Space and Time in Languages and Cultures: Language, culture, and cognition 9789027273604, 902727360X

This is an interdisciplinary volume that focuses on the central topic of the representation of events, namely cross-cult

344 27 12MB

English Pages 363 [376] Year 2012

Report DMCA / Copyright


Polecaj historie

Space and Time in Languages and Cultures: Language, culture, and cognition
 9789027273604, 902727360X

Table of contents :
Space and Time in Languages and Cultures. Language, culture, and cognition
Editorial page
Title page
LCC Data
Table of contents
Editors and contributors
Foreword: Space and time in languages, cultures, and cognition
Introduction: Linguistic, cultural, and cognitive approaches to space and time
1. Preliminary remarks
2. The contributions to the volume
3. Perspectives for future research
I. Linguistic and conceptual representation of events
1. Event-based time intervals in an Amazonian culture
1. Introduction
2. Calendars and time reckoning: Anthropological perspectives
3. Amondawa culture and society: An overview
4. Time intervals in Amondawa language and culture
5. Time and the human lifespan in Amondawa
6. Discussion
2. Vagueness in event times: An epistemic solution
1. Introduction
2. Temporal vagueness vs predicate vagueness
3. What is it that is vague?
4. Selecting a theory for temporal vagueness
5. Conclusions
3. Aspectual coercions in content composition
1. Introduction
2. Background
3. From type presupposition to coercion
4. A sketch of a formal theory of lexical meaning
5. Modality, aspect, and the verbal complex
6. Conclusion
4. Back to the future: Just where are forthcoming events located?
1. Introduction
2. Spatio-kinetic metaphors for time and the role of Ego
3. On the nature of the future tense
4. Future locations
5. Epistemic metaphors
6. Conclusion
II. Cultural perspectives on space and time
5. The “Russian” attitude to time
1. Introduction
2. Semantics of the syntactic phraseme X v X ‘X to X’
3. Pragmatics of the syntactic phraseme X v X ‘X to X’
4. Conclusions
6. Two temporalities of the Mongolian wolf hunter
1. Introduction
2. Pleasing the White Father and producing hiimor’
3. Wolf, hiimor’, and predation
4. Hiimor’ and temporality
5. Concluding remarks
7. Koromu temporal expressions: Semantic and cultural perspectives
1. Introduction
2. Geographic and sociolinguistic setting
3. Time and Koromu morphosyntax
4. Semantic primes and semantic molecules
5. Basic temporal adverbs and temporal primes in Koromu
6. Time, days, and parts of a day
7. Counting the days - a deictic centred system
8. Traditional perspectives on asi ‘months’
9. Conclusion
Appendix 1
Appendix 2
Appendix 3
Appendix 4
8. Universals and specifics of ‘time’ in Russian
1. Introduction
2. The Natural Semantic Metalanguage (NSM) as a tool of semantic description
3. Russian temporal expressions and cultural attitudes to time
4. Concluding remarks
III. Conceptualizing spatio-temporal relations
9. Linguistic manifestations of the space-time (dis)analogy
1. Preliminaries
2. The extensive noun-verb parallelism
3. Asymmetries
4. The multiple roles of time
5. The dynamic conception of space
6. Summation
10. Vectors and frames of reference: Evidence from Seri and Yucatec
1. Introduction
2. Orientation and frames of reference
3. Data collection and methods
4. Seri and Yucatec data
5. Frames of reference and vectors
6. Conclusion
11. Verbal and gestural expression of motion in French and Czech
1. The conceptual category of motion and language diversity
2. Impact of language on the conceptualisation of motion
3. Co-speech gesture
4. The present study and hypotheses
5. Participants
6. Materials
7. Procedure
8. Speech and gesture coding
9. Results
10. Discussion and conclusion
12. Language-specific effects on lexicalisation and memory of motion events
1. Introduction
2. The current study: Recognition and verbalisation with increased memory load
3. Methods and materials
4. Results
5. Discussion
6. Conclusions
13. Space and time in episodic memory: Philosophical and developmental perspectives
1. Introduction
2. Digging deeper: The Kantian roots
3. Why this is a ‘minimalist’ view of episodic memory
4. Objections: This minimalist account of episodic memory is neither (a) about memory, nor (b) episo
5. Testing these ideas
6. The Tim and Tom experiment
7. Discussion and coda
14. Conceptualizing the present through construal aspects: The case of the English temporal construc
1. Introduction
2. Analytical tools: A general introduction
3. The analysis
4. Conclusions
15. From perception of spatial artefacts to metaphorical meaning
1. Introduction
2. Method and material
3. Mental imagery for paths or roads
4. Metaphorical and non-metaphorical uses of path or road
5. Discussion
6. Conclusion
Contents of the companion volume: Linguistic diversity
Name index
Subject index
Language index

Citation preview

Space and Time in Languages and Cultures Language, culture, and cognition Edited by

Luna Filipović University of East Anglia

Kasia M. Jaszczolt University of Cambridge

John Benjamins Publishing Company Amsterdam / Philadelphia

Human Cognitive Processing (HCP)

Cognitive Foundations of Language Structure and Use This book series is a forum for interdisciplinary research on the grammatical structure, semantic organization, and communicative function of language(s), and their anchoring in human cognitive faculties. For an overview of all books published in this series, please see

Editors Klaus-Uwe Panther

Nanjing Normal University & University of Hamburg

Linda L. Thornburg

Nanjing Normal University

Editorial Board Bogusław Bierwiaczonek

University of Economics and Humanities, Poland

Mario Brdar

Josip Juraj Strossmayer University, Croatia

Barbara Dancygier

University of British Columbia

N.J. Enfield

Max Planck Institute for Psycholinguistics, Nijmegen & Radboud University Nijmegen

Elisabeth Engberg-Pedersen University of Copenhagen

Ad Foolen

Radboud University Nijmegen

Raymond W. Gibbs, Jr.

University of California at Santa Cruz

Rachel Giora

Tel Aviv University

Elżbieta Górska

University of Warsaw

Martin Hilpert

Freiburg Institute for Advanced Studies

Zoltán Kövecses

Eötvös Loránd University, Hungary

Teenie Matlock

University of California at Merced

Carita Paradis

Lund University

Günter Radden

University of Hamburg

Francisco José Ruiz de Mendoza Ibáñez University of La Rioja

Doris Schönefeld

University of Leipzig

Debra Ziegeler

Paul Valéry University, France

Volume 37 Space and Time in Languages and Cultures. Language, culture, and cognition Edited by Luna Filipović and Kasia M. Jaszczolt



The paper used in this publication meets the minimum requirements of the American National Standard for Information Sciences – Permanence of Paper for Printed Library Materials, ansi z39.48-1984.

Library of Congress Cataloging-in-Publication Data Space and time in languages and cultures : language, culture, and cognition / edited by Luna Filipovi´c, Kasia M. Jaszczolt. p. cm. (Human Cognitive Processing, issn 1387-6724 ; v. 37) Includes bibliographical references and index. 1. Space and time in language. 2. Psycholinguistics. 3. Cognition. 4. Language and culture. I. Filipovi´c, Luna. II. Jaszczolt, Kasia M. P37.5.S65S56   2012 401--dc23 2012016426 isbn 978 90 272 2391 3 (Hb ; alk. paper) isbn 978 90 272 7360 4 (Eb)

© 2012 – John Benjamins B.V. No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the publisher. John Benjamins Publishing Co. · P.O. Box 36224 · 1020 me Amsterdam · The Netherlands John Benjamins North America · P.O. Box 27519 · Philadelphia pa 19118-0519 · usa

Table of contents Language, culture, and cognition

Editors and contributors Foreword: Space and time in languages, cultures, and cognition Kasia M. Jaszczolt and Luna Filipović Introduction: Linguistic, cultural, and cognitive approaches to space and time Luna Filipović and Kasia M. Jaszczolt

vii xi


Part I.  Linguistic and conceptual representation of events 1. Event-based time intervals in an Amazonian culture Vera da Silva Sinha, Chris Sinha, Wany Sampaio and Jörg Zinken


2. Vagueness in event times: An epistemic solution Minyao Huang


3. Aspectual coercions in content composition Nicholas Asher and Julie Hunter


4. Back to the future: Just where are forthcoming events located? Alan Wallington


Part II.  Cultural perspectives on space and time 5. The “Russian” attitude to time Valentina Apresjan


6. Two temporalities of the Mongolian wolf hunter Bernard Charlier


7. Koromu temporal expressions: Semantic and cultural perspectives Carol Priestley


8. Universals and specifics of ‘time’ in Russian Anna Gladkova



Space and Time in Languages and Cultures: Language, culture, and cognition

Part III.  Conceptualizing spatio-temporal relations 9. Linguistic manifestations of the space-time (dis)analogy Ronald W. Langacker


10. Vectors and frames of reference: Evidence from Seri and Yucatec Jürgen Bohnemeyer and Carolyn O’Meara


11. Verbal and gestural expression of motion in French and Czech Kateřina Fibigerová, Michèle Guidetti and Lenka Šulová


12. Language-specific effects on lexicalisation and memory of motion events Luna Filipović and Sharon Geva


13. Space and time in episodic memory: Philosophical and developmental perspectives James Russell and Jonathan Davies


14. Conceptualizing the present through construal aspects: The case of the English temporal constructions  Grzegorz Drożdż


15. From perception of spatial artefacts to metaphorical meaning Marlene Johansson Falck


Contents of the companion volume


Name index


Subject index


Language index


Editors and contributors

Editors Luna Filipović School of Language and Communication Studies Faculty of Arts and Humanities University of East Anglia Norwich Research Park Norwich, NR4 7TJ, United Kingdom [email protected]

Jürgen Bohnemeyer Department of Linguistics University at Buffalo The State University of New York 609 Baldy Hall Buffalo, NY 14260, USA [email protected]


Bernard Charlier The Mongolia and Inner Asia Studies Unit Department of Social Anthropology University of Cambridge The Mond Building Free School Lane Cambridge CB2 3RT, United Kingdom [email protected]

Valentina Apresjan Department of Linguistics Philological Faculty Higher School of Economics Khitrovsky per., 2/8, korp. 8 109028 Moscow, Russia [email protected]

Vera da Silva Sinha Universidade Federal de Rondônia Departamento de Letras e Lingüística (Lingüística e Línguas Indígenas) CEP: 76801-059 Porto Velho-RO Brazil [email protected]

Nicholas Asher CNRS, Laboratoire Institut de Recherche en Informatique de Toulouse Université Paul Sabatier Toulouse, France [email protected]

Jonathan Davies University of Bristol Faculty of Medicine First Floor South, Senate House Tyndall Avenue Bristol, BS8 1TH, United Kingdom [email protected]

Kasia M. Jaszczolt Department of Theoretical and Applied Linguistics University of Cambridge Cambridge CB3 9DA, United Kingdom [email protected]

viii Space and Time in Languages and Cultures: Language, culture, and cognition

Grzegorz Drożdż Institute of English University of Silesia Ul. Grota-Roweckiego 5 41-205 Sosnowiec, Poland [email protected]

Anna Gladkova School of Behavioural, Cognitive and Social Sciences University of New England Armidale NSW 2351, Australia [email protected]

Kateřina Fibigerová Department of Psychology Faculty of Arts Charles University of Prague Celetná 20 110 00 Praha 1, Czech Republic & Laboratoire Octogone-ECCD Pavillon de la Recherche Université de Toulouse-Le Mirail 5, Allées A. Machado 31058 Toulouse Cedex 9, France [email protected]

Michèle Guidetti Laboratoire Octogone-ECCD Pavillon de la Recherche Université de Toulouse-Le Mirail 5, Allées A. Machado 31058 Toulouse Cedex 9, France [email protected]

Luna Filipović School of Language and Communication Studies Faculty of Arts and Humanities University of East Anglia Norwich Research Park Norwich, NR4 7TJ, United Kingdom [email protected] Sharon Geva Developmental Cognitive Neuroscience Unit UCL Institute of Child Health 30 Guilford Street London WC1N 1EH, United Kingdom [email protected]

Minyao Huang Department of Theoretical and Applied Linguistics Faculty of Modern and Medieval Languages University of Cambridge Sidgwick Avenue Cambridge CB3 9DA, United Kingdom [email protected] Julie Hunter Universite de Pau et des Pays de l’Adour, Pau Institut Jean Nicod and Ecole des Hautes Études de Sciences Sociales Paris, France [email protected] Marlene Johansson Falck Department of Language Studies Umeå University 901 87 Umeå, Sweden [email protected]

Editors and contributors

Ronald W. Langacker Department of Linguistics University of California, San Diego 9500 Gilman Drive La Jolla, CA 92093-0108, USA [email protected]

Chris Sinha Lund University School of Languages and Literature (Linguistics) PO Box 201, 221 00 Lund, Sweden [email protected]

Carolyn O’Meara Seminario de Lenguas Indígenas Instituto de Investigaciones Filológicas Universidad Nacional Autónoma de México Circuito Mario de la Cueva Ciudad Universitaria, 04510 México, D.F. [email protected]

