objects and information structure - wordpress.com · objects and information structure ... j....

O B J E C T S A N D I N F O R M AT I O N S T R U C T U R E

In many languages, the objects of transitive verbs are either marked bygrammatical case or agreement on the verb, or they remain unmarked: thisis differential object marking. This book is a cross-linguistic study of howdifferential object marking is affected by information structure, the structur-ing of the utterance in accordance with the informational value of its elementsand contextual factors. Marked objects tend to be associated with old infor-mation or information that the sentence is about, while unmarked objectstend to express new information. The book also sheds light on grammaticalpatterning in languages with differential object marking: in some languagesmarked and unmarked objects have identical grammatical properties, whereasin other languages marked objects are more active in syntax. Finally, it pro-vides a theory of the historical changes that led to the emergence of variouspatterns of differential object marking.

M A RY D A L RY M P L E is Professor of Syntax in the Faculty of Linguistics,Philology and Phonetics at the University of Oxford.

I R I N A N I K O L A E VA is a lecturer in Endangered Languages in the Depart-ment of Linguistics at the School of Oriental and African Studies, Universityof London.

https:/www.cambridge.org/core/termshttps://doi.org/10.1017/CBO9780511993473https:/www.cambridge.org/core

In this series

99 D I A N E B L A K E M O R E : Relevance and linguistic meaning: the semanticsand pragmatics of discourse markers

100 I A N R O B E R T S A N D A N N A R O U S S O U : Syntactic change: aminimalist approach to grammaticalization

101 D O N K A M I N K O V A : Alliteration and sound change in early English102 M A R K C . B A K E R : Lexical categories: verbs, nouns and adjectives103 C A R L O T A S . S M I T H : Modes of discourse: the local structure of texts104 R O C H E L L E L I E B E R : Morphology and lexical semantics105 H O L G E R D I E S S E L : The acquisition of complex sentences106 S H A R O N I N K E L A S A N D C H E R Y L Z O L L : Reduplication:

doubling in morphology107 S U S A N E D W A R D S : Fluent aphasia108 B A R B A R A D A N C Y G I E R A N D E V E S W E E T S E R : Mental spaces

in grammar: conditional constructions109 H E W B A E R M A N , D U N S T A N B R O W N A N D G R E V I L L E G .

C O R B E T T : The syntax-morphology interface: a study of syncretism110 M A R C U S T O M A L I N : Linguistics and the formal sciences: the origins of

generative grammar111 S A M U E L D . E P S T E I N A N D T . D A N I E L S E E L Y : Derivations in

minimalism112 P A U L D E L A C Y : Markedness: reduction and preservation in phonology113 Y E H U D A N . F A L K : Subjects and their properties114 P . H . M A T T H E W S : Syntactic relations: a critical survey115 M A R K C . B A K E R : The syntax of agreement and concord116 G I L L I A N C A T R I O N A R A M C H A N D : Verb meaning and the lexicon:

a first-phase syntax117 P I E T E R M U Y S K E N : Functional categories118 J U A N U R I A G E R E K A : Syntactic anchors: on semantic structuring119 D . R O B E R T L A D D : Intonational phonology, second edition120 L E O N A R D H . B A B B Y : The syntax of argument structure121 B . E L A N D R E S H E R : The contrastive hierarchy in phonology122 D A V I D A D G E R , D A N I E L H A R B O U R A N D L A U R E L J .

W A T K I N S : Mirrors and microparameters: phrase structure beyond freeword order

123 N I I N A N I N G Z H A N G : Coordination in syntax124 N E I L S M I T H : Acquiring phonology: a cross-generational case-study125 N I N A T O P I N T Z I : Onsets: suprasegmental and prosodic behaviour126 C E D R I C B O E C K X , N O R B E R T H O R N S T E I N A N D J A I R O

N U N E S : Control as movement127 M I C H A E L I S R A E L : The grammar of polarity: pragmatics, sensitivity,

and the logic of scales128 M . R I T A M A N Z I N I A N D L E O N A R D O M . S A V O I A :

Grammatical categories: variation in romance languages129 B A R B A R A C I T K O : Symmetry in syntax: merge, move and labels130 R A C H E L W A L K E R : Vowel patterns in language131 M A R Y D A L R Y M P L E A N D I R I N A N I K O L A E V A : Objects and

information structure

Earlier issues not listed are also available


CAMBRIDGE STUDIES IN LINGUISTICS

General Editors: P. AUSTIN, J . BRESNAN, B. COMRIE,S. CRAIN, W. DRESSLER, C. J . EWEN, R. LASS,

D. LIGHTFOOT, K. RICE, I . ROBERTS, S. ROMAINE, N. V. SMITH

Objects and Information Structure


O B J E C T S A N DI N F O R M AT I O NS T R U C T U R E

M A R Y D A L R Y M P L EUniversity of Oxford

I R I N A N I K O L A E VAUniversity of London


C A M B R I D G E U N I V E R S I T Y P R E S S

Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore,So Paulo, Delhi, Dubai, Tokyo, Mexico City

Cambridge University PressThe Edinburgh Building, Cambridge CB2 8RU, UK

Published in the United States of America by Cambridge University Press, New York

www.cambridge.orgInformation on this title: www.cambridge.org/9780521199858

c Mary Dalrymple and Irina Nikolaeva 2011

This publication is in copyright. Subject to statutory exceptionand to the provisions of relevant collective licensing agreements,no reproduction of any part may take place without the writtenpermission of Cambridge University Press.

First published 2011

Printed in the United Kingdom at the University Press, Cambridge

A catalogue record for this publication is available from the British Library

Library of Congress Cataloguing in Publication dataDalrymple, Mary. Objects and information structure / Mary Dalrymple,Irina Nikolaeva.

p. cm. (Cambridge studies in linguistics; 131)ISBN 978-0-521-19985-8 (hardback)1. Grammar, Comparative and generalSyntax. 2. Grammar, Comparative andgeneralTopic and comment. 3. Semantics. 4. Focus (Linguistics)I. Nikolaeva, Irina. II. Title.P291.D35 2011415dc22

2011012721

ISBN 978-0-521-19985-8 Hardback

Cambridge University Press has no responsibility for the persistence oraccuracy of URLs for external or third-party internet websites referred toin this publication, and does not guarantee that any content on suchwebsites is, or will remain, accurate or appropriate.


Contents

Preface and Acknowledgements page 11List of Abbreviations 13

1 Introduction 11.1 The phenomenon 11.2 Previous work 2

1.2.1 Marking as distinguishing arguments 21.2.2 Marking as coding features 51.2.3 DOM in transformational syntax 8

1.3 Criteria for marking 111.4 Our proposal 13

1.4.1 Information structure 141.4.2 Syntax 171.4.3 Diachrony and grammaticalisation 181.4.4 Limits of our analysis 19

1.5 Structure of the book 202 Syntactic assumptions 22

2.1 Grammatical functions 222.1.1 Grammatical function diagnostics 232.1.2 Grammatical functions and semantic roles 25

2.2 Objects 262.3 Levels of syntactic representation 282.4 Describing linguistic structures 322.5 Grammatical agreement 342.6 Agreement and pronominal incorporation 362.7 Casemarking 392.8 Nonsyntactic critera for casemarking patterns 412.9 Conclusion 44

3 Information structure in grammar 453.1 The content of information structure 453.2 Information-structure roles 47

7


8 Contents

3.2.1 Focus 473.2.2 Topic 483.2.3 Topicality and topic-worthiness 503.2.4 Secondary topic 53

3.3 Conclusion 574 Syntax and information structure 58

4.1 Alternative views of information structure 584.1.1 Tree-based representations 584.1.2 Early work in LFG 61

4.2 Information structure and its role in grammar 654.2.1 Content of information structure: The LFG view 654.2.2 Linguistic encoding of information structure

relations 674.2.3 Information structure in relation to semantics 68

4.3 Our architecture 714.3.1 Glue and the syntax-semantics interface 714.3.2 Information structure features and relations 754.3.3 Levels and equations 784.3.4 A short text 864.3.5 Information structure and its place in grammar 90

4.4 Conclusion 935 Topicality and grammatical marking 94

5.1 Topical marking for different grammatical functions 945.2 Grammatical encoding of topical subjects 995.3 Subjects and topichood 1025.4 Conclusion 106

6 Topical marking of nonsubjects 1076.1 Casemarking of topical nonsubjects 107

6.1.1 Persian 1076.1.2 Tariana 1136.1.3 Topical nonsubject casemarking 115

6.2 Agreement with topical nonsubjects 1166.2.1 Itelmen 1166.2.2 Tabassaran 1206.2.3 Topical nonsubject agreement 122

6.3 Conclusion 1247 Topicality and DOM 125

7.1 Objects as grammaticalised secondary topics 1257.2 Agreement with topical objects: Tundra Nenets 127


Contents 9

7.3 Casemarking of topical objects 1327.3.1 Tigre 1337.3.2 Dolakha Newar 137

7.4 Conclusion 1398 Primary and secondary objecthood and DOM 140

8.1 Grammatical marking and grammatical function 1408.2 Object agreement and grammatical function: Ostyak 1428.3 Object casemarking and grammatical function 150

