 Hubers, Frank 209 Humboldt, Wilhelm von 26, 29–30, 131 Hundt, Marianne 66 Igartua, Iván 207–208 Ihsane, Tabea 74 imperfective 28, 58, 60–63, 66–67, 73, 105, 180, 200, 213 Imperfective Cycle 59–60, 67, 73, 90, 93, 199, 205 Inceptive Cycle 201 Indo-European 22, 54, 58, 90, 109, 139, 159, 162 inflection 13, 42, 103, 105, 139, 202 Ingham, Richard 4, 95 Inglese, Guglielmo 83 INST see instrumental instrumental 30, 36, 134, 156–157 intention 7, 17, 75 internal reconstruction 12, 25 Interrogative Cycle 106, 109, 203, 215 IPF see imperfective IRR see irrealis irrealis 18, 59, 73, 96–97, 124 Ishiyama, Osamu 121–122 Jackendoff, Ray 188 Jäger, Agnes 209 Jakubowicz, Celia 163

Index

Jarad, Najib Ismail 60 Jelinek, Eloise 142 Jespersen, Otto 2, 4, 11, 14–15, 94–95, 121 Jespersen’s Cycle 2, 12, 18–19, 94, 96–98, 100, 102, 106–107, 123, 126, 190, 201, 209 Ji, Ming Ping 107 Johannessen, Janne 44 Jones, William 26–27, 29 Kahr, Joan 158, 161 Kari, James 104 Karlgren, Bernhard 137–138 Katerine 23 Katz, Aya 53 Katz, Jerrold 15 Kaufman, Terrence 179 Kellogg, Samuel H. 101 Kiparsky, Paul 6, 18, 36, 97, 207 Kirby, Ian 154 Kiss, Katalin 209 Kittilä, Seppo 82 Klamer, Marion 150 Klima, Edward 36 Koh, Ann 84 Kóna-Tas, András 145 König, Christa 17, 160 Korostovcev, Mihail Aleksandrovič 137 Koyukon 12 Kranich, Svenja 66, 92 Kulikov, Leonid 160 Kuryłowicz, Jerzy 5, 26, 34–35, 161–162 Kuteva, Tanya 17, 19, 53, 58–59, 109, 158, 178, 180 LA see labeling algorithm labeling algorithm 197 Lahiri, Aditi 139 Lam Chit Yu, Cherry 98 Lambrecht, Knud 142, 144, 163 Langacker, Ronald 174 Larjavaara, Meri 163 Larrivée, Pierre 97 Lash, Elliot 177 Late Merge Principle 37, 190 Ledgeway, Adam 131 Leech, Geoffrey 76 Leer, Jeff 104 Lefèvre, André 1, 26, 31–32, 34 Lehmann, Christian 17, 26–28, 34–36, 46, 209

251

Lehmann, Winfred 27, 74, 146, 162 Leibniz, Gottfried Wilhelm 26–27, 29 Leiss, Elizabeth 41, 180 Lemoine, Kevin 163 Leslie, R. 23 Lewis, Diana 121 lexical: elements 2, 6; expression 10, 42; item 1–2, 9, 11, 17, 91, 192, 210; verb 7, 9, 11, 25, 72, 81, 86, 99, 102, 137, 161 Li, Charles 53, 151, 161 light verb 62, 138 Lightfoot, David 36 Lim, Lisa 6 linguistic cycle: history 26, 31–32, 34–35, 40; origin 1, 3, 7, 10, 16–19, 117, 174; tug-of-war 15, 175 Lloyd, Albert 68 LMP see Late Merge Principle LOC see locative locative 44–45, 103, 157–158, 163, 191; adverb 50–51, 53–54, 56 Longobardi, Guiseppe 191–192 Los, Bettelou 112–113 Lucas, Christopher 96, 177 Lyons, John 45 M see masculine MacLeod, Morgan 71 macro cycles: controversial 8, 10; explanations 172; interactions 162, 167; language typology change 128–131 Magistro, Guiseppe 190 Mair, Christian 203 Mandarin Chinese 19, 23, 55, 98, 107, 136 Mansfield, John 205 Maroldt, Karl 179 Martinet, André 39 Martínez-Areta, Mikel 46 Marušič, Franc 77–78 masculine 47–48, 159, 180 Mayberry, Rachel 209 McCawley, James 15 McDonald, Brittany 208 McElvenny, James 27, 30, 32 McGregor, William 160 McWhorter, John 10, 19, 132 Mei, Tsu-lin 138 Meillet, Antoine 2, 14–15, 18, 26, 32–33, 35–36, 201, 209