Lenka Šulová Charles University of Prague Faculty of Arts Department of Psychology Celetná 20 110 00 Praha 1, Czech Republic [email protected]

Carol Priestley Languages and Linguistics Griffith University 170 Kessels Road Nathan, QLD 4111 [email protected] James Russell Department of Experimental Psychology University of Cambridge Downing Street Cambridge CB2 3EB, United Kingdom [email protected] Wany Sampaio Universidade Federal de Rondônia Departamento de Letras e Lingüística (Lingüística e Línguas Indígenas) CEP: 76801-059 Porto Velho-RO Brazil [email protected]

Alan Wallington School of Linguistics and English Language Bangor University College Road Bangor Gwynedd LL57 2DG, United Kingdom [email protected] Jörg Zinken University of Portsmouth Department of Psychology King Henry Building King Henry I Street Portsmouth, Hampshire PO1 2DY United Kingdom [email protected]



Space and time in languages, cultures, and cognition

The two volumes comprising Space and Time in Languages and Cultures, published as HCP 36 and HCP 37, originated as a selection of papers from Space and Time across Languages, Disciplines, and Cultures (STALDAC 2010) – an international conference organised by the editors of this collection at Newnham College, Cambridge, April 8–10, 2010. The conference gathered participants from various continents, presenting and discussing work on how humans represent space and time in various languages – including exotic and endangered – as well as how space and time are researched in linguistics, psychology, anthropology, philosophy, and various areas of cognitive science. The very intricate nature of the relationship between space and time is confirmed by the diversity of the areas of research that are represented by the contributions to the two volumes. This multifaceted approach to spatial and temporal constructs in human language, cognition, and culture enables us to shed new light on the interaction between potentially universal and language-specific/culturespecific features that shape the way people interact with each other and with their environment. Language as a uniquely human phenomenon provided a unifying platform for the discussions in the present volumes. The principal aim we have with this collection of contributions is to show that an all-encompassing understanding of space and time in language is not achievable in isolation, within a single discipline, but can be attained only through the study of linguistic habits, social contexts, scientific knowledge, and philosophical interpretation. The chapters in the collection follow several leading themes. The first volume, HCP 36, focuses on language diversity and presents research on, among other things, how location in space and time is conveyed in various languages; space and time in language acquisition; and speaking about motion, with its universal and language-specific aspects (see the Introduction to HCP 36). The second volume, HCP 37, devoted to language, culture, and cognition, focuses on the central topic of the representation of events; cross-cultural differences in representing time and space; and various aspects of the conceptualisation of space and time (see the Introduction to HCP 37). For the reader’s convenience, the tables of contents of both HCP 36 and 37 are listed in each volume.

xii Space and Time in Languages and Cultures: Language, culture, and cognition

Looked at more summatively, in juxtaposing the conceptual domains of spatial and temporal thought, the present two-volume collection contributes to various interrelated domains of research and types of research methods. Thinking and speaking about space and time frequently requires mobilising both linguistic and extralinguistic means of expression and hence these two domains are particularly conducive to fulfilling a role as the testing ground for theories of interaction, and therefore division of labour, between lexicon, grammar, and pragmatics. Several contributions address this question of the lexicon/grammar/pragmatics tradeoffs, for example in the domain of spatial deixis, time/tense mismatches, aspect, and language acquisition. Chapters in this category also contribute valuable data and theorising to the debate on linguistic relativity vs universalism. The topic of event construction can be safely regarded as pervading all sections in both volumes. Event type, its internal structure, boundaries, or the language-dependence of the construal are taken up in most contributions. Space is frequently addressed through cross-linguistic or cognitive analyses of motion events. Similarly, temporality, both external (tense) and internal (aspect) to the event, yields easily to contrastive, developmental, and psychological analyses. As far as methods are concerned, the theme of spatial and temporal reference is particularly conducive to experimental and other empirical testing: databased studies prevail in the collection. Formal semantic, philosophical, and theoretical contrastive linguistic approaches are also represented. They contribute to the discussion of event structure, tense, and aspect, among other things. The first volume collects many pertinent examples of contrastive linguistic research, both synchronic and diachronic, and both experimental and non-experimental. The second volume exemplifies interdisciplinary research methods, crossing the boundaries both within and between linguistics, anthropology, psychology, and philosophy. There are, of course, many aspects of time and space research that have not been covered in this collection, for example in the areas of neuroscience, formal syntax, metaphysics of time, or tense logic, to name a few. The field is indeed vast. The present collection, albeit cross-disciplinary, has language and cognition as its uniting theme but even in this domain, broad in itself, it contributes merely a selection of ideas that are currently in the focus of attention. We hope it will galvanise the emergence of new research questions, ideas, and solutions. We would like to acknowledge our gratitude to our colleagues who assisted us in the preparation of this collection. First, we would like to thank Malcolm Todd for his careful and thoughtful copy-editing. Next, our thanks go to Jos Tellings for editorial assistance in the early stages of the project and to our Cambridge STALDAC team for their help in organising the event from which the papers stem: George Walkden (Conference Secretary), Alistair Appleton, Jesper Carlson, Chris

Foreword: Space and time in languages, cultures, and cognition xiii

Cummins, Chi-Hé Elder, Minyao Huang, Eleni Kapogianni, Jane McDonnell, and Federico Pianzola; as well as to the sponsors: John Benjamins, Cambridge University Press, Oxford University Press, Cambridge ESOL, and the Department of Linguistics, University of Cambridge. Kasia Jaszczolt would like to thank Newnham College and the School of Arts and Humanities at the University of Cambridge for research grants that partially supported the preparation of the volumes. Last but not least, our thanks go to the series editors of Human Cognitive Processing, Linda Thornburg and Klaus-Uwe Panther, and to two anonymous referees for their invaluable comments and suggestions during the preparation of the final draft of the collection.  

Kasia M. Jaszczolt Luna Filipović


Cambridge, UK February 2012


Linguistic, cultural, and cognitive approaches to space and time Luna Filipović and Kasia M. Jaszczolt University of East Anglia / University of Cambridge

1. Preliminary remarks The interaction among language, culture, and cognition is vibrant, dynamic, and multidirectional. Language impacts culture in numerous ways, for example by being the medium for preservation of culture. It also impacts cognition by providing the most efficient system of categorisation, aid to memory, or spatial orientation. Similarly, culture impacts cognition through entrenching culture-specific preferences for the understanding of the environment. For instance, both spatial and temporal configurations may bear culture-specific traits in addition to cognitively universal ones, which may restrict which linguistic and cultural parameters become more, or less, important under different circumstances and conditions. It is this multifaceted nature of the relationships that are forged in and via language, culture, and cognition that inspired the equally multi-layered but highly interconnected collection of chapters in the current volume. Space and time are central conceptual domains in human experience. They constitute a large proportion of topics in daily conversations and they are omnipresent in human reasoning and planning even when language is not explicitly used. They provide an insight into spatial and temporal understanding of culturally and linguistically different communities. This second volume brings together a variety of approaches to interactions between language and culture, conceptualisation, development, memory, and human interaction with the environment. The authors approach the topic of spatial and/or temporal categories from a variety of interdisciplinary angles in an attempt to show both unity and diversity in thinking and talking about these domains within and across languages and cultures. Crucially, it is not just a matter of the same spatial or temporal features being expressed differently in different languages, but it may be the case that because

Luna Filipović and Kasia M. Jaszczolt

they are expressed in different ways, speakers may differ with respect to how they think about them. Linguistic diversity does not necessarily preclude cognitive diversity although it does not completely exclude the latter as a possibility. This possibility can be seen in the revival of neo-Whorfian strands in linguistics and psychology, which report evidence that language-specific effects on cognition (e.g. categorisation of objects or events and memory) are more than just a speculative possibility. With regard to relativist claims, we can notice a cline in the strength of linguistic relativity effects (cf. Malt et al. 2003; Levinson 2003a; Slobin 2000, 2003, 2006). There are good reasons for combining research on space with research on time in one state-of-the-art collection. The main question here is whether the conceptualisations of space and time are interrelated. There is some evidence that speakers are unable to ignore spatial concepts while representing time (e.g. Casasanto and Boroditsky 2008; Vallesi et al. 2008), which strengthens the currently widely researched hypothesis that the human concept of time is (asymmetrically) dependent on the concept of space. By the same token, the experience of duration is claimed to be asymmetrically dependent on the experience of spatial length. On the other hand, when it comes to left-right spatial orientation, relativity can be observed: some cultures conceptualise time vertically, some horizontally, and if horizontally, there are some interesting differences in the use made of directionality. Cyclic time, such as recurring seasons, can also dominate in some languages and in some contexts (see Jaszczolt 2009). The arrow of time often maps onto the spatial left-to-right dimension but this mapping is not universal. Next, even when cultures share the left-to-right conceptualisation, there may be differences that lie elsewhere. While in Mayan codices time was represented as running from left to right, it moved, so to speak, “in leaps and bounds”: it was not measured by units of equal duration. In recent research (see Aveni 2010), the method of the Maya is called ‘intervallic time reckoning’: an important event is followed by another memorable event but they are separated by an interval of a length particular to these adjacent events. And it is to this variable duration of a unit that the Maya were turning their attention. In the Christian calendar, we have a mixed system: sometimes it is the interval that takes precedence in conceptualisation, for example Lent is always of the same duration of forty days, while for some other parts of the calendar events determine the duration, as in the case of the interval between Christmas Day and Easter Sunday. The aspects and levels of possible diversity are therefore themselves numerous. Moreover, experiments strongly suggest that in addition to not being universal, the time-space mapping is not automatic either (Ulrich and Maienborn 2010) but rather proceeds as a simple metaphorical mapping. Next, there is a question

Introduction: Linguistic, cultural, and cognitive approaches to space and time

as to how deeply ingrained this relativity is; in other words, whether linguistic relativity precludes the existence of a common level of concepts (Levinson 2003a; Boroditsky 2001) and whether the effect of language on concepts can be overcome by a conceptualisation specific for L2 (January and Kako 2007). There are important metaphysical questions inherent here, such as whether time and space are metaphysically analogous; for example, whether we can have empty time just as we have empty space (Rundle 2009). The unidirectionality of metaphorical mapping seems to provide one plausible answer – or, more generally, linguistic semantic investigations help provide answers for big metaphysical questions. This is where the importance of this interdisciplinary collection lies – in the emphasis on language and language diversity on the one hand, and on the wider cognitive perspective on the other, both applied to the space-time relation. Both time and space conceptualisation rely on a well-delineated theoretical construct that pertains to some abstraction over situations. This role is normally fulfilled by events – or eventualities, where the latter subsume static and dynamic events, that is, events, states, and processes. Events are abstract entities; they can be construed “thinly”, with fine-grained distinctions, whereby Tom’s walking to the newsagent is considered a different event from Tom’s walking fast to the nearby newsagent, or “thickly”, with coarse-grained distinctions, whereby these two count as descriptions for the same event. But they have to conform to one necessary characteristic: they have to obey temporal continuity and manifestation of the same object. For example, Tom’s playing football and Tom’s playing hockey cannot be construed as one event, and neither can Tom’s putting on his coat and Jim’s putting on his coat, although syntactically they can easily be combined into simple sentences, respectively, as in (1) and (2). (1) On Saturday Tom played football and hockey. (2) Tom and Jim were putting on their coats.

. But it has to be remembered that when abstract entities were first introduced into semantics by Frank Ramsey in the 1920s they were contrasted with abstract facts: the fact that Brutus stabbed Caesar is abstract, while the event of Brutus’s stabbing Caesar is real and has its spatio-temporal location. Nevertheless, the ensuing discussion on the individuation of events and criteria for ‘eventhood’ threw up different principles of classification. See e.g. Davidson (1967), Chisholm (1970), Kim (1973) for these early discussions, and Montague (1960) on events as properties of time. The abstract nature of events is well elaborated in Asher (2000), who points out that facts have to be true and are eternal (e.g. the event of raining in Cambridge on 8 November 2011), while events, such as raining in Cambridge, acquire temporal specification externally. For further discussion see Jaszczolt (2009), Section 3.2.