8.3.1 Mongolian 1508.3.2 Chatino 1558.3.3 Hindi 159

8.4 Objects and markedness 1648.5 Conclusion 167

9 Multiple objects and grammatical alignment 1699.1 On the typology of multitransitive constructions 1699.2 No ditransitive construction 172

9.2.1 Nenets 1729.2.2 Ostyak 1739.2.3 Mongolian 175

9.3 Ditransitive constructions 1779.3.1 Chatino 1779.3.2 Hindi 179

9.4 Other multitransitive constructions 1809.4.1 Dolakha Newar 1819.4.2 Tigre 1859.4.3 Applicatives: Upper Necaxa Totonac 186

9.5 Topicality and goals 1919.6 Conclusion 193

10 Semantic features, topicality and grammaticalisation 19410.1 Case studies 194

10.1.1 Uralic 19410.1.2 Persian and the Iranian languages 20110.1.3 Hindi and the Indo-Aryan languages 203

10.2 Paths of grammaticalisation 20710.2.1 Spreading of DOM 20810.2.2 Narrowing of DOM 211

10.3 Towards a typology 21510.4 Conclusion 217


10 Contents

11 Conclusion 219

References 223Author Index 240Language Index 243Subject Index 245


Preface and Acknowledgements

Our work on this topic began in connectionwith our participation in the AHRCproject Noun Phrase Agreement and Coordination, a joint project which in-volved, besides us, Louisa Sadler, Ryo Otoguro, and Aline Villavicencio at theUniversity of Essex (AHRC grant MRG-AN10939/APN17606, 20042006).We are grateful to Louisa, Ryo, and Aline for helpful comments and discus-sion as our ideas developed over the course of the project.Earlier versions of the work were presented at the 2005 International Lex-

ical Functional Grammar Conference (LFG05) in Bergen, the 2005 Linguis-tic Association of Great Britain conference in Cambridge, the 2005 Colloquede Syntaxe et Semantique a Paris (CSSP 2005), the University of Leipzig inDecember 2007, the University of Manchester in April 2008, the workshopArctic Languages: Syntax,Morphology, Lexicon at the University of Tromsin September 2009, the workshop Variation and Change in Argument Realisa-tion at the University of Naples in May 2010, the Surrey Linguistics Circle inDecember 2010, and the State of the Art Seminar on Differential Object Mark-ing and Information Structure at the University of Konstanz in December 2010.An early version of our work appeared as Dalrymple and Nikolaeva (2005),though it will be clear to readers of that paper that our ideas have evolvedconsiderably since its publication. We have profited from discussion of theseearlier versions with Farrell Ackerman, Ash Asudeh, David Beck, BalthasarBickel, Miriam Butt, Elizabeth Coppock, Greville Corbett, David Cram, An-nahita Farudi, Ron Kaplan, Tracy Holloway King, Aditi Lahiri, Olivia Lam,Louise Mycock, Tara Mohanan, Louisa Sadler, Devyani Sharma, Nigel Vin-cent, Yogendra Yadava, and Annie Zaenen.For very helpful comments on drafts of this book as it took shape, we thank

Farrell Ackerman, Ash Asudeh, Miriam Butt, Dolgor Guntsetseg, Dag Haug,Sigve Berge Hofland, Bozhil Hristov, Nazareth Kifle, Aditi Lahiri, LouiseMycock, and Louisa Sadler. We are grateful to our language consultants fortheir generous assistance: Ghazaleh Kad and Shamsi Saber for Persian, Dol-gor Guntsetseg and Tuya Bavuu for Khalkha Mongolian, Galina Koreneva andAnna Lamdo for Tundra Nenets, and Devyani Sharma and Tara Mohanan for

11

https:/www.cambridge.org/core/termshttps://doi.org/10.1017/CBO9780511993473.001https:/www.cambridge.org/core

12 Preface and Acknowledgements

Hindi. Nazareth Kifle provided crucial assistance with the Tigre data. Finally,we thank our husbands, Ken Kahn and Simon Carne, for their love and supportduring the writing of this book.


List of Abbreviations

We have regularised some of the glosses and the transcription of the originalsources. We use the following abbreviations:

Abs absolutiveAcc accusativeAdv adverbAll allativeAN action nominalArt articleAsp aspectAttr attributiveAug augmentativeAux auxiliaryC completiveCaus causativeCl classifierCom comitativeConv converbCop copulaDat dativeDecl declarativeDet determinerDim diminutiveDir directiveDur durativeDu dualErg ergativeExpl expletive

Ez ezafeFem feminineFoc focusFut futureGen genitiveHab habitualHon honorificImp imperativeImpf imperfectiveImprs impersonalIndef indefiniteIndic indicativeInf infinitiveInstr instrumentalInterj interjectionIntr intransitiveIrr irrealisLoc locativeMasc masculineMod modalNeg negationNMLZ nominalisationNom nominativeObj object marker or

object agreement

Obl oblique agreementP potentialPart participlePas passivePerf perfectivePl pluralPoss possessivePostess postessivePrep prepositionPres presentProg progressiveProh prohibitivePron pronominalPurp purposiveQ questionRel relative form or

relative pronounRep reportativeSg singularSpec specificSubj subject agreementSub subordinateSupess superessiveTop topicTr transitive

13


1 Introduction

1.1 The phenomenon

Many languages exhibit non-uniform grammatical marking of objects. Varia-tions can occur within one and the same language with objects of one and thesame verb. For example, in Turkish (Altaic)1 the object of the same verb eithertakes the accusative suffix or remains unmarked:2

(1) a. AliAlibirone

kitab-1book-Acc

ald1buy.Past.3Sg

Ali bought a (certain) book.b. AliAlibirone

kitapbook

ald1buy.Past.3Sg

Ali bought a book. (Enc 1991:5)

In Palauan (Austronesian), the object of the same verb either does or does nottrigger agreement on that verb:

(2) a. Te-illebed-iiSubj.3Pl-Perf.hit-Obj.3Sg

athebilisdog

atherengalekchildren

The kids hit the dog.b. Te-illebedSubj.3Pl-Perf.hit

athebilisdog

atherengalekchildren

The kids hit a dog/the dogs/some dog(s). (Woolford 2000:5)

Such patterns are widely known under the rubric of differential object mark-ing or DOM (a term introduced by Bossong 1985).We understand DOM as covering both agreement and casemarking (case or

adpositional marking on the object). Though we recognise that agreement andcasemarking differ both historically and synchronically, as noted by Comrie(1979) and Croft (1988:167168), among many others, we believe that theyshare commonalities in the context of DOM, and we will use the cover term1 Language families are provided according to the classifications in Ethnologue (Grimes 1999).2 References are provided for examples that do not come from our own fieldwork.

1


2 Introduction

(grammatical)marking to refer to them. This approach is in line with Nichols(1986), who analyses agreement and casemarking as alternative strategies forencoding the relation between the head and a dependent, as well as some gen-erative literature, where case and agreement are inherently linked.The aim of this book is to provide a new view of DOM which encompasses

syntactic, semantic, and information-structural differences between markedand unmarked objects. We will make the following claims:

Marked objects are associated with the information-structure role of topic.The association may be either synchronic or historical.Where the direct con-nection between marked objects and topicality has been lost through gram-maticalisation,3 marked objects in some languages become associated withsemantic features typical of topics (animacy, definiteness, specificity).

In some languages, marked and unmarked objects display an identical be-havioural profile and can be assigned to the same grammatical function.Other languages distinguish syntactically between marked and unmarkedobjects: marked objects are primary objects, while unmarked objects aresecondary objects. This reflects the tendency for topical arguments to ap-pear high on the grammatical function hierarchy.

We begin our discussion with a review of previous work; we then present theessential aspects of our claims, and conclude this chapter with an overview ofthe book.

1.2 Previous work

DOM has been studied from a formal, generative perspective as well as afunctional-typological perspective, and has been discussed and analysed byLazard (1984), Bossong (1985, 1991), de Hoop (1992), Aissen (2003a,b),Nss(2004), and de Swart (2007), among many others. Many of these analyses con-centrate either on differential object agreement or on differential object case-marking, including both case and adpositional marking.

1.2.1 Marking as distinguishing argumentsAnalyses of grammatical marking (and in particular casemarking) have of-ten appealed to two types of functional motivation: coding/indexing and dis-criminatory/disambiguating/distinguishing (Comrie 1979, 1989, de Hoop3 We use the term grammaticalisation in an extended sense, to refer to a process wherebymarking indicating a pragmatic contrast comes to be associated with syntactic or semanticrather than pragmatic features: cf. the broad understanding of grammaticalisation as the waygrammatical forms arise and develop through space and time (Heine 2004:575).