252

Index

micro cycle: definition 8–10, 35, 42, 90; interaction 122, 124; polarity 93–94; see also Negative Cycle MID see middle voice Middle English 2, 4, 64, 69–72, 74, 95, 115, 170, 194, 205 middle voice 83, 87–88, 213 Millar, Robert McColl 45 Mithun, Marianne 4, 9, 17–19, 105, 118–119, 160, 209 modal auxiliary 24, 42, 73, 75, 77–78, 82, 214 Modal Cycles 17, 73–74, 82, 90 Modern English 2, 4, 19, 24, 61, 63–64, 69, 95, 129, 157, 171 Molinelli, Piera 119–120 Mood Cycle 74, 199, 205, 211 More, Thomas 65–66 Morin, Yves-Charles 164 Moroz, George 200–202 morphology 4, 11, 16, 128; derivational 130, 138; inflectional 139, 214; voice 87, 90 Morris, Richard 23 Mossé, Fernand 67, 73 Munro, Pamela 57 Mus, Nikolett 209 Mustanoja, Tauno 70, 183 N see noun Narrog, Heiko 11, 16, 154 Navajo 10, 12, 13, 118–119, 122, 132, 140, 147, 155 NEC see Negative Existential Cycle NEG see negative negative: marker 58, 94, 97, 99, 104, 107, 213; particle 18, 55, 78, 103–104, 106, 124; reinforcement 4–5; weakening 8 Negative Cycle 1, 95, 97, 100, 106, 163; history 2, 194–195; stages 4, 6, 8, 12, 18, 43, 177 Negative Existential Cycle 55, 94, 99–100, 123 Nerlich, Brigitte 27 Nesselhauf, Nadja 7, 17 Newmeyer, Frederick 36, 207 Nichols, Johanna 130, 132–135, 140, 156, 168, 200–201, 206, 214 Noailly, Michèle 163 NOM see nominative nominative: article 150; case forms 143–144, 160; first person

pronoun 21, 159; masculine 180; on the predicate 57 Norde, Muriel 207 noun: generic 83, 89; marker 44, 46; masculine 48; verb interaction 39; word placement 28, 49, 51, 213 noun phrase 22 NP see noun phrase OB see object object: demonstrative 112; direct 96; marker 136, 149–150, 160–161, 202; non-human 151–153 Object Agreement Cycle 148-154, 200 object marker 136, 149–150, 160–161 OE see Old English OED see Oxford English Dictionary Öhl, Peter 174 Oinas, Felix 158 Old English: explanations 179–181, 187; head preference principle 188, 194–195; macro cycles 129, 185; micro cycle 42, 48, 63–64, 67–71, 74–75, 111–112, 114; negative cycles 1–2, 4, 9, 94; texts 22–24 Old Norse 12, 19, 50, 84, 91, 213 OM see object marker Orléans corpus 164–165, 167 Orosius (Alfred) 23 OTA see Oxford Text Archive Owens, Jonathan 77 Oxford English Dictionary 1, 23, 109–111, 115, 182–182, 212 Oxford Text Archive 22 P see preposition Padilla-Moyano, Manuel 73 Parodi, Claudia 157 Parry, Mair 126 participle 62–64, 70–71, 78, 81, 86, 90 particle: inflected/uninflected 105–106; negative 55, 78, 103–104, 106, 124 past: simple 70–71; tense 58–59, 69, 73 Pedersen, Holger 27, 40, 159 Perfective Cycle 59, 68–70, 72–73, 179, 199, 205 Persian 19, 23, 149–150 Persson, Maria 77 Pfau, Roland 209 phonetic weakening 6, 97, 145, 196, 204 phonological: impact 2; reduction 5–7, 176; weakening 6, 176