Luna Filipović and Kasia M. Jaszczolt

The literature on this topic is ample (see e.g. Higginbotham et al. 2000). In the current collection events are in the focus of attention for several reasons. Firstly, and most importantly, they are units in which one discusses the passage of time and motion in space. Next, they are units of our memories, experiences, and anticipations. Thirdly, by looking at how events are expressed in languages we have access to perception, conceptualisation, including mechanisms for lexicalisation and grammaticalisation of concepts, and, generally, human cognition. Next, discourse about events provides material for addressing semantic questions such as ambiguity, vagueness, or reference. All of these topics are addressed in Part II of this volume. Crucially, the domains of space and time are instrumental in our search for the understanding of language-specific and universal aspects of human language and cognition. In the spatial domain we notice that while universal categories can be posited (e.g. à la Jackendoff 1995, 2003), an impressive variation is also evident (e.g. Levinson and Wilkins 2006) and at times this variation appears to undermine claims that all languages are the same underneath. What is “underneath” is still highly contested in the literature. Strong universalist claims are not based on empirical knowledge nor can they be since universalists in the Chomskyan vein for example are not interested in processing data, which are accessible for examination, but rather believe that universality lies in internalised computations, sadly not accessible for any kind of examination (see Hawkins, forthcoming, for a definitive overview and insightful discussion). The domain of space is widely researched for the purpose of finding universals in language and thought. Conflicting evidence is offered when it comes to how humans perceive space, interact with it, and talk about it. For instance, Jackendoff (1995, 1997, 2003) proposes a set of universal categories that feature in the human conceptualisation of space, such as thing, place, direction, event, manner, path, amount. For Jackendoff, cross-linguistic variation is merely a matter of complex packaging at a different level (see Filipović 2012 for discussion). However, this “packaging” of information at the language-specific level is not to be dismissed as less relevant or less consequential for the understanding of how spatial language and spatial cognition relate to each other. Levinson (2003a) and Levinson and Wilkins (2006) argue that universals in spatial language and cognition should be sought at the atomic levels of conceptualisation, where notions such as contact, vertical relation, adhesion, containment, and perhaps others reside. Complex concepts, termed “molar” by Levinson (2003a) like on or in are not good candidates for universals as previously thought (cf. Landau and Jackendoff 1993). Furthermore, it is at the molar level, the level of languagespecific “packaging” of information, that we operate in our everyday lives when communicating our thoughts and experiences. It is also at this level that we can

Introduction: Linguistic, cultural, and cognitive approaches to space and time

detect language effects on the processing of spatial information. For example, it is known that frames of reference vary across languages and cultures and those are not translatable across different modalities, i.e. the preferred frame of reference in a language pervades other cognitive activities such as orientation in space (see Levinson 2003b). Language-specific lexicalisation patterns in the expression of motion events are also shown to have an effect on how speakers of different languages remember what they see (see Filipović 2011; Filipović and Geva, this volume) or what kind of detail they provide in their accounts of events (Filipović 2007a, 2007b; Slobin 2006). Importantly, the search for language-specific features informs in turn the search for language universals and provides insights into how to tease these two apart (Filipović 2010). Similarly, when it comes to the expression and conceptualisation of time, we can also see the dynamism between universal and language- or culture-specific features. The psychological arrow of time, as it is popularly called by Stephen Hawking – that is, the direction in which humans experience the passing of time – is at the same time constrained by universal and culture-specific tendencies. Concerning the first, the psychological arrow has to be compatible with the direction in which the universe is expanding and with the direction of the increasing disorder (entropy) in the world. Concerning the second, various cultures adopt various conceptualisations of the direction, the unit of time, and the past, present, and future. But even in itself that domain of psychological time has a solid universal core: we think about time as we do because we are aware of the finiteness of life, of the fact that we were born and will die (see Heidegger 1953). In some cultures this awareness of one’s beginning and end is translated into calendrical systems. For example, in Shang China (1600–1044 BC), the days of sacrifices were named by the names of ancestors who, although dead, provided the continuity seen as essential for the survival of the human race. As a result, the sequence of such named days also shaped the present and the future, endowing them with analogous units (see Lewis 2010).

2. The contributions to the volume 2.1

Linguistic and conceptual representation of events

Part I of this volume, Linguistic and conceptual representation of events, is opened by “Event-based time intervals in an Amazonian culture”, in which Vera da Silva Sinha, Chris Sinha, Wany Sampaio, and Jörg Zinken discuss the difference between human cross-cultural universals such as experience of time on the one hand and its linguistic encoding and cultural variability in conceptualisation on

Luna Filipović and Kasia M. Jaszczolt

the other. Their ethnographic and field-experiment-based study of time intervals in a Tupi language and culture of Amazonia, Amondawa, shows that Amondawas have no time-based time interval systems, no lexicalised concept of “time as such”, and no practices of “time reckoning” as conventionally understood in the anthropological literature. Next, Minyao Huang in “Vagueness in event times: An epistemic solution” investigates vagueness in the temporal boundaries of events (onsets and endpoints) and argues that such vagueness is epistemic in nature, whereby the formation of knowledge lags behind physical occurrence of events. She concludes that the construal of temporal vagueness should be based on gradation of epistemic commitment to discrete physical changes in the real world. In “Aspectual coercions in content composition”, Nicholas Asher and Julie Hunter sketch a formal theory of lexical meaning, using type composition logic (TCL), the system developed in Asher (2011), with a particular focus on aspectual coercion. They put it in the context of the latest developments in dynamic semantics and computation modelling, addressing the issue of the interaction between lexical semantics and discourse structure, the role of presuppositions, and interaction of aspect with modality. In “Back to the future: Just where are forthcoming events located?”, Alan Wallington discusses the role and functioning of location metaphors in futuretime reference. He analyses temporal locations as indicating the degree of epistemic detachment which in turn is understood as the conceptual foundation for temporality. 2.2

Cultural perspectives on space and time

Part II, Cultural perspectives on space and time, opens with Valentina Apresjan’s contribution “The ‘Russian’ attitude to time”, in which she argues that a pragmatic approach to a temporal construction in Russian is necessary in order to understand Russian cultural perceptions of how events occur. Two specific constructs are evoked, namely that (1) events do not usually happen when they had been planned and that (2) if two unrelated events happen at the same time, this triggers the understanding that there is some fateful connection between such events. In particular, she looks at the ‘X to X’ construction (e.g. minuta v minutu ‘minute to minute’ meaning ‘to the minute’), using corpus data, and points out that the construction obtains culture-specific interpretations in English and in Russian, triggered by cross-cultural differences in attitudes to events. Culture-specificity in understanding temporality is also evident in the chapter by Bernard Charlier. In “Two temporalities of the Mongolian wolf hunter”,

Introduction: Linguistic, cultural, and cognitive approaches to space and time

Charlier describes what he calls “two temporalities” of Mongolian hunters, cyclical and single-event-based, and discusses the many different concepts Mongols have for the English word ‘fortune’ in the context of wolf-hunting as a culturally significant marker of temporal reference. Next, in “Koromu temporal expressions: Semantic and cultural perspectives”, Carol Priestley tackles time in Koromu, a Papuan language. The author argues that we can trace some formative semantic primes of meaning that are also evident in other languages but that the Koromu cultural context provides the backdrop for the understanding of semantic molecules, i.e. non-primitive meanings that occur within the meaning of other concepts. Finally, in “Universals and specifics of ‘time’ in Russian”, Anna Gladkova employs Natural Semantic Metalanguage (NSM) in order to demonstrate that Russian has exponents of the universal human concepts time and now, namely vremja and sejčas respectively, while the terms pora, teper’, and nynče are semantically complex. The NSM idea of universal semantic primes is supported, while also allowing for some cultural specificity in that lexical alternatives have been identified. 2.3

Conceptualizing spatio-temporal relations

Part III follows up with more cognitively-oriented work on Conceptualising spatiotemporal relations. Ronald Langacker, in the chapter “Linguistic manifestations of the space-time (dis)analogy”, shows that while space and time are comparable and can be plausibly viewed as being subject to unified treatment, time, if not fundamentally different in nature, has special status. He illustrates how the conception of time leads to the dynamic conception of space, which in turn enables the metaphorical conception of time itself. Next, in “Vectors and frames of reference: Evidence from Seri and Yucatec”, Jürgen Bohnemeyer and Carolyn O’Meara show that frames of reference play an equally important role in representations of the orientation of entities as they do in representations of their location and direction of motion. They propose that orientation is conceptually encoded in terms of vectors, which are a separate type of primitive conceptual function. The linguistic evidence in support of their claims comes from English, Yucatec Mayan, and Seri – the latter two indigenous languages spoken in Mexico. “Verbal and gestural expression of motion in French and Czech” by Kateřina Fibigerová, Michèle Guidetti, and Lenka Šulová is a cross-disciplinary study in the field of semantic typology and acquisition. The authors looked for languagespecific effects in the way French and Czech speakers express motion events in

Luna Filipović and Kasia M. Jaszczolt

language and gesture. They found considerable differences in verbalisation but the gestures were far more similar than expected. They further discussed the role of gesture in the expression of speakers’ intentions. Next, “Language-specific effects on lexicalisation and memory of motion events” by Luna Filipović and Sharon Geva presents the results of an experimental psycholinguistic study in which the memory of motion events in monolingual speakers of English and Spanish was tested. The authors found evidence for thinking-for-speaking, whereby speakers of English and Spanish performed differently in the recognition memory task. They discuss the results in the context of linguistic relativity. Continuing on the topic of the memory of events, the chapter “Space and time in episodic memory: Philosophical and developmental perspectives” by James Russell and Jonathan Davies is an experimental developmental study of spatiotemporal episodic memory. It offers insights into nonconceptual abilities that are nonetheless temporal in the sense of tracking order. Namely, children begin to be able to bind the spatiotemporal elements of a recollected event to the semantic elements at about the age of two-and-a-half, with this ability becoming well-established at the age of 3. Following Kant in the view that spatiotemporal content is constitutive of something being an experience, they suggest that this content is present in children who are too young to possess the concept of time. Next, Grzegorz Drożdż, in the chapter “Conceptualising the present through construal aspects: The case of the English temporal constructions” uses the framework and tools of Langacker’s Cognitive Grammar and illustrates the interactions between structured temporal concepts and temporal constructions in English. He emphasises the important role language has to play in conceptualising temporal distinctions relevant for the understanding of temporal boundaries both in language and in the conceptualisation of time. The section and this volume conclude with a chapter by Marlene Johansson Falck, “From perception of spatial artefacts to metaphorical meaning”, in which she combines a psycholinguistic investigation of mental imagery for paths and roads with a corpus linguistic analysis of the terms ‘path’ and ‘road’. She finds functional restrictions relevant for metaphorical contexts that are not operational in non-metaphorical ones.

3. Perspectives for future research This volume provides an interdisciplinary platform for further investigation into the relationship between language and thought. First of all, cutting edge research in various disciplines in cognitive science makes use of the concept of event, for

Introduction: Linguistic, cultural, and cognitive approaches to space and time

instance experimental psychology, formal semantics, cognitive linguistics or philosophy of language. By using event as a construct for discussing time and space a number of contributions to this volume offer new insights into, as well as pose new questions about, both universal and language-specific aspects of the construct itself. Another very important area of research is that of cultural diversity. By bringing together linguistic, psychological, anthropological and philosophical aspects of this diversity the chapters offer new perspectives on culture, approaching it through the study of social interaction, language and cognition. While in Volume 1 the question of diversity and universality appeared on the level of language systems and language use, here the same question is approached on the level of human behaviour. At the same time the evidence and arguments presented in what follows clearly indicate that further research effort is needed in order to answer fundamental questions such as whether there is a culture-specific, or language-specific, conceptualisation of space and time, and what the extent of this variability may be in cognitive processing, memory, social interactions, perception, and the use of language.

References Asher, N. 2000. Events, facts, propositions, and evolutive anaphora. In Speaking of Events, eds., J. Higginbotham, F. Pianesi, & A. C. Varzi, 123–150. Oxford: Oxford University Press. Asher, N. 2011. Lexical Meaning in Context: A Web of Words. Cambridge: Cambridge University Press. Aveni, A. F. 2010. The measure of time in Mesoamerica: From Teotihuacan to the Maya. In The Archaeology of Measurement: Comprehending Heaven, Earth and Time in Ancient Societies, eds., I. Morley & C. Renfrew, 203–215. Cambridge: Cambridge University Press. Boroditsky, L. 2001. Does language shape thought? Mandarin and English speakers’ conceptions of time. Cognitive Psychology 43: 1–22. Casasanto, D., and L. Boroditsky. 2008. Time in the mind: Using space to think about time. Cognition 106: 579–593. Chisholm, R. M. 1970. Events and propositions. Nous 4: 15–24. Davidson, D. 1967. The logical form of action sentences. In The Logic of Decision and Action. ed., N. Rescher. Pittsburgh: University of Pittsburgh Press. Reprinted in D. Davidson, 1980, Essays on Actions and Events, 105–122. Oxford: Clarendon Press. Filipović, L. 2007a. Language as a witness: Insights from cognitive linguistics. Speech, Language and the Law 14 (2): 245–267. Filipović, L. 2007b. Talking about Motion: A Crosslinguistic Investigation of Lexicalization Patterns. Amsterdam & Philadelphia: Benjamins. Filipović, L. 2010. Typology meets witness narratives and memory: Theory and practice entwined in cognitive linguistics. In Cognitive Linguistics in Action: Theory to Application and Back, eds., E. Tabakowska, M. Choinski, & L. Wiraszka, 235–248. Berlin: Mouton de Gruyter.