1.2 Previous work 3

and Narasimhan 2005, de Hoop and Malchukov 2007, Nss 2007, Malchukov2008, and others).Discriminatory/disambiguating casemarking serves to distinguish between

different categories: for example, between the two arguments (the subject andthe object) of a transitive clause. It encodes the relation between two argu-ments rather than the properties of an individual argument. The discriminatoryfunction of casemarking has been argued to provide a functional motivationfor the fact that in many languages casemarking is missing on the single argu-ment of intransitive verbs and on one of the two arguments of transitive verbs.Silverstein (1976) and Comrie (1977) argue that since the basic purpose offormal marking on core arguments is to distinguish the subject from the ob-ject, the need to overtly mark the object is greater in some cases than in othersbecause an object with subject-like semantic properties for example, an an-imate, specific, or definite object is more likely to be confused with thesubject. Therefore, objects whose semantic features are typical of subjects aremore likely to be overtly marked. This approach relies on the concept of thetransitive prototype, in which the object is prototypically inanimate, indefinite,and/or nonspecific (Comrie 1989), and maintains that the function of DOM isto signal deviation from the prototype. It also stands in conformance with thewidespread functional view that infrequent (and therefore functionallymarked)categories receive more formal marking, whereas frequent (and therefore func-tionally unmarked) categories tend to remain formally unmarked; the explana-tion for this is based on economy considerations (Haiman and Thompson 1985,Du Bois 1987) and the relation between functional and formal markedness. Onthis view, DOM is essentially iconic: formal marking on objects reflects theirstatus as atypical or infrequent objects, and thus their functional markedness.The idea that marking serves to distinguish or differentiate between argu-

ments of a predicate has been pursued in depth in the influential work of Ais-sen (2003a,b). In her approach, languages may appeal to different factors inDOM, but in all cases the resulting patterns reflect the tension between twofunctional principles: (i) iconicity between functional and formal markednessof objects, and (ii) economy, the pressure to avoid excessive marking. Follow-ing much work in functional typology (Silverstein 1976, Givon 1976, Comrie1977, 1979, 1989, Du Bois 1987, Dixon 1994), Aissen claims that there isa prototypical association involving grammatical functions and features suchas animacy, humanness, definiteness and specificity/referentiality. Subjects areprototypically high in these features, while objects are low. In other words,properties that are unmarked for subjects are marked for objects, a relationknown asmarkedness reversal. On this view, unmarked subjects are animate,human, definite and specific, while marked subjects are inanimate and/or non-


4 Introduction

specific. For objects, the opposite markedness patterns are at work: inanimateand indefinite/nonspecific objects are unmarked, while nonhuman definite an-imate objects are more marked, and human objects are most highly marked.Aissen formalises these correlations as Optimality Theoretic constraints.

Referential properties of animacy, humanness, definiteness and specificity areorganised into two Prominence Scales, the Animacy Scale and the DefinitenessScale.

(3) Animacy Scale: Human > Animate > InanimateDefiniteness Scale: Personal pronoun> Proper name > Definite NP >

Indefinite specific NP > Nonspecific NP

Humans are located higher on the Animacy Scale than nonhuman animates,which in turn are higher than inanimates, and so on. In addition, Aissen in-troduces a binary Relational Scale, where the subject outranks the object, aswell as several harmonic (or markedness) hierarchies representing the relationbetween the Prominence Scales and the Relational Scale. The harmonic align-ment constraints for animacy and definiteness features are displayed in (4) and(5), respectively.

(4) *Su/Inan *Su/Anim *Su/Hum

*Oj/Hum *Oj/Anim *Oj/Inan

(5) *Su/NSpec *Su/Spec *Su/Def *Su/PN *Su/Pro

*Oj/Pro *Oj/PN *Oj/Def *Oj/Spec *Oj/NSpec

The most highly ranked constraints in (4) penalise inanimate subjects and hu-man objects; the constraints ranked one step lower penalise animate nonhumansubjects and animate nonhuman objects, and so on. The definiteness alignmentconstraints in (5) work similarly.These hierarchies predict the most and least marked patterns of subject and

object marking across languages. Constraints higher on the hierarchy incurmore costly violations than constraints lower on the hierarchy. This means thatif an object at a certain point in the hierarchy is overtly marked, then any objectthat is higher on the relevant scale will also be overtly marked. DOM ariseswhen some but not all objects are marked; this is implemented by correlatingharmonic constraint hierarchies with the constraint *STRUCC, motivated bythe needs of economy, which penalises a value for the morphological categoryCASE below a certain point on the hierarchy.The position of *STRUCC in the hierarchy leads to different patterns of ob-

ject marking across languages. If *STRUCC dominates all the constraints on


1.2 Previous work 5

both scales, then marking is banned for all objects. If *STRUCC is ranked at thelowest point on the hierarchy, all objects receive grammatical marking. Suchlanguages do not display DOM. In languages with DOM, object marking canbe based either on the Animacy Scale or on the Definiteness Scale. For exam-ple, Aissen (2003b) shows that in Catalan (Romance) only personal pronounobjects are casemarked. This is captured in an Optimality Theoretic accountby positioning *STRUCC lower in the Definiteness Scale than the constraintpenalising pronominal objects. Similarly, if *STRUCC is ranked lower thanthe top-ranked constraint *Oj/Hum in the Animacy Scale, casemarking is pe-nalised for all objects except the most highly ranked human objects. Accordingto Aissen (2003b:456), such languages are difficult to find, although there aremany languageswhere marking is penalised for all objects except animates (in-cluding humans): for example, Sinhala (Indo-Aryan). Further demotion of theeconomy constraint *STRUCC yields other patterns of object marking. In Pit-jantjatjara (Pama-Nyungan), pronominal and proper name objects are marked,while other objects, including definite objects, are unmarked; this is treated bypositioning *STRUCC below the constraints penalising marking on pronom-inal and proper name objects. In Hebrew (Semitic), only definite objects re-quire the object marker et, while indefinite objects are always unmarked; thismeans that the economy constraint is ranked lower than the constraint penalis-ing definite objects. Simultaneous reference to multiple features involves morecomplicated multidimensional crossing of the scales, but the basic principleremains the same.These pioneering proposals have inspired much subsequent work and dis-

cussion, including an exploration of patterns that do not fit neatly into Aissenscross-linguistic picture. For example, Yang and van Bergen (2007) argue thatin Mandarin Chinese (Sino-Tibetan), objects that are obligatorily marked inthe ba-construction are either highly prominent in terms of animacy or, sur-prisingly, low in prominence in terms of definiteness; for a small category ofobjects in-between, casemarking is optional. Importantly, the effect of animacyand definiteness is only noticeable on scrambled objects; Yang and van Bergenpropose that the syntactic position of the object introduces an additional di-mension into the prominence hierarchy which can influence marking. Otherworks following on from Aissens work, including Morimoto (2002), proposevarious readjustments of the original constraint hierarchy, but do not questionthe general spirit of the prototype deviation approach to DOM.

1.2.2 Marking as coding featuresThe coding/indexing perspective on marking differs from the discriminatoryperspective in that marking is taken to signal specific semantic and/or prag-


6 Introduction

matic properties of the relevant argument, rather than a particular relation be-tween one argument and another. In fact, de Hoop and Narasimhan (2005)claim that a purely discriminatory function for casemarking is rare, and thatin most instances casemarking serves to signal some aspect of the interpre-tive content of a phrase. This is particularly clear for obliques, where casecan bear an important semantic load in signalling the meaning contributionof the casemarked phrase. In the analysis of DOM, the coding/indexing per-spective assumes that the presence of marking is connected to the presence ofcertain properties of the object. This view goes back to Hopper and Thomp-sons (1980) classic study, in which DOM is taken to be one of the signalsindicating high transitivity, rather than deviation from a transitive prototype.Nss (2004, 2006, 2007) argues that Aissens approach contradicts the es-

tablished notion of transitivity, and proposes that the prototypical transitiveclause is one in which the two participants are maximally semantically dis-tinct. Her definition of semantic distinctness includes several parameters, butthe parameter that is especially relevant for cross-linguistic patterns of DOMis affectedness, understood roughly as involvement in the verbal event andchange of state of the participant as a direct result of this event. On this view,prototypical objects are those that are highly affected by the transitive event.Formally marked objects are not functionally marked; instead, they are proto-typical highly affected and individuated objects, which tend to receive moregrammatical marking than non-prototypical, less affected objects.Other researchers have also appealed to affectedness as a factor in analyses

of DOM: for example, Cetinoglu and Butt (2008) discuss the role of affected-ness in object casemarking in Turkish. However, the relevance of affectednessfor DOM has been questioned by de Hoop (2008), who shows that in manycases object marking is present in sentences in which the object participant isnot affected by the verbal event. For instance, in Mandarin Chinese the objectmarker ba, often treated as an instance of DOM, is required on objects of theverbs forget and lose, although the forgotten or lost thing is not usuallyaffected by the forgetting or losing event. Nss (2004) claims that definite ob-jects are more affected than indefinite ones because the event affects the wholerather than a part (cf. I drank the milk and I drank some milk). However, inmany cases it is difficult to argue that definite or animate objects are more af-fected than indefinite or inanimate ones, if affectedness is understood in termsof a change of state. De Hoop cites the following Hindi (Indo-Aryan) examplesfrom Mohanan (1994), involving the object marker ko:


1.2 Previous work 7

(6) a. Ilaa-neIla-Erg

haarnecklace

ut.haayaalifted.Past.MascSg

Ila lifted a necklace.b. Ilaa-neIla-Erg

haar-konecklace-Obj


Ila lifted the necklace. (de Hoop 2008, citing Mohanan 1994:80)