Index

Pinkster, Harm 157 Platts, John 102 Polotsky, Hans Jakob 137 Post, Mark 6, 176, 208 postposition 6, 18, 158–159 Poussa, Patricia 179 PP see preposition phrase PR see present Pragmatic Cycle 2, 36, 41, 93, 117–118, 122, 124–125, 200, 201, 205 prediction 7 preposition 61, 64, 111, 129, 145, 161–162, 181, 183; locative 157–158 preposition phrase 22, 111, 116 present tense 54–55, 58–61, 63, 73, 187 PROG see progressive progressive: markers 53, 61; obligatory 60, 63, 67, 73; special 60, 65 pronoun: isolated 29; subject 18, 141, 148, 151, 154, 168, 202 Pronoun Cycle 13, 18–19, 28, 93, 142, 156, 167, 178, 192 Pronoun to Agreement Cycle 10, 130, 156, 168, 195 PRT see participle; particle PST see past Putnam, Michael 45, 53 Pye, Clifton 18 Pysz, Agnieszka 180–181 Q see question QUA see qualifier qualifier 43 question 14, 18–20, 39, 49; markers 108, 110, 190 Quirk, Randolph 76 Raidt, Esther 52 Rauth, Philipp 45, 53 Rautionaho, Paula 67 Raynouard, Marie 44, 177 reconstruction 11–12, 25, 47, 58, 87 REFL see reflexive reflexive 84–85, 89–90, 213 Reflexive Cycle 209 Reh, Mechthild 5, 35, 209 Reinöhl, Uta 206 REL see relative relative 35, 118 Relative Cycle 209 renewal: of the demonstrative 46, 53; deontic modals 42, 44, 75, 82; loss and 26, 46, 54, 73, 82, 204;

253

negative 98, 107, 124, 194; process 6–8, 13, 36; pronoun 128, 156; study 15–16 Rice, Keren 104 Rijk, Rudolf de 72 Rissanen, Matti 114 Roberts, Ian 37, 176, 189 Robins, Robert H. 27, 29 Roehrs, Dorian 45, 53 Romance 12, 30, 33, 44–46, 51, 53, 120, 130, 176–177 Romero, Sergio 96 Rose, Françoise 209 Rosés Labrada, Jorge Emilio 123 Ross, Malcolm D. 67, 87, 153 Roussou, Anna 37, 176, 189 Rupp, Laura 45, 48–49 Rutten, Gijsbert 209 S see singular Saddour, Inés 60–61 SAE see Standard Average European SAI see Subject Auxiliary Inversion Said, Edward 27 Sampson, Tory 209 Sanskrit 29–30, 62, 101, 125, 138–139, 157, 160, 162 Sansò, Andrea 83–85 Sapir, Edward 26, 33, 131 Sasse, Hans-Jürgen 160 Scandinavian 12, 42–43, 50, 53, 83–84, 179 Scheffer, Johannes 66 Schlegel, August Wilhelm von 26, 44, 132 Schlegel, Friedrich von 29–30 Schiering, René 6, 21, 176 Schwegler, Armin 27, 131–133 Scivoletto, Giulio 120 second person 122, 147, 149, 196; pronouns 24, 121, 126, 140, 145–146, 153, 163, 196, 200; singular 151; subjects 164 semantic: content 6–7, 18, 20, 58, 111, 118; feature 16, 193; impact 2, 97; loss 5 Semantic-Pragmatic Cycles 117–118 Seržant, Ilja 140, 142, 200–202, 209 Shakespeare, William 23, 66 Shibaska, Reijirou 121 Shields, Kenneth 54 Siewierska, Anna 21, 82, 140–142, 148, 155, 215 Simon-Vandenbergen, Anne-Marie 120 Sims, Lynn 201