Luna Filipović and Kasia M. Jaszczolt

Filipović, L. 2011. Speaking and remembering in one or two languages: Bilingual vs. monolingual lexicalization and memory for motion events. International Journal of Bilingualism 15 (4): 466–485. Filipović, L. 2012. Spatial reference in discourse. In The Cambridge Handbook of Pragmatics, eds., K. Allan & K. Jaszczolt, 403–421. Cambridge: Cambridge University Press. Hawkins, J. A. Forthcoming. Cross-linguistic Variation and Efficiency. Oxford: Oxford University Press. Heidegger, M. 1953. Sein und Zeit. Tubingen: Max Niemeyer. Transl. by J. Stambaugh as Being and Time. 1996. Albany: State University of New York Press. Higginbotham, J., F. Pianesi, and A. C. Varzi, eds. 2000. Speaking of Events. Oxford: Oxford University Press. Jackendoff, R. 1995. Semantics and Cognition. Cambridge, MA: MIT Press. Jackendoff, R. 1997. The Architecture of the Language Faculty. Cambridge, MA: MIT Press. Jackendoff, R. 2003. Foundations of Language. Oxford: Oxford University Press. January, D., and E. Kako. 2007. Re-evaluating evidence for linguistic relativity: Reply to Boroditsky (2001). Cognition 104: 417–426. Jaszczolt, K. M. 2009. Representing Time: An Essay on Temporality as Modality. Oxford: Oxford University Press. Kim, J. 1973. Causation, nomic subsumption and the concept of event. Journal of Philosophy 70: 217–236. Reprinted in J. Kim, 1993, Supervenience and Mind: Selected Philosophical Essays, 3–21. Cambridge: Cambridge University Press. Landau, B., and R. Jackendoff. 1993. “What” and “where” in spatial language and spatial cognition. Behavioral and Brain Sciences 16: 217–238. Levinson, S. C. 2003a. Space in Language and Cognition: Explorations in Cognitive Diversity. Cambridge: Cambridge University Press. Levinson, S. C. 2003b. Language and Mind: Let’s get the issues straight! In Language in Mind, eds., D. Gentner & S. Goldin-Meadow, 25–45. Cambridge, MA: MIT Press. Levinson, S. C., and D. P.  Wilkins. 2006. Towards a semantic typology of spatial description. In Grammars of Space: Explorations in Cognitive Diversity, eds., S. C. Levinson & D. Wilkins, 512–552. Cambridge: Cambridge University Press. Lewis, M. E. 2010. Evolution of the calendar in Shang China. In The Archaeology of Measurement: Comprehending Heaven, Earth and Time in Ancient Societies, eds., I. Morley & C. Renfrew, 195–202. Cambridge: Cambridge University Press. Malt, B. C., S. A. Sloman, and S. P.  Gennari. 2003. Speaking versus thinking about objects and actions. In Language in Mind, eds., D. Gentner & S. Goldin-Meadow, 81–111. Cambridge, MA: MIT Press. Montague, R. 1960. On the nature of certain philosophical entities. The Monist 53: 159–194. Reprinted in R. Thomason, ed., 1974, Formal Philosophy: Selected Papers of Richard Montague, 148–187. New Haven, CT: Yale University Press. Rundle, B. 2009. Time, Space, and Metaphysics. Oxford: Oxford University Press. Slobin, D. I. 2000. Verbalised events: a dynamic approach to linguistic relativity and determinism. In Evidence for Linguistic Relativity, eds., S. Niemeier & R. Dirven, 107–138. Amsterdam & Philadelphia: Benjamins. Slobin, D. I. 2003. Language and thought online: Cognitive consequences of linguistic relativity. In Language in Mind: Advances in the Investigation of Language and Thought, eds., D. Gentner & S. Goldin-Meadow, 157–191. Cambridge, MA: MIT Press.

Introduction: Linguistic, cultural, and cognitive approaches to space and time

Slobin, D. I. 2006. What makes manner of motion salient? Explorations in linguistic typology, discourse and cognition. In Space in Languages: Linguistic Systems and Cognitive Categories, eds., M. Hickmann & S. Robert, 59–81. Amsterdam & Philadelphia: Benjamins. Ulrich, R., and C. Maienborn. 2010. Left-right coding of past and future in language: The mental timeline during sentence processing. Cognition 117: 126–138. Vallesi, A., M. A. Binns, and T. Shallice. 2008. An effect of spatial-temporal association of response codes: Understanding the cognitive representations of time. Cognition 107: 501–527.


part i

Linguistic and conceptual representation of events

chapter 1

Event-based time intervals in an Amazonian culture* Vera da Silva Sinha,# Chris Sinha,† Wany Sampaio# and Jörg Zinken‡ #Federal

University of Rondônia / †Lund University / ‡University of Portsmouth

We report an ethnographic and field-experiment-based study of time intervals in Amondawa, a Tupi language and culture of Amazonia. We analyse two Amondawa time interval systems based on natural environmental events (seasons and days), as well as the Amondawa system for categorising lifespan time (“age”). Amondawa time intervals are exclusively event-based, as opposed to time-based (i.e. they are based on event-duration, rather than measured abstract time units). Amondawa has no lexicalised abstract concept of time and no practices of time reckoning, as conventionally understood in the anthropological literature. Our findings indicate that not only are time interval systems and categories linguistically and culturally specific, but that they do not depend upon a universal “concept of time”. We conclude that the abstract conceptual domain of time is not a human cognitive universal, but a cultural historical construction, semiotically mediated by symbolic and cultural-cognitive artefacts for time reckoning. Keywords: Amazonia, artefacts, semiotic mediation, onomastics, time reckoning

* Our most important thanks go to the Amondawa community, who have shared their language with us. We wish especially to thank Chief Tari Amondawa and Arikan Amondawa, who is the indigenous teacher in the village school. Support for this study was provided by the European Union, as part of the collaborative project SEDSU, ‘Stages in the Evolution and Development of Sign Use’ under the 6th Framework NEST/Pathfinder programme ‘What it Means to be Human’; and by the Federal University of Rondônia and the University of Portsmouth.


Vera da Silva Sinha, Chris Sinha, Wany Sampaio and Jörg Zinken

1. Introduction Although both the phenomenological experience of time (Bergson 1910), and the linguistic encoding of temporal inter-event relationships in grammar and the lexicon, may be considered to be human transcultural universals, the conceptualisation and linguistic expression of time intervals (that is, lexicalised concepts of intervals of temporal duration) is widely culturally variable. Much anthropological linguistic research has addressed variability in calendric systems, and in the social practices of “time reckoning” (Evans-Pritchard 1939, 1940) that are dependent on, and realised through, such calendric systems. Numerically based calendric systems can be regarded as organising time-based time intervals. Timebased time intervals (such as “clock time” and “calendar time”: Levine 1997; Postill 2002) are those whose boundaries are constituted by the segmentation of a conceptual domain of “time” as an abstract and measurable entity: what we may call Time as Such. Examples of time-based time intervals are hours and weeks. Although time-based time intervals are based upon natural (astronomical) cycles of events, they are conventional and their duration is derived from counting in a number system. Time-based time intervals can be distinguished from eventbased time intervals. Event-based time intervals are those whose boundaries are constituted by the event itself. In this sense, there is no cognitive differentiation between the time interval and the duration of the event or activity that defines it, and from which in general the lexicalisation of the time interval derives. The reference event is often natural (such as ‘spring’, e.g. ‘Let’s take a holiday in the spring’), but sometimes conventional (such as ‘coffee break’, e.g. ‘Let’s discuss this during coffee break’). We report here an ethnographic and field-experiment-based study of time intervals in a Tupi language and culture of Amazonia, Amondawa. In the following section, we review anthropological research on time intervals. In Section 3, we provide an overview of the Amondawa culture and society. In Section 4, we report our analysis of two event-based Amondawa time interval systems (seasons and days), and in Section 5 we describe the Amondawa system for categorising lifespan . The event-based time interval may be characterised as a change of state (e.g. ‘sunrise’), as a stative event attribute (e.g. Amondawa ara, ‘daylight’), or as an activity. The lexicalisation may be metonymic or “pars pro toto”, as in Amondawa pojiwete, ‘when we start work, morning’ (Whitrow 1988: 15). . The fieldwork on which this paper is based was carried out by the first and third authors; the fourth author had primary responsibility for the fieldwork manual; the first, second, and fourth authors had primary responsibility for the data analysis; the second and first authors have primary responsibility for this text.

1.  Event-based time intervals in an Amazonian culture

time (“age”).We show that Amondawa has no time-based time interval systems, no lexicalised concept of Time as Such and no practices of “time reckoning” as conventionally understood in the anthropological literature (Evans-Pritchard 1939). In Section 6 we discuss the implications of our findings for proposals and assumptions that time is a fundamental cognitive domain or semantic primitive.

2. Calendars and time reckoning: Anthropological perspectives There is a considerable body of anthropological research dealing with culturally specific calendric systems. Calendric systems frequently possess a recursive structure such that different time intervals are embedded within each other, and/ or a structure of metrically overlapping intervals. These intervals are typically cyclical in nature, with both embedded and overlapping cycles. The most familiar to us is the now widely adopted lunar and solar (more strictly, monthly and annual) Gregorian calendar. A dramatic example of the complexity that such systems can attain is provided by the classical Mayan calendars. The Mayan civilisation used three different calendar systems. The so-called Long Count calendar organised the historical time of the classic period of Mayan in a fashion comparable to a car’s odometer, counting days in geared cycles of ascending size. The Long Count used the number 360 as an approximation of the year, multiplying the 20-day months by eighteen to arrive at a round-figure year of 360 days or one tun. Twenty tuns composed a katun, and twenty katuns formed one baktun. These time intervals (tun, katun, and baktun) could be used to specify any day in Maya history. The Long Count could also generate time references in an (in principle) infinite scale, a fact which both structured Mayan cosmology and was the main motivation and function for Mayan mathematical knowledge; this worked with place value and the number zero, both unknown to Mediterranean classical antiquity. The Haab was a civil calendar based on the tun, and five days were added at the end of the Haab year to approximately synchronise it with the solar year. The Tzolkin (counting days or Sacred Year) calendar was a ceremonial calendar, with 20 periods of 13 days, thus completing a ritual cycle every 260 days (Edmonson 1976; Wright 1991). Calendric systems are not purely quantitative systems of measurement and ordering. They are also expressive of cultural beliefs and values. The Western (Gregorian) calendric system, for example, conceptually superimposes on its . We restrict this discussion to time interval systems, rather than attempting to address the much wider topic of the anthropology of time in general. For reviews, see Gell (1992), Munn (1992).



Vera da Silva Sinha, Chris Sinha, Wany Sampaio and Jörg Zinken

cyclic structure a linear model of time as involving motion from an origin (the birth of Christ) to a notional endpoint (the End of Days). This dualistic cyclicallinear conceptualisation (with varying relations of dominance between cyclicity and linearity) is characteristic also of other calendric systems, such as the Mayan (described above), the Islamic, and the Vedic (Keyes 1975). Geertz (1973), in his classic paper “Person, time and conduct in Bali”, argues that temporality (and time interval measurement) in Balinese culture cannot be comprehended without recognising its contextual embedding within Balinese notions of personhood, social status, and social role. Personhood, social role, and time form a complex matrix in which, Geertz (as interpreted by Vickers 1990: 166) argues, “time in Bali is not linear, that is not quantitatively divided, but qualitative – organised in terms of degrees of malevolence and benevolence”. Calendric time is thus co-constituted with social norms of conduct and power (Bloch 1977). It is this interpretation that underlies Geertz’s hypothesis that Balinese time is “detemporalized”: the Balinese, claims Geertz (1973: 398), have “a classificatory, fulland-empty, ‘de-temporalized’ conception of time in contexts where the fact that natural conditions vary periodically has to be at least minimally acknowledged”. Gell (1992: 72) points out, however, that “the evidence for Balinese detemporalization is specifically connected with the permutational calendar … that it does not generate regular periodicities (such as solar years subdivide in lunar months, which subdivide into market weeks, etc.). Instead the permutational calendar specifies quantum units (days) in terms of combined product of independent five-, six- and seven-day cycles”. Alongside this Pawukon permutational calendar, which commutes a complex trinomial expression whose completion takes 210 days, the Balinese also employ a variant of the luni-solar Hindu (Vedic) calendar. Gell (1992: 73) summarises Geertz’s argument as being that “both Balinese calendars are non-metrical and ‘non-durational’, and thus correspond to the climaxless ‘steady state’ and non-progressive tenor of Balinese life”. Geertz’s analysis has been criticised on various grounds, ranging from its Durkheimian over-emphasis on ritualistic conduct (Bloch 1977) to its neglect of the significance in everyday time reckoning of the quantitative computations made possible by the Balinese calendar, and the degree of expertise displayed by Balinese in exploiting these possibilities. Without entering too deeply into this issue, we would make a very simple point: whatever cognitive and social significance we may wish to accord to cultural variations in calendric systems (see e.g. Charlier, this volume, on the use of the astrological calendar by Mongolian hunters; Keyes 1975, and Davis 1976 on the Northern Thai system), all such systems are quantificational, in the sense of being based upon a measurement system, and all can be considered as time-based, segmenting and measuring temporal duration in Time as Such. The speech practices of reckoning or telling time, with their

1.  Event-based time intervals in an Amazonian culture

etymological roots in Germanic words for counting (e.g. Dutch rekenen, ‘to count’), express and reproduce this quantificational view of time. Analogous arguments to those applying to calendric time can be made for clock time, that is the conceptualisation and measurement of time intervals in the diurnal cycle, although less attention has been paid to this in the anthropological and linguistic literature (see, however, Postill 2002). Not all societies employ either calendar or clock systems of the quantificational type. Evans-Pritchard (1939, 1940) described what he termed the Nuer “cattle clock” or “occupational time”. Time in Nuer society, he proposed, is based on environmental changes and associated social activities. The concept of time in Nuer society is thus a product of the interplay between “ecological time” and “social structure time”. In describing Nuer concepts of time we may distinguish between those that are mainly reflections of their relations to environment, which in a broad sense we may call ‘oecological time’ [sic], and those that are reflections of their relations to one another in the social structure, which we may describe as ‘structural time’ … time has therefore two movements, an oecological (or occupational) movement and a structural, or moral, movement. (Evans-Pritchard 1939: 189–190)

The Nuer ruon (year) divides time into two principal seasons, tot (rainy season) and mei (dry season). These two main seasons are supplemented by classifications based on activities. For example, Jiom (meaning ‘windy’) refers to the period when the cattle-camps are formed, and Rwil refers to the period of moving from camp to village, clearing cultivations and planting (op. cit. p. 196). Although there are names for (roughly) lunar months, Nuer society does not count or measure Time as Such; the language has no word either for the abstract notion of time, or for units of abstract time, and temporal reference points are provided by social activities. Nuer have no abstract numerical system of time-reckoning based on astronomical observations but only descriptive divisions of cycles of human activities (op. cit. p. 197) … since the months are anchored to oecological and social processes the calendar is a conceptual scheme which enables Nuer to view the year as an ordered succession of changes and to calculate to some extent the relation between one event and another in abstract numerical symbols (op. cit. p. 200).