According to de Hoop, there is no reason to think that the necklace that ispicked up in (6b) is more affected than the necklace that is picked up in (6a),even though the former is definite and the latter is indefinite.In fact, de Hoop and Narasimhan (2005), Nss (2007), and Malchukov

(2008) note that in the case of DOM, disambiguating and indexing approacheslead to roughly the same result (though they make different predictions with re-spect to differential subject marking). Affectedness normally presupposes highindividuation of the referent, while individuation in its turn is associated withdefiniteness. Animacy may also be relevant for affectedness because the ef-fect of an action on an animate entity is more salient for human cognition thanthe effect on an inanimate entity and is more likely to arouse empathy. Thus,affectedness is ultimately operationalised in terms of the same features ofprominence as are relevant on the disambiguating/discriminatorymarking per-spective: when a language decides what kinds of objects are affected enoughto be marked, it can choose on the basis of more easily measurable propertiessuch as animacy and definiteness.De Swart (2006, 2007) proposes an analysis which, in a sense, combines

the discriminatory and coding approaches. His model is based on the idea thatthe speaker takes the perspective of the addressee in order to ensure recov-erability of the intended interpretation. In some instances this implies thatthe speaker chooses to mark an object, rather than leaving it unmarked, whenhe/she wishes to emphasise a certain feature of the object: for instance, def-initeness in Hindi. If the speaker intends to highlight definiteness in order toensure that the addressee will interpret the object as definite, accusative koappears. If the speaker does not want to force a definite interpretation of theobject, casemarking is omitted. Thus, marking on the object is the result of aprincipled decision on the part of the speaker and has a coding function.Recoverability presupposes semantic distinctness between two arguments

(cf. Nss 2004). This explains why casemarking on objects can be influencedby the semantic properties of the subject and vice versa. However, de Swartdoes not account for these patterns in terms of transitivity and, unlike Ais-sen, does not appeal to prototypical features of subjects and objects. He illus-


8 Introduction

trates his analysis with data from languages which seem to best support thedisambiguating/discriminatory view of marking. In these languages, DOM isprimarily determined by the need to differentiate the subject from the object.De Swart argues that in such languages, sentences with no semantic contrastbetween agent and patient must show a morphological contrast between them,in order to avoid ambiguity. The relevant semantic dimensions involve familiarprominence features, but are largely language-specific. In Malayalam (Dravid-ian), for example, object marking is mostly found on animate objects. How-ever, in sentences which can in principle be interpreted in two different ways,it is also found on inanimate objects; the reason seems to be that without overtcasemarking, the sentence would be ambiguous.

(7) a. kappalship.Nom

tiramaalakal-ewaves-Acc

bheediccusplit.Past

The ship broke through the waves.b. tiramaalakalwaves.Nom

kappal-ineship-Acc

bheediccusplit.Past

The waves split the ship.(de Swart 2007, citing Asher and Kumari 1997)

Such systems are called global because marking depends on properties ofthe subject, properties of the object, and the relation between them, along thelines of the discriminating/disambiguating view of marking. In contrast, lo-cal systems are those in which the presence of object marking is only depen-dent on the features of the object itself, along the lines of the coding/indexingapproaches. As de Swart notes, global systems present a problem for Aissensmodel, since they depend on the simultaneous consideration of properties ofthe subject and object rather than properties of the object alone, but can beaccounted for within the framework of Bidirectional Optimality Theory.

1.2.3 DOM in transformational syntaxMany analyses of phrasal syntax within the transformational paradigm assumetwo distinct positions for objects, VP-internal and VP-external, and postu-late a correlation between object position and object marking (Diesing 1992,Dobrovie-Sorin 1994, van Geenhoven 1998, Torrego 1998, Ritter and Rosen2001, Woolford 1999, 2000, 2001, de Hoop 1992, among others). The dis-tinction is generally analysed as semantically driven, dependent on a specificvs. a nonspecific interpretation of the object, and it is generally assumed thatVP-internal indefinite/nonspecific objects are syntactically less visible thanVP-external definite/specific ones.


1.2 Previous work 9

De Hoop (1992) observes that a morphological difference in case in lan-guages with DOM is linked to a semantic difference in the interpretation ofindefinites. Her analysis has clear similarities to work by Diesing (1992) onthe interpretation of indefinite objects: Diesing suggests that arguments uni-versally excluded from VP-internal positions are specific (or, more precisely,presuppositional), while VP-internal arguments are interpreted as nonspecific.This is formally modelled by assuming that strong NPs must move out ofthe VP, and that existential closure applies at the VP level to weak, nonspe-cific indefinite NPs. De Hoop (1992) assumes two categories of NPs, strong(presuppositional) and weak. Strong NPs are anchors in conversation; theyare semantically characterised as generalised quantifiers and include referen-tial, partitive, and generic expressions. Weak NPs are analysed as existentiallyquantified. Additionally, there are two kinds of object case: Weak Case, as-signed VP-internally at D-structure, and Strong Case, assigned at S-structureto [Spec,AgrO]. A strong NP moves out of the VP to get Strong Case, whileWeak Case requires syntactic adjacency to the verb. Overt accusative markingon VP-external objects is analysed as the morphological realisation of abstractStrong Case. This is exemplified, for instance, in Turkish, as shown in (1),where the marked and unmarked object receive specific and nonspecific inter-pretations, respectively.These works deal only with casemarking, but since case and verbal agree-

ment are treated as two aspects of the same phenomenon in this framework,roughly similar analyses have been proposed for differential object agreement.Both case assignment/checking (depending on ones syntactic assumptions)and agreement are treated in terms of movement of the object to the spec-ifier position of the relevant agreement head. In object-agreement languages,agreement serves as a specificity licenser, as argued byMahajan (1992), amongothers.Subsequent work has made it clear that specificity is not the only feature

responsible for DOM. Adopting the premise that VP-internal and VP-externalobject positions may be associated with different semantic properties, Wool-ford (1999, 2000, 2001) aims to explain which objects occupy which of thesepositions and why, taking into account more complex patterns where there isno single semantic feature that triggers movement out of the VP and agree-ment. Instead, a VP-external object may have any of several distinct clustersof features.The basic premise of Woolfords proposal is that economy keeps objects

in their base VP-internal positions unless that would incur a violation of theExclusion Principles. Exclusion Principles are modified versions of Diesingsmapping principles, which exclude NPs bearing certain features from the VP-


10 Introduction

internal object position. Woolford assumes a family of Exclusion Principlesbased on different semantic features including specificity, humanness, ani-macy, and number; on her view, these are separate principles and cannot bereduced to one more general principle. Each Exclusion Principle can be in-dependently active in a language. The economy principle Avoid Movement,which prohibits moving objects out of their base position (Chomsky 1995),ensures that objects remain within the VP unless this violates one or moreExclusion Principles. Cross-linguistic differences in restrictions on agreementare dealt with in terms of different rankings of various Exclusion Principlesand the economy principle Avoid Movement.In some languages DOM seems to depend on aspectual features of the verb.

The idea that object marking correlates with aspect has been explored byRamchand (1997), Meinunger (1998), Woolford (2000) and others. Woolford(2000) claims that aspect in Palauan determines the ranking of Exclusion Prin-ciples which govern object casemarking. Ritter and Rosen (2001) provide amore sophisticated analysis, arguing that in languages with DOM the splitin object marking is determined by the presence or absence of the feature[QUANT(ISATION)]. Quantised objects (their Class I objects) must check theirQUANT feature, forcing such objects to move out of the VP and triggeringagreement or casemarking, while nonagreeing or noncasemarked objects (theirClass II objects) are not quantised and remain within the VP. The specificsemantic contribution of the feature differs from language to language. When[QUANT] is an inherent feature of the verb, it has aspectual meaning: it en-codes delimitedness or boundedness of the event. According to Ritter andRosen (2001), this situation is exemplified in Finnish (Uralic), where objects ofbounded events stand in the accusative case and objects of unbounded eventstake the partitive case; in Palauan, where object agreement correlates withboundedness of the event (as expressed through perfectivity); and in MandarinChinese, where the ba-construction is only possible with delimited events. Onthe other hand, when [QUANT] is a feature of the functional head Agr, it isuninterpretable and lacks inherent semantic content. In this case DOM is notsensitive to event type; instead, [QUANT] enters into a checking relation withdefinite/specific/animate objects, as in Turkish, Hebrew or Bantu. Ritter andRosen (2001) do not explain why different semantic types are involved in ob-ject split in these languages, but emphasise the importance of treating [QUANT]as a feature which bears on the interpretation of verbs as well as objects, andwhich can be realised either as object case or object agreement.