254

Index

singular 121, 139, 143, 151–153, 159; pronoun 84, 215; see also first person Skeat, Walter 23 Slade, Benjamin 62 Slobodchikoff, Tatyana 209 Smith, Aaron 7 Smith, Adam 26–28 Smitterberg, Erik 66 sociolinguistic 16, 177, 201 Song, Jae Jung 121 Specht, Franz 159 Speyer, Augustin 45, 53 stages: cycle 41, 156; flexionless 2, 11; older 10, 25, 63; synthetic 2, 136, 202 Standard Average European 177, 203 Starosta, Stanley 87 Stassen, Leon 201 Steinbach, Markus 209 Stiernhielm, Georg 27, 29 Strang, Barbara 66 Streitberg, Wilhelm 67–68, 73 SU see subject SUBJ see subjunctive mood subject: agreement 88, 102, 123, 138, 141; gender 54; marking 82–83, 130; pronouns 18, 141, 148, 151, 154, 168, 202; real 57, 66, 69, 196 Subject Agreement Cycle 140, 142, 195, 200 Subject and Object Pronoun Cycles 19 Subject Auxiliary Inversion 76, 80, 82 subjunctive mood 73 Sullens, Idelle 23 Summerer, Lotte 16, 39, 183, 185, 187 Sundquist, John 157 Sweet, Henry 23 syntax 4, 13, 16, 24, 118, 152; morphology 35, 43, 185, 190 synthetic: language 1, 4, 26, 30, 41, 129, 131–132, 135–136, 210; stages 2, 136, 139, 202; see also analytic systematic 3, 22, 27, 30, 37–38, 53, 208, 210 Szmrecsanyi, Benedikt 19, 132, 134–135 T see tense Tagliamonte, Sally 45, 48–49 Tanana (Lower/Upper) 43 Tauli, Valter 2, 26–27, 34–35, 159–160

tense: past 58–59, 69, 73; present 54, 58–60, 63, 73, 187 tense, mood, and aspect 9, 204–205; cycles 9, 90, 100, 201 Tesnière, Lucien de 106, 136 third person 21, 84, 104, 143, 150–152, 193, 196; pronoun 46, 53–55, 140, 163 Thomason, Sarah 179 Thompson, Sandra 53, 161 Thorpe, Benjamin 23 Tiedmann, Dieterich 173 TMA see tense, mood, and aspect TNS see tense Tohono O’odham 23, 46, 175 Traugott, Elizabeth 6–7, 16–17, 27, 34–36, 39, 121, 183, 185, 187, 207 Trips, Carola 157 Trousdale, Graeme 183 Turner, Ralph Lilley 102 Twain, Mark 23 Uhlenbeck, C.C. 159 v see light verb V see verb van de Velde, Mark 209 van der Auwera, Johan 2, 4, 9, 18–19, 96, 107, 207 van der Wal, Marijke 209 van Riemsdijk, Henk 20 verb: copula 55–56, 104; inflected 134; unaccusative 84–86; volition 9, 17, 75, 77 Veselinova, Ljuba 9, 18, 98 Viberg, Åke 56 Vincent, Nigel 162 Vindenes, Urd 45, 50 Visconti, Jacqueline 2, 6, 117 Visser, Fredericus Theodorus 56, 66, 70 Voice Cycle 13, 41, 83–85, 87, 90, 199, 205 Vossen, Frens 4, 9, 18–19, 96, 107 Walkden, George 107, 111, 125 Walker, Jim 72 Waltereit, Richard 106 weakening: general 4, 6–8, 14, 18, 95; phonetic 24, 77, 97, 145, 176, 204 Whitman, John 122

Index

Whitney, Willian Dwight 26, 30–31, 102 Wilcox, Phyliss and Sherman 209 Willis, David 4, 15, 26, 36, 96, 209 Wolff, John 87 Wood, Johanna 9, 209 word: negative 201, 211; order 3–4, 14, 19, 128–129, 157

Yang, Hui-Ling 105 Žaucer, Rok 77–78 Ziegeler, Debra 66 Zieglschmid, Friedrich 71 Zribi-Hertz, Anne 142 Zúñiga, Fernando 82

255