Nuer months are not strictly lunar (though the Nuer know the lunar cycle), nor based upon any other fixed number of days. Rather, they are conventionally, if indeterminately, based on both lunar and ecological cycles, and the associated rhythm of social activities.


20 Vera da Silva Sinha, Chris Sinha, Wany Sampaio and Jörg Zinken

Nuer would soon be in difficulty over their lunar calendar if they consistently counted the succession of moons, but there are certain activities associated with each month, the association sometimes being indicated by the name of the month. The calendar is a relation between a cycle of activities and a conceptual cycle, and the two cannot fall apart, since the conceptual cycle is dependent upon the cycle of activities from which it derives its meaning and function.  (Evans-Pritchard 1940: 100)

In summary, time for the Nuer is a schematised relation between socially and environmentally defined events, and Nuer time reckoning is not a calculation of, or in, Time as Such, but a rough estimate, only infrequently numerically expressed, based on social-structural relationships and activities. The Nuer seem, according to Levine’s (1997) terminology, to be living in event time rather than clock time: activities are not fitted into a schedule governed by the clock or calendar, rather the temporal structure of life emerges from participation in daily activities. Nuer time is not the only system of time intervals reported in the anthropological literature that employs lunar months in a non-quantified system. The time interval system of the Ainu culture of Southern Sakhalin, which in other respects (economy, social structure, and cosmological time) is quite different from the Nuer system, includes lunar months that regulate ritual as well as trapping and fishing activity. However, “the Ainu are quite oblivious to names of the months as well as the number of months in the year” (Ohnuki-Tierney 1973: 289), and the Ainu, whose basic number system (non-derived numbers) extends to five, rarely or never reckon time intervals numerically, using the opposition between two or three and the derived number six to contrast short with long durations. While the Nuer event-based time interval system can be thought of as quasi-calendric, permitting rough time-reckoning practices, the unnamed Ainu lunar months do not participate in anything resembling a yearly calendar. Ohnuki-Tierney concludes that “the Ainu concept of time is basically qualitative; quantitative measurement of time is little developed. Therefore, no temporal divisions represent measurable units; they are distinguished from other units in the same time scale by the special meaning which the Ainu attach to them” (op. cit. p. 292). These descriptions of Nuer and Ainu event-based time interval systems serve as a useful starting point for our ethnographic and field-experimentally based description of time intervals in Amondawa.

1.  Event-based time intervals in an Amazonian culture

3. Amondawa culture and society: An overview The Amondawa are an indigenous group living in the Uru-eu-wau-wau reservation, in the State of Rondônia in Brazilian Greater Amazonia. Amondawa is classified as a Tupi Kawahib language belonging to the family Tupi-Guarani, closely related to the other Kawahib languages (Diahoi, Karipuna, Parintintin, Tenharim, Uru-eu-wau-wau) of Amazonian Brazil (Sampaio 1996, 1999; Sampaio and Silva 1998). The population at the time at which the fieldwork here reported was conducted consisted of about 115 people. Before official contact in 1986 by the government agency FUNAI, the Amondawa population was almost 160 people; after contact, this number went down by more than 50%, according to contemporary reports. In 1991, the Amondawa population was no more than 45 people, living in the area surrounding the Trincheira post, which is also the current habitation. The main cause for the precipitate decline of the population was contact-induced disease, such as tuberculosis, colds, measles, malarial fever, chicken pox, and other viruses (Silva 1997). At present, the population is skewed towards the younger generation, which makes up more than half of the population. The Amondawa kinship system, in common with other Tupi Kawahib groups, is organised in terms of exogamous moieties. Descent is patrilineal. The woman does not lose her paternally derived name when she marries, but her children will be the descendents of her husband and adopt names from his moiety (Menendez 1989: 110). The Amondawa moieties are designated by the bird names Mutum and Arara. The mutum is a black bird living almost all the time on the ground and the arara is a colourful macaw that lives in the highest trees. Descent is reflected in the system of personal proper names, because each moiety has an inventory of masculine and feminine names. Amondawa people change their names during their life course, and these names are indicative of the person’s “age”/social role, gender, and moiety. The change of names occurs at the birth of a new baby and/or when the individual assumes a new position, attribute, or role in social life. We describe this system and its significance for the Amondawa cultural conceptualisation of time below. Amondawa productive activity is based around cultivation. The men work in the field planting corns, beans, rice, potatoes, and manioc. Traditionally, cultivation has been for subsistence but is now also for the market. Manioc flour is the . Amondawa is not the original pre-contact self-designation of this community, but is now the community usage. . The original indigenous name is Kanideia, but the term arara has become common usage post-contact.



Vera da Silva Sinha, Chris Sinha, Wany Sampaio and Jörg Zinken

most important commodity yielding monetary income for the community. Each nuclear family has its own field. The families from the same moiety sometimes share work and profit. This means that in effect each moiety decides how much will be produced each season. There is no culture of accumulation or of keeping produce or seed for the next season; everything produced is consumed or sold and the money is used for buying manufactured products, such as soap, clothes, shoes, TVs. Hunting and fishing, traditionally significant activities, remain the other main sources of food. The traditional mode of Amondawa education is oral and informal, but since 1994 formal schooling has also been provided by the State. Today the majority of the Amondawa people are bilingual in Amondawa and Portuguese. Portuguese has high status because it is the main vehicle for communicating with others outside the village. Communication between community members is still in Amondawa, and Amondawa is the language of first acquisition. Schooling is bilingual, with a predominance of spoken and written Amondawa as medium of instruction. The teacher is a trained community member supported by the specialist from the State Department of Education. The curriculum emphasises Amondawa history and tradition and knowledge of the local environment.

4. Time intervals in Amondawa language and culture Amondawa does not employ cardinal chronologies such as ages of individuals, or ordinal chronologies such as yearly or monthly calendars, since the Amondawa number system has only four numeral terms, of which the equivalents of ‘three’ and ‘four’ are derived. The non-derived terms are pe’i ‘one’ and monkõi ‘two’. Monkõiape’i or ape’imonkõi are alternative lexicalisations of ‘three’; monkõiuturaipei and monkõimeme are alternative lexicalisations of ‘four’. An abstract term for time does not exist in Amondawa. The word kuara (‘sun’) is preferentially used to denote time intervals in general, since it is the movement of the sun that governs the passage of both the time of day and the seasons. Our ethnographic research has failed to identify any co-occurrence of numerals with any time interval designation. These features of the Amondawa language mean that time reckoning simply does not occur in Amondawa discourse. This does not, however, mean that the language lacks a lexicon of time intervals. The two time interval systems on which, together with the personal proper name system, we focus in this section are the seasonal and diurnal systems. As far as we know, these are the only such systems.


1.  Event-based time intervals in an Amazonian culture


A field manual was developed, which consisted of elicitation games and questionnaires (Zinken, Sampaio, Silva Sinha and Sinha, 2005). The manual was specifically constructed to identify temporal expressions and their ranges of use in Amondawa. Two of the tasks in the field manual addressed the lexicalisation of time interval terms: The calendar questionnaire and the calendar installation. These tasks are described below. 4.1.1 Calendar questionnaire The aim of the calendar questionnaire was to provide data on the inventory of calendar event-types that are lexicalised in Amondawa. The questionnaire contains a list of interval terms in Portuguese, relating to time intervals based on the moon (the ‘month’ and its subdivisions), and on the sun (the ‘day’ and its subdivisions). It also contains questions about sowing, harvesting, and festivals. Participants. Data were collected during five field trips between September 2005 and January 2006. The participants were six adult bilingual native Amondawa language consultants (four male and two female), all of whom were familiar with the researchers administering the instruments and experienced in the role of language consultant. Procedure. The researcher started by asking direct questions in Portuguese about Amondawa calendar units, names of festivals, parts of the day, and time adverbials as the central topic of the conversation. The researcher did not ask for literal translations, but asked more general questions about broadly equivalent terms in Amondawa and developed on this basis a conversation. It was emphasised to the participants that there were no right or wrong answers and that it was the Amondawa cultural knowledge that was the focus of investigation. The participants’ responses were video and audio recorded and post-transcribed. Results. There is no word meaning ‘time’ in Amondawa. There are in Amondawa no words for weeks, months, and years, and there are no names for time-referenced festivals. In fact, there are no such festivals in contemporary Amondawa culture, only marriage parties and traditional ceremonies that are not

. The standard version of the Field Manual (Zinken et al. 2005) is written in English but was translated by the field researchers into Portuguese.



Vera da Silva Sinha, Chris Sinha, Wany Sampaio and Jörg Zinken

Table 1.  Amondawa temporal reference terms Nominals with temporal meaning

English translation

Kuara Jahya Ipytuna Ko’ema Ko’emameme

Sun Moon Night, black Morning “Tomorrow”

Other (adverbial) time referencing expressions Koro, koroite Tiro Tirove Awo Ki…ko Poti … nehe Emo Ramo Ki … i’i

Today, now, right now (fut.) Today, now, right now (fut.) Today, in the immediate past (earlier today) Here, now Past Future Past Past Past

referenced to specific time intervals. There are names for seasons and parts of the seasons, for the day and night and parts of the day and night, and some temporal deictic and adverbial terms. Some of these are listed in Table 1, which is not exhaustive. The remaining tasks reported here focused on the elucidation of the time intervals and their systematic organisation in lexical and conceptual fields. For further analysis of temporal reference in Amondawa, see Sinha, Silva Sinha, Zinken, and Sampaio (2011). 4.1.2 Calendar installation: Seasons This elicitation game gave participants the opportunity to build a map of their model or schema of the ‘year’ (or other interval longer than a month) and its subintervals or constituents, by placing a series of paper plates, each representing a conventional time interval, on the ground. The participants were requested by the researcher to “make a map of the year using the objects”.

. We know little of the deep pre-contact history of Amazonian cultures, especially before the Spanish/Portuguese conquest. The only thing of which we can be certain is that it would be a grave mistake to view the existing (surviving) cultures of indigenous groups as being representative of some “unchanging” primordial state “without history” (Hornborg and Hill, in press; Wolf 1982).

1.  Event-based time intervals in an Amazonian culture

Plate 1.  One participant’s representation of the Amondawa year Procedure. Four participants (all men) were interviewed in Portuguese with simultaneous translation into Amondawa. Paper plates were given to the participant who was then asked to “make a map of time in Amondawa with them”, in which each plate should represent one interval of time in Amondawa culture. The example provided was that in Portuguese each plate would represent a month. The participants’ responses were video and audio recorded and post-transcribed. Plate 1 shows the results of playing the game with one participant, who has used the plates to construct a schematic representation of the succession of seasons in Amondawa. Results. In Amondawa, there is no word for ‘year’. Linguistically, time is divided not into years, but into two seasons: the dry season Kuaripe (‘in the sun’) and the rainy season Amana (‘rain’). The term Kuaripe, referring to the hot, dry season, derives from the noun Kuara (‘sun’), with the locative postposition pe, meaning ‘in’ or ‘at’. The rainy season is designated simply by the noun Amana, which means rain. The passage of the seasons is marked by changes in the weather, and consequent changes in the landscape, and also by the rhythm of agricultural activities. Each season is further subdivided into three intervals corresponding to the beginning, middle (or “high”), and end parts of the season. Table 2 lists the Amondawa bi-seasonal lexical system.