1.3 Criteria for marking 11

1.3 Criteria for marking

As we saw in the previous section, most existing work on DOM appeals toreferential semantic features of the object noun phrase, such as animacy, defi-niteness, or specificity, to distinguish marked and unmarked objects. Aissens(2003a, 2003b) work is based on hierarchies of animacy and definiteness. ForNss (2004, 2006, 2007), the crucial notion of affectedness correlates withreferential properties of the object, although the correlation is indirect. Mosttransformational work on DOM is based on the premise that object markingpatterns are defined in terms of semantic features (Diesing 1992, van Geen-hoven 1998, and others). Work by Woolford (1999, 2000, 2001) relies on afamily of Exclusion Principles based on specificity, humanness, animacy, andnumber. Ritter and Rosen (2001) use a general notion of boundedness whichencompasses specificity and definiteness as well as event-boundedness.These criteria are indeed useful in analysing patterns of DOMwhere objects

that are characterised as semantically strong show more agreement with theverb, or more casemarking, than objects without these properties. For exam-ple, semantic factors have been argued to provide the clearest explanation ofpatterns of DOM in Turkish or Hebrew, where the distribution of marked andunmarked objects is fairly straightforward and definable in terms of referentialfeatures: Turkish marks specific objects, while Hebrewmarks definite objects.4

Yet these factors do not adequately account for languages in which objects withthe same semantic features can be either marked or unmarked. Such variationwas noted by Bossong (1991:152), who concluded that in these languages therules of DOM cannot be formulated precisely, but must allow for a certaindegree of variability across speakers and situations.Take, for example, Hindi. As discussed by Mohanan (1994:79 ff.), Singh

(1994), and Aissen (2003b), the object marker ko in modern Hindi appears onspecific human objects (in some dialects, all specific animate objects), whetherdefinite or indefinite, but is impossible on indefinite inanimate nouns.4 However, von Heusinger and Kornfilt (2005) and Kornfilt (2009) argue that the correlationbetween specificity and accusative marking in Turkish is imperfect, and Danon (2006) claimsthat the definiteness condition on casemarking in Hebrew is a purely syntactic condition,related to the presence of an article, rather than a semantic property.


12 Introduction

(8) a. Ilaa-neIla-Erg

bacce-ko/*baccachild-Obj/child


Ila lifted the/a child.b. Ravii-neRavi-Erg

kaccaaunripe

kelaabanana

kaat.aacut.Past.MascSg

Ravi cut the/a unripe banana.c. Ravii-neRavi-Erg

kacceunripe

kele-kobanana-Obj

kaat.acut.Past.MascSg

Ravi cut the/*a unripe banana.(Aissen 2003b, from Mohanan 1994)

In other words, for specific human/animate nouns ko-marking is obligatory,while for inanimate indefinites, marking is disallowed. However, with inani-mate/nonhuman definite objects, apparent optionality arises: such objects areeither marked or unmarked.On Aissens analysis of Hindi, the constraints that disallow ko on inanimate

nonspecific objects and indefinite inanimates are dominated by *STRUCC,which penalises a value for the morphological category CASE, so casemark-ing is absent. On the other hand, constraints that penalise the absence of caseon human/animate objects dominate *STRUCC, so accusative marking is re-quired. Optionality of marking on inanimate/nonhuman definite objects is han-dled in terms of constraint reranking: the relative ranking of *STRUCC againstthe constraints that penalise the absence of case on definite nonhuman or inan-imate objects is undetermined. On this view, the patterns of grammatical mark-ing are basically random and would be expected to reflect idiolectal and/or freevariation.On De Swarts (2007) analysis, definiteness in Hindi does not trigger case-

marking but is rather a side effect of the use of casemarking. De Swart claimsthat this eliminates optionality; however, his theory does not lead to straightfor-ward and testable predictions for the distribution of ko, and it remains unclearhow it can be formalised in an explicit grammar. Most importantly, no clearevidence is provided for the hypothesis that the presence of the accusative cor-relates with the speakers intention to highlight definiteness.Many languages exhibit similar patterns of optionality. In Sinhala, inan-

imate objects are never marked, whereas animate objects are optionally case-marked.

(9) mam@I

miniha(-w@)man-(Acc)

daekkasaw

I saw the man. (Gair and Paolillo 1997:32)


1.4 Our proposal 13

In Kannada (Dravidian), accusative marking is obligatory on animate and/ordefinite objects but optional on indefinite inanimates (Lidz 2006). Casemarkedindefinite objects are specific, while noncasemarked objects are ambiguousbetween a specific and nonspecific interpretation.

(10) a. NaanuI.Nom

pustakabook

huDuk-utt-idd-eenelook.for-Nonpast-be-1Sg

I am looking for a book/a certain book.b. NaanuI.Nom

pustaka-vannubook-Acc

huDuk-utt-idd-eenelook.for-Nonpast-be-1Sg

I am looking for a certain book. (Lidz 2006:11)

Lidz (2006) shows that the specific vs. nonspecific interpretation of noncase-marked objects correlates with their syntactic position, but not exactly in theway predicted by de Hoops (1992) analysis. Casemarked objects have a spe-cific reading even when they occur inside the VP, while noncasemarked objectsmay occur VP-externally, in which case they must receive a specific reading.Apparent optionality is also observed in languages with differential object

agreement. In Ostyak (Uralic), object agreement may but need not appear withdefinite as well as indefinite objects, as shown in (11) (see Nikolaeva 1999,2001 and the discussion of Ostyak in Chapter 8, Section 8.2):

(11) tamthis

kalaNreindeer

we:[email protected]

/ we:[email protected]

He killed this reindeer.a:mujsome

kalaNreindeer


/ we:[email protected]

He killed a/some reindeer.

These examples demonstrate that in some languages, semantic features suchas animacy, definiteness, or specificity alone are not enough to account forthe distribution of agreement or casemarking on objects. Further, there is noevidence that the properties of the subject participant (as in global systems)or the degree of affectedness play a role. Optionality is observed with exactlythe same subjects and exactly the same verbs.

1.4 Our proposal

We approachDOM from a different perspective.We believe that in many cases,the seeming unpredictability of variation in DOM is due to the limited set offeatures examined in previous work, and that a better account is possible if ad-ditional factors are taken into account. In particular, variations in marking can


14 Introduction

often be explained by reference to information structure, a level of sentencegrammar where propositions, as conceptual states of affairs, are structured inaccordance with the informational value of sentence elements and contextualfactors. We propose a theory of DOM which emphasises the role of informa-tion structure in the marking patterns of objects. The information-structuraldistinctions we make explain many of the unexplained areas in semantic anal-yses of DOM.

1.4.1 Information structureSpecifically, we propose that topicality plays an important role in DOM:markedobjects are often topical, while unmarked objects are nontopical. For example,in Ostyak, object agreement is required for topical objects:

(12) a. What did he do to this reindeer?b. tamthis

kalaNreindeer


/ *we:[email protected]

He killed this reindeer.

Nontopical, focused objects disallow agreement:

(13) a. Which reindeer did he kill?b. tamthis

kalaNreindeer


/ *we:[email protected]

He killed this reindeer.

This is true independent of semantic features of animacy and definiteness; inboth of these examples, the object is animate and definite.Following Lambrecht (1994) and others, we understand topicality as a prag-

matic relation that holds between a referent and the proposition expressed byan utterance: topicality has to do with the construal of the referent as prag-matically salient, so that the assertion is made about this referent. Topicalityis not an inherent property of a referent, and although it correlates with therole played by the referent in the preceding discourse, the correlation is im-perfect. It cannot be unambiguously established on the basis of the referentialfeatures of the object either; rather, it depends on the speakers assessment ofits saliency within a given communicative context, as we discuss in Chapter 3.Since we view information structure as part of grammar, we treat grammati-cality as including pragmatic well-formedness. For instance, object agreementis ungrammatical in (13) only if it is construed as an answer to the questionWhich reindeer did he kill?, though it is possible in the context in (12). In or-der to demonstrate the pragmatic role of a referent, we provide examples incontext whenever possible.


1.4 Our proposal 15

In some respects, our view most closely resembles the proposal of Nss(2004): like ours, her analysis relies not on the semantic features of the object,but rather on its perceptual role. In fact, Nss (2004) suggests that affected-ness, which plays a crucial role in her treatment of DOM, has two aspects:expressing a part/whole relation, and salience. Salience is a necessary propertyof topics, so the two notions, though distinct, seem to be closely related (seeCroft 1991:155 for discussion of the correlation between affectedness and top-icality). However, for Nss (2004) salience has to do with which entities areof more interest for human perception in general. Her analysis does not, then,address contextual factors. On our approach, salience relates to the role playedby a referent in the pragmatic structure of the proposition, established within agiven context.Our proposal provides a different functional motivation for DOM from the

discriminatory/disambiguating perspective: we argue that DOMwas originallymotivated by the need to highlight similarities between subjects and topicalobjects, which tend to be grammatically marked, as opposed to nontopical ob-jects. That is, it arises from the need to give overt expression to a property thatis common to subjects and (some) objects but less typical of other non-coregrammatical functions: languages tend to mark topics, whether subjects or ob-jects, either by agreement or by casemarking. This means that topical objectsare not functionally marked, atypical, or uncommon; in fact, we suggest thatobjects are just as likely to be topical as nontopical.The idea that topicality may play a role in DOM has been mentioned in

previous work (Aissen 2003b, Leonetti 2003, and others). However, it has notreceived extensive elaboration in the family of functional-typological and Op-timality Theoretic analyses of DOM, nor has it been standardly incorporatedinto most existing generative or transformational treatments. Some researchemploys the notion of prominence interpreted more broadly than AissensProminence Scales (de Hoop and Narasimhan 2005, de Hoop and Lamers2006, de Hoop 2008). De Swart (2007:138) defines prominence as a featurethat is concerned with the centrality of an entity in the discourse or with thereadiness with which an entity presents itself to the speaker as a topic of con-versation. His analysis distinguishes two levels of prominence: (i) discourseprominence, which reflects the status of the argument in the discourse, and (ii)semantic prominence, which depends on the intrinsic semantic features of theargument. The notion of discourse prominence seems to be comparable withour notion of topicality; however, de Swart does not provide a detailed char-acterisation of discourse prominence, noting only that semantic prominenceinfluences discourse prominence because an inherently prominent element ismore likely to be topical.