26 Vera da Silva Sinha, Chris Sinha, Wany Sampaio and Jörg Zinken

Table 2.  Amondawa seasonal time interval words AMONDAWA



Time of the sun (“SUMMER”)

O´an kuara

‘The sun is born’. The arrival of the sun (beginning of the time of the sun).

Itywyrahim kuara

‘Burning sun’. Very strong, hot sun, high summer.

Kuara Tuin Or Akyririn Amana

‘Small sun’. End of the time of the sun.


Rain / Time of the rain (“WINTER”)

Akyn Amana

‘Falling rain’. The arrival of the rain.

Akyrimba´U Amana Or Amana Ehãi

‘Heavy falling rain’. Time of the heavy rains.

Amana Tuin Or Akyririn Kuara

‘Small rain’. End of the rainy season.

‘Almost rain’. The time of falling rain is close.

‘Great rain’. Rain of long extent and duration.

‘Almost sun’. The time of the sun is close.

Figure 1 represents, approximately, the way the seasons were mapped by participants. It is based upon the constructions of all four participants, each of whom constructed a curvilinear representation that fitted into the available working space, more or less on the horizontal axis perpendicular to the direction in which the participant faced, in either a left-to-right or right-to-left order of placement. No participants attempted to create a circular, cyclic representation. It is unclear whether the curvilinear responses were a result of a compromise between an intended rectilinear configuration and the length of human reach, or signify that neither cyclicity nor rectilinearity are relevant to the Amondawa seasonal schema.

AMANA Akyn amana

Amana tuin

KUARIPE O’an kuara

Akyrimba’u ama Akyn amana Amana ehai

Figure 1.  The Amondawa season schema


Kuara tuin Akyn akyririn Itywyrahim kuara

1.  Event-based time intervals in an Amazonian culture

4.1.3 Calendar installation: Days This elicitation game gave participants the opportunity to build a map and/or installation of their model or schema of the diurnal cycle. The procedure was identical to that described above for the calendar installation. The day installation game was administered immediately after the calendar installation game. Results. The term for ‘day’ in Amondawa, Ara, refers only to the daylight hours and also has the meaning ‘sunlight’. There is no Amondawa term for the entire 24-hour diurnal cycle. Ara, ‘day’, contrasts with Iputunahim, ‘night’, which also means ‘intense black’. There is a major subdivision of Ara, ‘day’, into two parts, Ko´ema (morning), and Karoete (noon/afternoon). Thus, additionally to the binary day–night contrast, it is also possible to say that the 24-hour period is divided into three major parts, Ko´ema, Karoete, and Iputunahim. Both day and night are further subdivided into intervals that are conceptualised and named on the basis of the daily round of activities. Table 3 lists all time interval terms produced by the participants in the day installation game. Table 3.  Parts of the day in Amondawa Ara or ajia

Day (daylight)

Ko´Ema Pojiwete Kojawahim A´U Matera Ajia Katua Ajimbu´U Pyriete Kuara Ruwi Ajia katua

Morning ‘When we start work.’ Early morning. ‘When we feel hungry’ ‘When we eat’. Lunchtime. ‘Good morning time’. After lunch. ‘Heavy morning’. Late morning. ‘The sun is high’. High noon.

Karoete Pyryrym Kuara Momina Werin Kuara Momina Kuara

Noon; afternoon. ‘The sun is turning’. Early afternoon. ‘The sun is almost gone’. Late afternoon, dusk. ‘The sun is gone’. Early evening. Twilight.

Iputuna Opon Jahya Tiro Apehyiahim Apoji Katua Ypytunahim Pyriete Jahya Ruwi Jahya Pyryrym Ko´Ema Werin Opon Kuara Tiro

Night (black) ‘The moon leaps up now’. Moonrise. ‘No more work.intense’. Sleep time. ‘Good …. ’ ‘Intense darkness’. Middle of the night. ‘The moon is high in the sky’. ‘The moon is turning’. Dawn is coming. ‘Almost morning’. Dawn. ‘The sun jumps up now’. Sunrise.



Vera da Silva Sinha, Chris Sinha, Wany Sampaio and Jörg Zinken

The schematisation of the diurnal cycle does not seem to be cyclical or circular. In trying to explain this task, the researchers used a circular diagram resembling a clock, with light and dark areas. However, none of the participants produced a circular installation. Instead, they produced curvilinear representations similar to those produced in the calendar installation game.

5. Time and the human lifespan in Amondawa As we have noted, the age of an individual is not measured chronologically in Amondawa culture, which lacks a numerical system able to enumerate above four. Rather, individuals are categorised in terms of stages or periods of the lifespan based upon social status and role, and position in family birth order. As we have also noted, each Amondawa individual changes their name during the course of their life, and the rules governing these name changes form a strict onomastic system. The Amondawa onomastic system is based upon the cross-cutting category systems of life stage, gender, and moiety. It is obligatory for each individual to change his or her name when “moving” from one life stage to another, and each name is selected from a finite inventory of names, each of which has a semantic value indicating moiety, gender, and life stage. Thus, by knowing the name of an Amondawa person, one can infer these dimensions of their social status. The principal event that can cause a change of names is the birth of a new member of the family. The new baby will be given a “Newborn” name, and may even assume a name previously held by the youngest existing family member; who then takes a new name. Regardless of the name given to the newborn, all the existing children will acquire a new name. The other situation that can provoke the changing of names is a change in the role of the individual in the family or in the group. No individual can be a child forever, in other words no-one can have a child name beyond a certain life stage. They have to grow up and assume responsibilities in the family. For example, when an older son changes his name, the father will change his name too. An adult woman will change her name when she is married, and her previous name will go to the youngest sister (Peggion 2005: 132). The names do not appear to have spiritual significance, and in assuming a new name and new social identity, the individual does not become identified with the personality of previous living or dead bearers of the name. Table 4 gives examples of names in each Amondawa moiety with an indication of their status meanings, although it is important to note that this is only an approximation. Table 4 does not represent the entire name inventory.

1.  Event-based time intervals in an Amazonian culture

Table 4.  Amondawa names and life stages Arara (F)

Arara (M)

Mutun (F)

Mutun (M)

Life stage





Newborn to toddler





Child to pre-adolescent


Pure- Tebu



Adolescent (from puberty)





Young adult

Mande´I Adiju Umby

Purap Mboria Mboria

Mboropo Kunha´pó kunhaviju

Yvaka Moarimã Mboava







Table 5.  Generic nouns referring to categories of persons Kurumin


Kwambáea Kuñã

Man Woman

Amu Tiwi

Old man Old woman

The Amondawa language also has a number of generic nouns referring to categories of persons of a particular age (Table 5). Our own and others’ research (Sampaio 1996; Silva 2000; Peggion 2005) has not been able to identify any other age-based person categories such as “adolescent”. Although we are not fully certain of this, our research to date suggests that there is only one more general expression, namely etiawa’ea (‘old’, an adjective of quality or state applicable to any object), used for reference to life stage: (1) Aron jihe etiawa’ea waiting I old (Adj) ‘I am waiting for my old age’

In other cases, life stage is referred to by means of the relevant life stage category, e.g. (2) a-kuahaw-a-him jihe kurumin ga inguarai-awer-a 1s.imagine-ger-intens. I child he play-past-nom. ‘Imagining I played as a child’



Vera da Silva Sinha, Chris Sinha, Wany Sampaio and Jörg Zinken

In summary, the temporal intervals making up human life stages in the Amondawa culture and language are designated in the kinship-related onomastic conceptual system, and to a more limited extent in categories of person of a particular age. They are not related to any calendric or numeric system segmenting Time as Such, and they are not constituents of either exact or rough quantitative time reckoning.

6. Discussion We have found that: a. Amondawa time interval conceptualisation is not integrated or coordinated with the four-number Amondawa numeral system. This fact precludes numeric time reckoning as a cognitive and linguistic practice. b. The rhythms of the natural world dominate the seasonal and diurnal time interval systems. The prominence of the sun, in terms of the intensity of emitted heat and light in different seasons, and its position in the sky at different times of day, is reflected in language consultants’ choice of the lexeme kuara ‘sun/sunlight’ as the nearest Amondawa equivalent term for the Portuguese word tempo, ‘time’, for which no strict translation equivalent exists. c. Both the seasonal and the diurnal time interval systems involve division and subdivision. The superordinate level of the seasonal system is bi-partite (dry season–rainy season), while that of the diurnal system seems to have two alternative divisional structures, a primary bi-partite one (day–night) and a secondary tri-partite one (morning–afternoon–night). Beneath these superordinate divisions are lower level subdivisions. d. In both cases it is the subdivision level of organisation that is coordinated with the organisation of social and, in particular, labour activity, regulating planting and harvesting times and working times during the day. The seasonal and diurnal time interval systems can therefore properly be thought of as cognitive, cultural, and linguistic schemas, but they differ from more familiar calendric and clock schemas in that there is no evidence that they are conceptualised by speakers as being cyclical in structure. Cyclicity is schematically characterised in terms of a circular or orbital path of motion in which “moving time” returns recurrently to the positions that demarcate the time intervals. None of our language consultants either verbally described a temporal cycle or produced a physical schematic model (installation) that possessed a circular structure. Rather, the schematisation seems to be simply in terms of succession, which

1.  Event-based time intervals in an Amazonian culture

may be (as we have seen) spatially modelled as a line, though not necessarily a straight one. Amondawa seasonal and diurnal time intervals are best thought of as high-level event categories – “happenings”, as it were, in the natural and social world, with which other happenings may coincide, or to which other activities and events are indexed. The third time interval system that we have analysed above is the conceptual system of Amondawa life stages, as this is reflected in Amondawa onomastic practices and knowledge. Time intervals in this system are conceptually inseparable from the Amondawa kinship and descent system, and form the basis of the social identity of individuals within that system. The names themselves have at least in some cases a meaning derived from gender and social roles, e.g. Kunha´pó derives from Kunha (‘woman’) and po (‘make/do/work’), “doing as a woman”. The time intervals that co-constitute (with gender and moiety) the onomastic system are not linguistically independent concepts, that is, they are not (or not all) designated by nominals (although there are nouns for child, adult, and elder). Hence, we cannot say of these time interval concepts that they are “high level events” in the same way as are the seasonal and diurnal time intervals. In fact, from a linguistic point of view they are implicit or covert categories that are, in at least some cases, lexicalised only in conflation with other (gender and moiety) categories, and then only as personal proper names. Life-stage time intervals are thus even further removed from the calendric conception of a time interval than the event-based seasonal and diurnal time intervals. Kinship as a basis for temporal reference is widespread; historical time for the Nuer is largely defined in terms of the initiation-based “age-set system”, and is therefore conceptualised in terms of “the movement of persons, often as groups, through the social structure” (Whitrow 1988: 10). Amondawa time bears yet other similarities to Nuer time as described by Evans-Pritchard (1939, 1940). The social and linguistic construction of time is based upon the interplay between ecological facts in the natural environment, and social facts or structures. The basis for social structure time in Amondawa, as in Nuer culture, is twofold: first, the rhythm of activity, especially work, and second the stages of life constructed in social affiliation, although, whereas for the Nuer this is based upon cohort groups who experienced ritual initiation together, for the Amondawa it is based upon individual “movement” through a kin-defined onomastic system. In the terms that we have employed above, for both Amondawa and Nuer, time intervals are event-based and social, rather than time-based. There are also two notable differences between Nuer and Amondawa time intervals. First, the Nuer employ a “quasi-calendar” of twelve months. Second, the Nuer months can be enumerated, although “Nuer do not reckon [months] as



Vera da Silva Sinha, Chris Sinha, Wany Sampaio and Jörg Zinken

fractions of a [year] unit. They may be able to state in what month an event occurred, but it is with great difficulty that they reckon the relation between events in abstract numerical symbols” (Evans-Pritchard 1940: 103–104). Amondawa time intervals do not include months, and time reckoning is apparently entirely absent from the repertoire of cultural practices. We might hypothesise, then, that while both Amondawa and Nuer time interval systems are event-based, the Nuer system possesses more features potentiating an evolution to a time-based system. Amongst the symbolic resources necessary for the cultural emergence of time-based time interval systems, such as true calendric and clock systems, is the existence of a more elaborate number system than the restricted Amondawa quantificational system. However, comparison with the Nuer case suggests that while necessary, this, in itself, is not sufficient. The Amondawa seasonal and diurnal time interval systems exemplify an exclusively event-based schematisation of time intervals. We suggest that a cultural-historical precondition for the schematisation of time-based time interval systems is the material anchoring of quantified time intervals in symbolic cognitive artefacts for measuring, segmenting, and reckoning time, such as calendar notations and clocks. All human artefacts are in a broad sense cognitive, inasmuch as they embody human intentionality (Sinha 1988; Bloom 1996). However, there is a special subclass of what we can call symbolic cognitive artefacts, which can be defined as comprising those artefacts that support symbolic and conceptual processes in abstract conceptual domains. Examples of cognitive artefacts are notational systems (including writing and number), dials, calendars, and compasses. Cultural and cognitive schemas organising the relevant conceptual domains may be considered as dependent upon, and not merely expressed by, the employment of cognitive artefacts. A key property of cognitive artefacts is thus that they are conventional and normative. Cognitive artefacts may be motivated by natural facts and the human phenomenological experience of these facts (e.g. the orbit of sun or moon; the number of fingers on a human hand), but they are not determined by them (witness, for example, the variety of arithmetical bases for number systems). Symbolic cognitive artefacts are instances of the extended embodiment of cognition (Sinha and Jensen de López 2000), instantiating the intersection of material and symbolic cultural forms. The symbolic systems and conceptual schemas that they support are materially anchored (Hutchins 2005) in the artefacts that support the socio-cognitive practices (and the reproduction of these practices through inter-generational transmission) constituting a segment of the life world of individual and group (Schutz 1966). Symbolic cognitive artefacts are thus a crucial (and species-specific) exemplification of the “ratchet effect” (Tomasello 1999) in

1.  Event-based time intervals in an Amazonian culture

human cultural evolution and development; that is, they stabilize cultural invention and permit cumulative intergenerational transmission. What implications does this analysis hold for understanding time as a concept and as a conceptual domain? We advance two linked hypotheses. First, we suggest that time-based time interval systems and categories are in a fundamental way linguistically constructed, that is, they cannot be “thought” without thinking them through language and for speaking (Slobin 1996). The conceptual schematisation of time-based time interval systems is not based in pre-linguistic and pre-conceptual image schemas (Lakoff and Johnson 1999). Rather, the schemas are actually constituted by the use of linguistically organised, materially-anchored symbolic cognitive artefacts such as calendar systems. Our second hypothesis is that the conceptual domain of Time as Such (or time abstracted from events) is not a human cognitive universal, but a cultural and historical construction constituted by schematised time-based time interval systems, and semiotically mediated by symbolic and cultural-cognitive artefacts for time reckoning. In this respect our analysis converges with Jaszczolt’s contention (HCP 36) that while temporal “location”, temporal relations and (we would add) time intervals are expressed in all languages, this does not require the postulation of time as a conceptual “primitive” – or, to put it another way, slightly divergently from Jaszcolt’s formulation, there are many ways of conceptualising temporality, not all of which depend on a “concept of time”. Langacker (this volume) distinguishes between time as a domain within which a profiled relationship occurs, and time as a profiled object of conception or abstract “thing”, commonly designated as a noun. It is this latter, reified time that corresponds our notion of the conceptual domain of Time as Such, grounded in the construction of time-based time intervals. We did not find any time-based time intervals in Amondawa, and our finding that there is no Amondawa word meaning ‘time’ – speakers using the word kuara, ‘sun’, to translate the Portuguese word tempo – further tells against both the universality of Time as Such, and the contention by Priestley (this volume) that the lexical concept ‘time’ is a semantic primitive or prime. Both Langacker (this volume) and Wallington (this volume) address the phenomenon of spacetime analogical or metaphorical mapping, a phenomenon frequently asserted or assumed to be universal (Fauconnier and Turner 2008). We suggest that what Wallington describes as the (metaphorical) reification of events is concomitant with the cultural historical and linguistic construction of the cognitive domain of Time as Such, the metaphorical “space” containing event-objects including timebased time intervals. Elsewhere (Sinha et al. 2011), we have reported, on the basis of our analysis of the expression of temporal inter-event relations in Amondawa, that linguistic space-time mapping, and the recruitment of spatial language for



Vera da Silva Sinha, Chris Sinha, Wany Sampaio and Jörg Zinken

structuring temporal relations is absent from Amondawa. Our contention is that this constellation of linguistic facts – the absence of time-based time intervals, the absence of lexicalisation of ‘time’, and the absence of space-time analogical mapping – is not accidental, but attests to the still largely neglected importance of socio-cultural processes in language and cognition.

References Bergson, H. L. 1910. Time and Free Will: An Essay on the Immediate Data of Consciousness. Transl. F. L. Pogson. London: Macmillan Bloch, M. 1977. The past and the present in the present. Man 12 (2): 278–292. Bloom, P. 1996. Intention, history, and artifact concepts. Cognition 60: 1–29. Davis, R. 1976. The Northern Thai Calendar and its uses. Anthropos 71: 3–32. Edmonson, M. S. 1976. The Mayan reform of Current Anthropology 17 (4): 713– 717. Evans-Pritchard, E. E. 1939. Nuer time-reckoning. Africa: Journal of the International African Institute 12 (2): 189–216. Evans-Pritchard, E. E. 1940. The Nuer: A Description of the Modes of Livelihood and Political Institutions of a Nilotic People. Oxford: Clarendon Press. Fauconnier, G., and M. Turner. 2008. Rethinking metaphor. In The Cambridge Handbook of Metaphor and Thought, ed., R. Gibbs, 53–66. Cambridge: Cambridge University Press. Geertz, C. 1973. The Interpretation of Cultures: Selected Essays. New York: Basic Books. Gell, A. 1992. The Anthropology of Time: Cultural Constructions of Temporal Maps and Images. Oxford: Berg. Hornborg, A., and J. Hill. In press. Ethnicity in Ancient Amazonia. Boulder: University of Colorado Press. Hutchins, E. 2005. Material anchors for conceptual blends. Journal of Pragmatics 37: 1555– 1577. Keyes, C. F. 1975. Buddhist pilgrimage centers and the twelve-year cycle: Northern Thai moral orders in space and time. History of Religions 15 (1): 71–89. Lakoff, G., and M. Johnson. 1999. Philosophy in the Flesh: The Embodied Mind and its Challenge to Western Thought. New York: Basic Books. Levine, R. 1997. A Geography of Time: The Temporal Misadventures of a Social Psychologist, or: How every culture keeps time just a little bit differently. New York: Basic Books. Menéndez, M. A. 1989. O Tenharim: Uma Contribuiçao ao Estudo dos Tupi Centrais. PhD thesis, Universidade de São Paulo. Munn, N. D. 1992. The cultural anthropology of time: A critical essay. Annual Review of Anthropology 21: 93–123. Ohnuki-Tierney, E. 1973. Sakhalin Ainu Time Reckoning. Man New Series 8 (2): 285–299. Peggion, E. A. 2005. Relações em Perpetuo Desequilibrio: a Organização Dualista dos Povos Kagwahiva da Amazonia. PhD thesis, Universidade de São Paulo. Postill, J. 2002. Clock and calendar time: A missing anthropological problem. Time & Society 11 (2/3): 251–270.

1.  Event-based time intervals in an Amazonian culture

Sampaio, W. 1996. Estudo Comparativo entre Linguas Tupi-Kawahib. PhD thesis, University of Campinas. Sampaio, W. 1999. A referência remissiva número-pessoal nos prefixos verbais da língua urueu-uau-uau. Unpublished manuscript, Federal University of Rondônia. Sampaio, W. & V. da Silva. 1998. Os povos indígenas de Rondônia: contribuições para com a compreensão de sua cultura e de sua história. 2nd ed. Porto Velho: UNIR. Schutz, A. 1966. Collected Papers III: Studies in phenomenological philosophy. The Hague: Martinus Nijhoff. Silva, V., da. 1997. Amondava: Uma Historia de Perdas. Ouro Preto do Oeste: Grupo de Apoio aos Povos Indigenas. Silva, V., da. 2000. Mboxuaxian: Uma Leitura Etnografica da Escola Amondawa. MA thesis. Recife: Universidade Federal de Pernambuco. Sinha, C. 1988. Language and Representation: A Socio-naturalistic Approach to Human Development. Hemel Hempstead: Harvester-Wheatsheaf. Sinha, C., and K. Jensen de López. 2000. Language, culture and the embodiment of spatial cognition. Cognitive Linguistics 11: 17–41. Sinha, C., V. da Silva Sinha, J. Zinken, and W. Sampaio. 2011. When Time is not Space: The social and linguistic construction of time intervals and temporal event relations in an Amazonian culture. Language and Cognition 3: 137–169. Slobin, D. 1996. From “thought and language” to “thinking for speaking”. In Rethinking Linguistic Relativity, eds., J. Gumperz & S. Levinson, 70–96. Cambridge: Cambridge University Press. Tomasello, M. 1999. The Cultural Origins of Human Cognition. Cambridge, MA: Harvard University Press. Vickers, A. 1990. Balinese texts and historiography. History and Theory 29 (2): 158–178. Whitrow, G. J. 1988. Time in History: Views of Time from Prehistory to the Present Day. Oxford: Oxford University Press. Wolf, Eric, R. 1982. Europe and the People Without History. Berkeley: University of California Press. Wright, R. 1991. Time among the Maya. New York: Henry Holt. Zinken, J., W. Sampaio, V. da Silva Sinha, and C. Sinha. 2005. Space, Motion and Time in Amondawa: Field Manual 2005–2006. Portsmouth: University of Portsmouth.


chapter 2

Vagueness in event times An epistemic solution* Minyao Huang University of Cambridge

Vagueness in event times pertains to the observation that one usually finds it difficult to slice the continuous flux of space-time into a series of events with clear-cut temporal boundaries. I argue that such vagueness originates from our ignorance of discrete changing points wherein states of affairs begin or cease to obtain. Applying the epistemic view on vagueness (Williamson 1994) to vagueness in event times, I contend that the lagging nature of knowledge prevents one from knowing the abrupt changes taking place between contiguous minimal time intervals. In turn, vagueness in the conception of event times is construed in terms of epistemic modality (Jaszczolt 2009), viz. a gradation of a thinker’s epistemic certainty towards the occurrence of a physical change. Keywords: discrete changing points, epistemic certainty, lagging nature of knowledge, temporal boundaries

1. Introduction Vagueness in event times pertains to the empirical observation that we usually fail to pinpoint when an event begins and when it ends. While the perception of an event often involves figuring out what has been put into action and/or what has been achieved, we often find ourselves unable to locate where the onset and the endpoint lie. For example, the reader may find it difficult to reflect on when she started reading the current chapter. Moreover, if the conclusion at the end is

* I am indebted to Kasia Jaszczolt for shaping my view in this chapter and providing numerous suggestions for its improvement. This chapter was presented as a paper at Space and Time across Languages, Disciplines and Cultures (STALDAC), 2010. I thank the audience for valuable comments, especially Keith Allan, Nicholas Asher, and Peter Ludlow. I also thank Ian Roberts and Luna Filipović for discussions and suggestions on a previous version of this chapter.


Minyao Huang

correct, the reader would see how difficult it is to know when she would finish the reading. The aim of this chapter is to investigate such vagueness in the temporal boundaries of events. I argue that the nature of the temporal vagueness under discussion is epistemic, rather than semantic or metaphysical. Event boundaries appear to the mind as vague partly because, given the intentional character of events, the conception of event times represents a cognitive parsing of the physical time, and partly because the epistemic process underlying the online conception of event times takes on a lagging character. In turn, epistemic vagueness in the conception of event times would be construed as a gradation of epistemic awareness towards the occurrence of real-world states of affairs. There are two main imports of the current research. On the one hand, since events are the basic units in space and time, the current discussion on the conceptualisation of events along the temporal dimension contributes to our understanding on the conceptualisation of time in general. On the other, if events are also the basic units in the linguistic representation of worldly situations, the source of vagueness in event times sheds new light on what form the linguistic representation of event times could take. To begin with, the best place to illustrate the temporal vagueness in question would be Frege’s example of the yellow-turning tree. In “Thought” (1918– 1919: 343) Frege discussed the thought one entertains in high summer that the deciduous tree in one’s garden is covered with green leaves. In six months’ time, the tree will turn yellow and the above-mentioned thought will be false, if one still entertains it, come winter. Frege’s point there is that without a specification in the temporal dimension, as, say, in high summer or in winter, a thought about the colour of the tree is incomplete in the sense that it cannot be associated with either truth or falsehood (an essential feature of a Fregean thought). That is, if the changing of the colour in time constitutes a truth maker for the thought that encodes it, to entertain the thought in the complete sense would involve keeping track of how the tree changes its colour with the passage of time. However, if the tree changes its colour day by day in a minute and generally indiscernible fashion, one may fail to take notice of any of its vicissitudes. The puzzle is then how could one grasp the passage of time whereby a thought about the luxuriance of the tree changes from being true to the opposite, when the change of truth status along the temporal dimension consists of ungraspable daily changes? Just when does the tree turn from green to yellow? Note that even if one could approach the tree’s colour only vaguely, it does not follow that the object of one’s thought, i.e. the metabolic system of the tree which controls its colour, fails to change precisely.

. See also Chapter 3 in this volume.