16 Introduction

Another exception is Meinunger (1998, 2000): working within the Mini-malist Program, he suggests that the distinction between topical and nontopicalobjects is manifested syntactically, and that the difference is formally encodedin terms of association with Agr. He also assumes an inherent relationship be-tween verbal agreement and Case, which are seen as two aspects of the samephenomenon, in that both are effects of the movement of an argument to thespecifier position of an agreement head. Languages differ in how they markscrambled topical objects: some languagesmark topical and nontopical objectswith morphologically different cases, while in other languages the differencein the interpretation of the objects is linked to object agreement, and in a thirdtype of language the difference is only positional. Meinunger views topical ar-guments as discourse-linked/familiar and contextually salient. In addition, thespeaker must intend to construe them as topics, as opposed to presentational fo-cus or comment, which normally contains discourse-new elements. However,his definition of topic and the relationship between topicality and the refer-ential features of the object are not clearly spelt out. For instance, Meinungerclaims that topicality is realised as specificity in Turkish and animacy in Hindi,but does not explain this difference, though object marking clearly depends onvery different semantico-pragmatic conditions in these languages.Escandell-Vidal (2009) presents a detailed analysis of DOM by means of

the preposition a in Balearic Catalan (Romance). Although differential objectcasemarking in many varieties of Catalan is explained by appeal to semanticfeatures of animacy or definiteness, in Balearic Catalan referential propertiesof the object play a secondary role. Pronominal objects are always casemarked,but for lexical objects casemarking is (partly) determined by topicality. As inour analysis, Escandell-Vidal understands topicality as aboutness, and there-fore Balearic Catalan seems to behave very similarly to a number of the lan-guages we discuss in this book. Her analysis differs from ours in that she doesnot discuss the alignment of grammatical functions of marked and unmarkedobjects with information-structure roles, though she shows that there are posi-tional differences between topical and nontopical objects, since Balearic Cata-lan is a language with designated topic and focus positions. Topical objectscannot occur in the canonical focus position, and instead must be either left-or right-dislocated. However, only discourse-old topics, and not all topical ob-jects, are casemarked. This implies that both marked and unmarked objectscan occur in topic position, and raises the question of whether they displayidentical syntactic behaviour.


1.4 Our proposal 17

1.4.2 SyntaxThis brings us to the second point. Our view differs frommost previous propos-als in another respect: we believe that existing typological analyses of DOM,which mainly concentrate on its functional motivation and the cross-linguisticdistribution of morphological marking, do not pay sufficient attention to thesyntactic side of the phenomenon. The behavioural properties and syntacticstatus of marked vs. unmarked objects are rarely discussed, and to our knowl-edge, there are few systematic cross-linguistic studies of the differences be-tween them.We have found that in some languages marked and unmarked objects do not

differ syntactically and, arguably, realise the same grammatical function, theobject. In other languages they not only have different information-structureroles, but also exhibit different behavioural syntactic profiles. For example,marked and unmarked objects in Ostyak differ in their ability to control coref-erence with the subject of an action nominal dependent clause, allow possessortopicalisation, control possessor reflexivisation, and launch floated quantifiers.In other languages as well, a number of syntactic tests distinguish marked ob-jects and unmarked objects, with the marked object displaying a larger numberof properties associated with core arguments. We suggest that in these lan-guages, marked and unmarked objects bear different grammatical functions.Our proposal is cast within the framework of Lexical Functional Grammar(LFG), which distinguishes between the primary objectOBJ and the secondaryor semantically restricted object OBJ (Bresnan and Kanerva 1989). Thesetwo types of objects are usually discussed in connection with double objectconstructions, but in many languages, both types are available in single ob-ject constructions as well, with the choice between them determined by theirinformation-structure role.As mentioned above, many analyses of DOM assume two distinct positions

for objects, VP-internal and VP-external, and posit a correlation between theposition of the object and grammaticalmarking (Diesing 1992,Dobrovie-Sorin1994, Ramchand 1997, van Geenhoven 1998, Torrego 1998, Ritter and Rosen2001, Woolford 1999, 2000, 2001, de Hoop and Malchukov 2007, amongothers). Issues of word order and the positions of objects have been studiedthoroughly within the LFG framework by Butt and King (1996), Choi (1999)and many other researchers: their work has clearly shown that information-structure role can be relevant for word order constraints, and that informationstructure has a strong effect on where objects can appear. However, we donot posit a direct relation between DOM and phrase structure position in alllanguages, because at least in some of the languages we analyse there is no ob-


18 Introduction

vious connection between the position of the object, its grammatical marking,and its grammatical function. More generally, we do not assume that gram-matical functions or information-structure roles must be identified configura-tionally. LFG analyses grammatical functions as syntactic primitives, and doesnot define them in terms of phrase structure position. Generalisations concern-ing the behaviour of different grammatical functions or arguments bearing dif-ferent information-structure roles can be easily modelled within LFGs pro-jection architecture (Kaplan 1987, Asudeh 2006), which allows constraintsto refer to and relate different aspects of the structure of an utterance.

1.4.3 Diachrony and grammaticalisationOur analysis also provides a diachronic explanation for why referential seman-tic features such as animacy, definiteness, and specificity often play a role inDOM. We propose that these features are involved in grammaticalisation oftopichood.Diachronic evidence shows that DOM often originates as a marking device

for topics. We take this situation to be historically primary for some languages.One possible direction of change involves spreading of topical marking. In alanguage with topical object marking, the marking can generalise or extend,applying to nontopical objects which have features typical of topics. When thishappens, marking patterns become automatic consequences of distinctions atother levels of structure. As a result, the role of information structure in objectmarking is diminished: marking is obligatory not only for topical objects, butalso for objects with certain referential semantic features, independent of theirinformation-structure role.The opposite direction of change involves narrowing of topical marking:

marking becomes specialised for topics which bear certain semantic features.In the relevant languages, only a subset of topical objects are formally marked,while nontopical objects remain unmarked. Narrowing usually involves themost typical members of the set. Objects ranked high on the prominence hier-archies are frequent topics, so topical marking can become restricted to them.At the next stage of grammaticalisation, the connection to information struc-ture may be completely lost, so that object marking becomes dependent onsemantic features alone.Thus, we propose that in many languages, different patterns of DOM arise

as a result of different directions of grammaticalisation of topic marking onobjects. Features that are typical of topics come to be required, or sufficient,for object marking.


1.4 Our proposal 19

1.4.4 Limits of our analysisOur theory of DOM does not extend to all instances of variable marking ofobjects. Since we are interested in differences in interpretation that dependon the presence or absence of grammatical marking, we only consider lan-guages where a subset of objects is grammatically marked (by either case oragreement), while another subset receives no marking at all, as in the examplescited above. We will not address those instances of DOM where objects cantake two alternative casemarkers, as attested, for example, in Icelandic (NorthGermanic):

(14) a. Hannhe.Nom

kloraDiscratched

migI.Acc

He scratched me.b. Hannhe.Nom

kloraDiscratched

merI.Dat

He scratched me. (Nss 2007:89, citing BarDdal 2001:146)

The Icelandic accusative-dative alternation has been reported to reflect a seman-tic contrast: in (14a) the scratching is perceived as a forceful act of violence,possibly painful for the patient participant, while in (14b) the interpretationinvolves volitionality on the part of the patient. In other languages, such asRussian (Slavic) or Finnish, the case of the object differs depending on aspec-tual characteristics of the verb. As interesting as these phenomena may be, webelieve that they deserve separate treatment and do not necessarily fall underthe same generalisations as DOM in languageswhich contrast formallymarkedand unmarked objects.The focus of this book is the grammatical function of object, not the ex-

pression of the patient/theme argument. We limit the scope of our study toobject marking casemarking or agreement in order to more fully explorethe relation between the presence and absence of marking and the status ofobject arguments at different levels of linguistic representation. We excludelanguages where some patient/themes are not syntactic objects at all, but par-ticipate in various detransitivising constructions and remain syntactically un-expressed. In other words, we only deal with proper syntactic arguments. Like-wise, we do not deal with instances where a patient/theme argument is realizedas a grammatical function other than object. In this way, our work differs fromsome previous research on DOM such as, for example, Nss (2007), whichtreats alternative encodings of patient/theme arguments independently of theirsyntactic status.Furthermore, our analysis does not address the distribution of verbal cli-


20 Introduction

tics or incorporated pronouns, although in many languages they are associatedwith some but not all objects, as discussed by Bresnan and Mchombo (1987)for Chichewa (Central Bantu), Culy (2000) for Takelma (Penutian), Bowern(2004) for Bardi (Nyulnyulan), Donohue (2004) for Tukang Besi (Austro-nesian), and Jaeger and Gerassimova (2002) and Jaeger (2004) for Bulgarian(Slavic), among others. The status of object markers in a number of Bantu lan-guages is debatable (see, for example, Seidl and Dimitriadis 1997 and Mori-moto 2002 on Swahili and Woolford 2001 on Ruwund), so we will not discusssuch languages unless we can demonstrate that the marking on the verb is actu-ally agreement and not pronominal incorporation (see Bresnan and Mchombo1987 and Chapter 2, Section 2.6 for detailed discussion of this issue).We also limit our scope to the examination of nominative-accusative lan-

guages, and have little to say about languages with an ergative-absolutive case-marking system, where the object is absolutive. Malchukov (2006) claims thatDOM is more typical of nominative-accusative systems, and proposes a prin-ciple of argument marking that holds of both types of languages: languagestend to avoid manipulation of the casemarking of the unmarked argument, i.e.the nominative argument in nominative-accusative systems, and the absolutiveargument in ergative-absolutive systems. Deemphasis of the absolutive objecttends to give rise to voice alternations such as antipassive, rather than case al-ternations. We do not take a stand here on whether this view is correct, but thefact remains that DOM seems to be more frequent in nominative-accusativelanguages, and we will mainly concentrate on such languages in this book.

1.5 Structure of the book

The book is structured as follows.Chapter 2 gives a brief introduction to LFG (Bresnan 2001, Dalrymple 2001,

Falk 2001), the syntactic framework employed in this book. Chapter 3 presentsan overview of the model of information structure which we assume, and inChapter 4 we present our proposal for the treatment of information structure inLFG.In Chapter 5 we discuss the expression of what we refer to as primary

topics in syntax, suggesting that although there is no unique alignment be-tween information-structure roles and grammatical functions, there are impor-tant cross-linguistic tendencies in the grammatical expression of primary top-ics.Chapter 6 addresses languages in which grammatical marking for nonsub-

ject topics occurs with a variety of syntactic roles: it is possible for objects, pos-sessors, instruments, and other syntactic functions. Comrie (2003) discusses


1.5 Structure of the book 21

agreement patterns in such languages under the rubric of trigger-happy agree-ment, and we show that casemarking can work in a similar manner. These dataconstitute the main evidence for the relevance of topichood in the grammaticalmarking of nonsubjects.Differential object agreement and differential object casemarking are dealt

with in Chapter 7, where we examine languages in which grammatical markingof nonsubject topics is syntactically restricted to objects. In such languages, (asubset of) topical objects receive special agreement or casemarking, while (asubset of) nontopical/focused objects remain unmarked. However, an exami-nation of the grammatical properties of unmarked and marked objects in theselanguages indicates that they represent the same grammatical function.We then take a closer look at syntactic differences between marked and un-

marked objects in languages in which grammatical marking is confined to ob-jects, and discuss cross-linguistically typical patterns of alignment betweensyntax and information structure. We show in Chapter 8 that in some such lan-guages, a number of syntactic tests distinguish marked objects and unmarkedobjects, with the marked (topical) object displaying more properties associatedwith core grammatical functions. Chapter 9 suggests some generalisations con-cerning ditransitive constructions in languages with DOM, and demonstratesthat although various semantic roles may correspond to the primary object inmultiple object constructions, DOM tends to be associated with topicality.Chapter 10 proposes a historical explanation for semantic patterns of DOM,

which can arise as a result of different directions of grammaticalisation ofnonsubject topic marking. Since features such as definiteness, specificity oranimacy are known to be associated with topics, grammatical marking mayspread to nontopical objects with these semantic features, or may narrow toinclude only topical arguments which also have these features.Chapter 11 concludes the book.


2 Syntactic assumptions

Nontransformational, constraint-based theories of grammar such as Construc-tion Grammar (Croft 2001, Goldberg 2006), HPSG (Sag et al. 2003), Roleand Reference Grammar (Van Valin 2003), Simpler Syntax (Culicover andJackendoff 2005), and Lexical Functional Grammar (Bresnan 2001, Dalrymple2001, Falk 2001) represent different aspects of the structure of an utterance asseparate but related grammatical modules. Such theories assume that syntac-tic structure is related to semantics, information structure, and other linguisticlevels not by means of transformational operations, but by constraints involv-ing one or more levels of structure. Information-structure roles may be associ-ated with particular phrase structural positions, but these positions do not de-fine the roles (as they often do in tree-based, transformational theories such asPrinciples and Parameters Theory or Minimalism: Chomsky and Lasnik 1993,Chomsky 1995). Instead, the relations between grammatical, semantic, andinformation structural roles are specified in terms of constraints involving dif-ferent levels of linguistic representation.In LFG, these different aspects of linguistic information are represented by

structures that may be of different formal types. The phrasal structure of thesentence is represented by a phrase structure tree, the constituent structure orc-structure. Grammatical functions like subject and object are represented bythe functional structure or f-structure. Information structure (Choi 1999,Butt and King 1996, 2000) is related to other grammatical levels within theprojection architecture of LFG (Kaplan 1987, Asudeh 2006). LFG providesthe tools needed to analyse the relation between these structures and to give aformal treatment of the grammar of object marking and its relation to semanticsand information structure.

2.1 Grammatical functions

A central tenet of LFG theory is the dual nature of syntactic structure. Theabstract grammar of all languages is organised around syntactic relations likesubject, object, and adjunct: all languages distinguish arguments from adjuncts,

22


2.1 Grammatical functions 23

and subjects from nonsubjects, for example. These abstract grammatical rela-tions are overtly expressed in phrasal structure in very different ways in dif-ferent languages. Some languages have rigidly fixed word order, while othersallow a great deal of freedom in word order, and languages vary in terms ofwhere the verb must appear in relation to its arguments. The level at whichword order and phrasal constituency is represented has its own internal or-ganisational principles, different from the level at which abstract grammati-cal relations are represented. The two syntactic structures of LFG, constituentstructure and functional structure, reflect this duality.

2.1.1 Grammatical function diagnosticsLFG assumes that abstract grammatical functions such as subject, object, com-plement, and adjunct are theoretical primitives, not determined by other levelsof structure. That is, the subject of a sentence is not defined as a phrase appear-ing in a particular phrase structural position or as an entity bearing a particularsemantic role; instead, subject, object, and other grammatical functions arebasic concepts of the grammar, around which various grammatical propertiestend to cluster (Falk 2006).In LFG, diagnostics identifying particular grammatical functions appeal to

abstract behavioural grammatical properties defined at functional structure. Inmany languages, for example, the subject is the argument which agrees withthe verb, or is required or disallowed as the antecedent of a reflexive or pro-noun, or is absent in a subordinate clause under identity with a matrix clauseargument. Mohanan (1994) shows that binding relations for pronouns in Hindiare constrained by the grammatical function of the antecedent. The antecedentof a non-reflexive pronoun cannot be the subject of its clause:

(1) Vijay-neVijay-Erg

Ravii-koRavi-Obj

uskiihis

saikilbicycle

parLoc

bit.haayaasit.Caus.Perf

Vijayi seated Ravij on hisi,j bike. (Mohanan 1994:126)

Matsumoto (1996) discusses honorification in Japanese (Japonic), showingthat some verbs take special forms to honour the subject referent, and not othergrammatical functions. The verb form o-V ni naru is used to honour the subjectsensei teacher, and cannot be used to honour a nonsubject:



(2) a. senseiteacher

waTop

honbook

oAcc

o-yomiHon-read

niCop

narimashitabecome.Polite.Past

The teacher read a book. (Matsumoto 1996:27)b. *JonJohn

waTop

senseiteacher

nobyo-tasukerareHon-help

niCop

nattabecome.Past

John was saved by the teacher. (Matsumoto 1996:28)

None of these properties is definitional of subjecthood, and indeed all of theseproperties have been found to be associated with nonsubjects as well as sub-jects, depending on the language. In fact, these properties tend to target the topof the grammatical function hierarchy in (3), originally proposed by Keenanand Comrie (1977), and so distinguish subjects from other arguments in somelanguages, and core arguments from noncore arguments, or arguments fromadjuncts, in other languages.

(3) subject > object > oblique> adjunct

Because the subject is the argument that stands at the head of the grammaticalfunction hierarchy, it is often the only grammatical function that can participatein certain processes, and can be reliably identified in this way.Diagnostics targeting nonsubject grammatical functions, specifically objects,

also vary from language to language. For example, some languages have objectagreement, which can function as a test for objecthood, and some languageshave a special relativisation strategy for objects that is different from the strat-egy used for other grammatical functions.Diagnostics for grammatical functions that are stated in terms of nonsyntac-

tic criteria are more difficult to use, and less reliable. Unlike transformationaltheories, LFG does not assume that grammatical functions are defined in termsof positions in a phrase structure tree. Thus, in many languages the appear-ance of a phrase in a particular position is not criterial for identification of thephrase as bearing a particular grammatical function, although in some fixedword order languages, a particular phrase structure position may be reservedfor a particular grammatical function. Grammatical function alternations arealso unreliable tests; it is often claimed that objects can be identified as thosearguments of an active verb which appear as the subject of the correspondingpassive verb, but although this is often true, there are passive subjects whichare not objects of the active verb (Bill is said to be available/*Someone saidBill to be available). For more discussion of grammatical function diagnostics,see Keenan (1976), Andrews (1985) and, for objects, Hudson (1992); for anLFG perspective, see Dalrymple (2001: Chapter 2) and, for objects, Borjarsand Vincent (2008) and Lam (2008).


2.1 Grammatical functions 25

2.1.2 Grammatical functions and semantic rolesPredicates encode an association between grammatical functions and particu-lar semantic roles: for example, the active version of the verb kick associatesits subject with the semantic role of kicker, and its object with the semanticrole of the entity that is kicked (The man kicked the ball). Mapping theory(Levin 1986, Bresnan and Kanerva 1989, Butt 1995, Alsina 1996) is a theoryof the relation between argument structure (the level at which semantic rolesare represented) and functional structure (the grammatical functions associatedwith those roles).Researchers in mapping theory agree that argument structure encodes a cer-

tain amount of semantic information about the arguments of a predicate, butdifferent researchers have adopted different views of the content and represen-tation of argument structure. Much research is based on the assumption thatargument structure makes available a set of abstract semantic roles like agent,source, theme, patient, and goal, which are ranked relative to one another on athematic hierarchy. Bresnan and Kanerva (1989) propose the thematic hier-archy shown in (4):

(4) Thematic hierarchy (Bresnan and Kanerva 1989):

AGENT > BENEFACTIVE> RECIPIENT/EXPERIENCER> INSTRUMENT> THEME/PATIENT > LOCATIVE

These semantic roles are systematically related to grammatical functions (see,for example, Bresnan and Kanerva 1989 and Bresnan and Zaenen 1990): onthis view, mapping theory relates the agent role (kicker) of kick to SUBJ, andthe patient role (object kicked) to OBJ. Other versions of mapping theory adopta different view of argument structure, often in an attempt to define or pro-vide a clearer characterisation of semantic roles. Butt (1995) assumes Jack-endoffs (1983, 1990) Conceptual Semantics representation of argument struc-ture, which involves the use of a set of conceptual primitives such as CAUSE,GO, and BE to encode basic semantic relations. Alsina (1996) works within aversion of the proto-role theory of Dowty (1991), which provides a set of en-tailments classifying the arguments of a predicate according to the degree towhich they exhibit proto-agent or proto-patient properties. Our theory of ar-gument marking does not depend on the details of these theories of argumentstructure, and so we will appeal when necessary to semantic roles like agent,patient, theme, source, and goal, without presupposing any particular theory oftheir representation or definition at argument structure.Whatever theory of argument structure is adopted, LFG researchers agree

that mapping theory must include several very general principles governing



the relation between argument structure and functional structure. A commonly-assumed principle is function-argument biuniqueness (Bresnan 1980), whichrequires each semantic role to correspond to exactly one grammatical function,and no grammatical function to correspond to more than one semantic role.1

For example, there are no predicates whose argument structure contains twodifferent semantic roles (an agent and a theme, for example) with both rolesmapped to a single grammatical function, and there are no predicates whoseargument structure contains a single semantic role mapped to two differentgrammatical functions:

(5) Function-argument biuniqueness:

ALLOWED: DISALLOWED: DISALLOWED:Argument structure: 1 2 1 2 1

Functional structure: GF1 GF2 GF1 GF1 GF2

Another general constraint is the subject condition, which stipulates that averbal predicate must have a subject. This is a more controversial conditionwhich has been challenged by some researchers (for discussion, see Berman1999), and it may be that the subject condition holds in only some languagesrather than being a universal principle of mapping theory. Though (as far as weknow) the subject condition holds in all of the languages that we will discuss,we will not take a position on the universality of the subject condition in thiswork.

2.2 Objects

Since our focus is the differential marking of objects, we will be most con-cerned with object grammatical functions. Here, we provide a brief overviewof the standard LFG view of objects. Recent LFG work has explored the objectfunctions from a variety of perspectives: see Lam (2008) for a thorough dis-cussion of object functions with a focus on constructions with multiple objects,and Borjars and Vincent (2008) for an LFG-based perspective that differs fromthe traditional view in interesting ways, particularly with regard to the analy-sis of the English double object construction. We return to a discussion of theobject functions and object mapping in Chapter 9.LFG assumes that more than one object function can be manifested in a

language. In early LFG, these object functions were called OBJ (object) and1 The principle of function-argument biuniqueness has been challenged by some researchers, inparticular by Alsina (1996) and Mohanan (1994).


2.2 Objects 27

OBJ2 (second object: Bresnan 1980, Kaplan and Bresnan 1982). In a doubleobject construction, the object which shares more properties with the objectof a monotransitive verb was called the (primary) object, OBJ, and the otherargument was called the second object, OBJ2.2

In subsequent work, Bresnan and Kanerva (1989) proposed to characterisegrammatical functions in terms of a decomposition of features3 +R and +O,where +R(estricted) indicates whether the grammatical function must corre-spond to a particular semantic role (+R) or may correspond to any semanticrole (R), and +O(bjective) indicates whether the grammatical function is ob-jective (object-like: +O) or not (O). On this view, grammatical functions arecross-classified as shown in (6):

(6) R +R

O SUBJ OBL

+O OBJ OBJ(Bresnan and Kanerva 1989:25)

The subject SUBJ and unrestricted/primary object OBJ can correspond to anysemantic role (R), while the (+R) grammatical functions OBJ and OBL mustcorrespond to particular semantic roles. OBJ stands for the family of object-like (+O) functions OBJTHEME, OBJLOC, and so on, while OBL stands for thefamily of non-object-like (O) oblique functions OBLLOC, OBLGOAL, OBLAGENT,and so on. This decomposition recasts the OBJ/OBJ2 opposition in terms of adistinction between OBJ and the family of grammatical functions abbreviated asOBJ, restricted/secondary object functions that are associated with a particularsemantic role. Languages can vary as to which OBJ functions they express:like many languages, English expresses only OBJTHEME (that is, the OBJ2 ofprevious LFG work is now treated as a restricted OBJ which is required to bearthe semantic role of theme), while other languages may have a wider range ofsemantically restricted objects.With the distinction in hand between OBJ as an unrestricted/primary object

and OBJ as a restricted/secondary object, Bresnan andMoshi (1990) examinedlanguages like Kichaga (Bantu), which allow promotion of oblique argumentsto objects via applicativisation, as in examples like (7):2 Dryer (1986) uses the terms primary object and secondary object to draw nearly the samedistinction; see Chapter 9 for further discussion of LFGs object functions in double objectconstructions.

3 This assumption entails that grammatical functions are not atomic, since they can bedecomposed into simpler featural components; grammatical functions are still primitives ofthe theory, in that they are not defined in terms of concepts from other levels of structure.



(7) a. n-a--ly-aFoc-1Subj-Pres-eat-FinalVowel

k-elya7-food

He is eating food.b. n-a--ly--aFoc-1Subj-Pres-eat-Applicative-FinalVowel

m-ka1-wife

k-elya7-food

He is eating food for the benefit of/to the detriment of his wife.(Bresnan and Moshi 1990:148)

Example (7a) shows the monotransitive verb eat with a single object, the class7 noun food. With the addition of the applicative affix in (7b), a benefac-tive/malefactive applied OBJ wife is added, and food becomes a secondaryOBJTHEME.Bresnan and Moshi (1990) show that in Kichaga, applicative constructions

with more than two objects are possible, establishing conclusively that a simpletwo-way distinction between OBJ and OBJ2 is not sufficient for grammaticaldescription. Consider example (8):

(8) n-a-l!e-ku-sh-k-kor.--aFoc-1Subj-Past-17Obj-8Obj-7Obj-cook-Applicative-FinalVowelShe/he cooked it with them there. (Bresnan and Moshi 1990:151)

This example contains three object markers, representing a locative object, aninstrumental object, and a patient object. The grammatical structure of this ex-ample cannot be analysed using only the grammatical functions OBJ and OBJ2,since this example contains three object-like functions. Bresnan and Moshishow that in (8), the instrumental OBJ is the unrestricted OBJ, while the locativeand patient arguments bear semantically restricted OBJ functions OBJLOC andOBJPATIENT.In much previous LFG literature, the OBJ/OBJ distinction was assumed to

play a role primarily or exclusively in the analysis of multitransitive construc-tions, but we will see in Chapter 8 that the distinction is also relevent for mono-transitive predicates. For further discussion of languages in which monotran-sitive verbs can take either OBJ or OBJ, see Cetinoglu and Butt (2008) forTurkish and Dahlstrom (2009) for Meskwaki (Algonquian).

2.3 Levels of syntactic representation

The following brief sketch presents the formal syntactic

objects and information structure - wordpress.com · objects and information structure ... j....

Documents