2.  Vagueness in event times

The question to be asked is precisely what is responsible for our being vague about how things change? At the outset, it is perhaps worth mentioning that the topic of temporal vagueness has somehow been neglected in contemporary semantic studies. There are several reasons for this neglect. One is the tendency to confine the discussion of event times to the relative relations in which one event stands to other events, in terms of succession and overlap (as in Discourse Representational Theory). Such a bias naturally bypasses the need to tackle temporal vagueness in the absolute relation in which an event stands to its own temporal profile. Another is the difficulty in satisfactorily defining such problematic notions as “event” and “time” in conceptual terms. Although it is generally held that events constitute the contents of time, the elusiveness of the concept of time inevitably causes considerable controversy with respect to what the content of time consists in. Here it is hoped that the following discussion on vagueness in the conception of event times will shed new light on the general issues pertaining to the concept of time. To anticipate, the structure of the exposition goes as follows: in Section 2, I compare temporal vagueness with the much-discussed phenomenon of predicate vagueness. The similarity of these two types of vagueness suggests that major theories in the domain of predicate vagueness automatically become theoretical candidates to model temporal vagueness. Divergent intuitions pertaining to predicate vagueness and temporal vagueness, however, indicate that major approaches to predicate vagueness will fare differently in the context of temporal vagueness when they are judged on the basis of pre-theoretical intuitions, as will be seen in Section 4. In Section 3 I propose a construal of events as a species of intentional objects, which enables us to appreciate its cognitive role in the parsing of the physical space-time. Accordingly, vagueness in the segmentation of event times could be construed as vagueness in the predictability of events, which induces the segmentation in question. Section 4 examines the major theoretic candidates for temporal vagueness by combining the criteria put forward in the preceding two sections, namely the pre-theoretical intuitions pertaining to event boundaries and the requirement of certain cognitive mechanisms responsible for the gradual change of event predictability. The conclusion reached towards the end is that the epistemic view on vagueness, according to which the source of vagueness is our ignorance of precise boundaries (Williamson 1994), fares best in the context of temporal vagueness, as it both befits our pre-theoretical intuitions and implies a plausible mechanism in the form of epistemic modality to account for the gradual change of predictability in the vague conception of events. Section 5 concludes.

. See Kamp and Reyle (1993).


40 Minyao Huang

2. Temporal vagueness vs predicate vagueness By ‘temporal vagueness’ I mean vagueness in the temporal boundaries of events. It is in many ways analogous to the phenomenon of predicate vagueness extensively discussed in the philosophical literature. Crudely speaking, the latter pertains to vagueness about what things are (e.g. whether you are bald), whereas the former is concerned with vagueness about how things go (e.g. whether the tree is turning yellow). Since what things are and how things go are different aspects that go into the determination of truth conditions, both types of vagueness point to indeterminacy in truth conditions. More specifically, both types of vagueness conform to the standard diagnoses of vagueness, namely the existence of borderline cases, the susceptibility to the Sorites Paradox, and the satisfaction of the tolerance condition proposed by Crispin Wright (1975). For example, just as a man with 67,100 hairs on his head represents a borderline instance of the property denoted by the predicate “bald”, a day when half of a tree’s leaves have turned yellow counts as a borderline case wherein it is unclear whether the event described by “the tree’s being covered with green leaves” obtains or not. Moreover, just as one does not change the opinion about a man’s baldness with the growth (or loss) of one hair on his head, one does not stop describing a tree as “being covered with green leaves” after the discolouring of one more leaf. In other words, vague predicates and descriptions are semantically tolerant in the sense that marginal differences in the respects relevant to their applications do not affect one’s verdict with respect to their applicability (Wright 1975: 333). That the tolerance condition gives rise to the Sorites Paradox is standard philosophical lore: it is plainly false to argue, from a deciduous tree’s being covered with green leaves in high summer and its changing colour indiscernibly day by day thereafter, to the paradoxical conclusion that the tree will be covered in green all year round. Hence, a theory of temporal vagueness shares with a theory of predicate vagueness these aims: to explicate what goes wrong in the Sorites reasoning, to propose an appropriate logical framework to model the use of vague terms, and ultimately, to address the source of vagueness. In view of the aforementioned commonalities, major theories in the domain of predicate vagueness, namely fuzzy logics (Zadeh 1965),

. Cf. Smith (2008), where a Closeness condition is proposed in place of the Tolerance condition. The Tolerance condition states that if objects a and b differ little in the aspects relevant to the application of predicate P, P(a) and P(b) do not differ in truth values. In contrast, the Closeness condition states that if objects a and b differ marginally in the aspects relevant to the application of predicate P, likewise P(a) and P(b) differ marginally in truth values. Apparently the Closeness condition is biased towards a fuzzy logical approach to vagueness whereas the Tolerance condition is more neutral in its characterisation of the vagueness phenomenon.

2.  Vagueness in event times

supervaluationism (Fine 1975), contextualism (Kamp 1981; Raffman 1994), epistemicism (Williamson 1994), etc., would be automatically considered as rival theories in the realm of temporal vagueness. Consequently, the focus of Section 4 of the current chapter is to weigh up the aforementioned frameworks with respect to their applicability to temporal vagueness, on the basis of our intuitions and common practices in the event sphere. When selecting an appropriate theoretic framework to model either type of vagueness, it is natural to appeal to our pre-theoretical intuitions. Here it should be noted that our intuitions appear to differ between predicate vagueness and temporal vagueness. For instance, while it is intuitively plausible to say of a borderline bald man that “he is neither bald nor hirsute”, or that “he is sort of bald”, etc., it sounds rather weird to say of a borderline interval that “an event neither happened nor not happened”, or that “an event sort of happened”. Therefore, the intuitive plausibility of the truth-value gap allegedly available in the discussion of predicate vagueness may not be duplicated in the case of temporal vagueness (see Section 4.1). Furthermore, while it is very probable that one cannot conceive a non-arbitrary boundary between n and n+1 hairs as the division between baldness and non-baldness, for some natural events such as explosion, birth, etc., it is conceivable that one particular second (or millisecond) marks the onset and/or endpoint of an event. Even for certain non-natural events, such as World War Two, it is not implausible to take one final stage, such as the end of a surrender announcement, as its endpoint, on the basis of its causal significance in the course of WWII. In this respect, the apparent “boundaryless-ness” (Sainsbury 1990) characteristic of predicate vagueness does not seem to apply readily to temporal vagueness. I come back to the conceivability of event boundaries in the discussion on a supervaluationist approach to temporal vagueness in 4.1. At any rate, if we are to systemise our pre-theoretical intuitions pertaining to events, we need to look closer at the baffling notion of “event”, before deciding on the best model for its temporal vagueness. In particular, we need to examine the relation that the temporal boundaries of events bear to the continuously evolving space-time, before developing an account on the vagueness of the former. In other words, we need to address, in what follows, the question of what it is that is vague in the context of temporal vagueness. . Here the list of major theories on the market is regrettably short. Readers are referred to Smith (2008: Ch. 2) for a state-of-the-art overview of the recent literature. . It should be mentioned that the intuitions appealed to throughout this chapter have not been formally tested. However, they have been informally tested among native speakers and are robust in their intuitive judgements.



Minyao Huang

3. What is it that is vague? To start with, my definition of event is fairly standard. According to Kim (1966, 1973, 1976), an event consists in the exemplification of a property (or n-tuple relation) by an object (or n-tuple objects) in time. My main claim in this section is that events are to be taken primarily as intentional objects. By an intentional object I mean that which a thought is about. For instance, Cleopatra is an intentional object in the sense that one could entertain a thought about Cleopatra, i.e. by thinking of Cleopatra as being such-and-such. Similarly, to say that an event is an intentional object is just to say that one could entertain a thought about the event, or that one could think of the event as taking place in such-and-such manner. My argument for treating events as intentional objects comes mostly from Davidson’s pioneering work on the semantics of event-describing sentences (1967). According to Davidson, the thematic grid of event-describing sentences contains a hidden argument slot ranging over events (1967: 118). Therefore, the logical form is projected as involving the existential quantification over events. Take (1) for example. In Davidson’s proposal, the predicate ‘played’ is translated into a three-place predicate with an extra event variable (omitting tense issues), bound by existential quantification, as in (1a): (1) Lim played the guitar. a. ∃e (Played (Lim, guitar, e))

As a consequence, the existential quantification over events implies that events constitute a category in the conceptual space as the referential base of event descriptions (Lepore 1985: 159). In other words, if event-describing propositions are the linguistic correlates of the event-induced thoughts, that those propositions are about (the existence of) events entails that events are intentional objects that thoughts could be about. Staying close to the original spirit in Davidson’s proposal, one is justified to view events as objects in the intentional states of mind, whatever the latter turn out to be. Further still, given certain assumptions about the representational character of intentional objects, viz. that intentional objects are representations of the world in certain ways, events could be regarded as mental representations of external states of affairs as perceived in one way or another.

. The idea of intentional object here is taken from the concept of intentionality in Husserlian phenomenology. . In post-Davidsonian so-called subatomic semantics (Parsons 1990), more event predicates, e.g. Subject and Object, are introduced into the logical form. For instance, (1a) is revised as ∃e (Played (e) ^ Subject (Lim, e) ^ Object (guitar, e)).

2.  Vagueness in event times

One immediate objection to the above claim is that it is vacuously justified. My answer to this objection is that while assimilating events to the category of intentional objects does not solve the knotty issue about the metaphysical nature of the latter, this move has at least two advantages. Firstly, in doing so we could at least arrive at some sort of minimal consensus about the thought-inducing nature of events, upon which various strands of theorisation on the metaphysics of events could agree. This would in turn ensure that different attempts to explicate the metaphysical nature of events have a shared target of theorisation; otherwise the debate on event metaphysics would become one of mere terminological dispute. Secondly, and more importantly, treating events as a species of intentional objects implies taking them as mind-dependent entities. That is, to say that events are mind-dependent is to say that the individuation of events is relative to the way external states of affairs are comprehended in the mind, i.e. to their modes of presentation in philosophical jargon. To illustrate, compare events to the mindindependent species of facts.10 To say that facts are mind-independent is to say that the individuation of facts pertains only to what is happening in the external world, without appeal to how the external happenings are represented in the internal world. For example, given the physical relation between the earth and the sun with respect to their relative positions and movements, the fact that the sun is rising at 7.00 (local time) in Guangzhou is the same fact as the sun’s setting at 19.00 (local time) in Buenos Aires. It is dictated by astronomical laws that the two events correspond to one and the same happening, hence to one single fact, despite the two different ways we describe it. On the other hand, to the extent that what is being observed relative to one’s horizon varies, the event of the sun’s rising 7.00 (local time) as observed by the people in Guangzhou diverges from the event of the sun’s setting at 19.00 (local time) as observed by the residents in Buenos

. See Crane (2001) for an overview of the metaphysical implications of intentional objects. . Kamp (1979: 142) contains similar arguments when it is said that “The concept of an event is not absolute. What can be singled out as a separate event depends on the tools available for singling it out. There are no good reasons to suppose these tools – the components of some particular conceptual scheme … – to be completely determined by the intrinsic features of the phenomena for whose structural perception they are employed … All we can meaningfully refer to is the class of events relative to a given conceptual scheme; and where it is not mentioned explicitly it is the conceptual scheme of the speaker which must be understood as presupposed”. 10. The event/fact distinction was pioneered by Vendler (1967). See also Asher (2003). My definition of events and my reasons for discussing the distinction differ from these two authors. Asher also suggests that the distinction appears to be less significant “in a Kimean view of events” (2000: 133).


44 Minyao Huang

Aires. In other words, the two events do not coincide in terms of their modes of presentation. Note that the different presentational modes in event perception do not always correlate with the different ways of describing events in linguistic terms (see Chapter 13 in this volume). In the above example, the discrepancy between the two events pertains to how the fact is perceived in relation to one’s horizon, with respect to the moving-up-versus-moving-down classification. The perceptual discrepancy may or may not be encoded in linguistic terms. Therefore, holding events to be intentional objects merely commits one to the viewpoint that perception pre-empts the recognition of events, and ipso facto, the individuation of events. Nonetheless, the empirical fact that the information encoded in linguistic terms often exhausts the information processed in perceptual terms indicates that the individuation of events often coincides with the individuation of event descriptions. In any case, one could hold on to the weaker view of events as intentional objects, even if one is not sympathetic to the stronger position of conflating events with event descriptions.11 Once we accept that events are mind-dependent, a fortiori, the temporal boundaries of events are mind-dependent. It follows further that the division of the continuously evolving space-time into a series of events is best construed as a type of “cognitive imposition” (Wynn 1996: 165). In other words, the perception of events plays the role of parsing the space-time continuum, via the cognitive mechanisms whereby external states of affairs are presented to the mind as events. Consequently, vagueness in the segmentation of the physical space-time by means of the cognitively imposed spatio-temporal boundaries may be due to some sort of cognitive mechanisms responsible for the perception of events. In this regard, psychological evidence might give us a clue about the mechanisms we are looking for that give rise to the vague appearance of temporal boundaries in perceptual terms. According to recent experimental findings on event perception, the appearance of a temporal boundary may arise when the course of external happenings becomes less predictable (Shipley 2008: 12). It is commonly held that the visual processing of external situations generally observes the perceptual cycle (Neisser 1976), wherein information uptake from the environment evokes similar experiences in the past; in turn, the activated background knowledge serves as a guide for further information pickup. Simply put, we are constantly observing, recalling what had followed previously, and predicting what 11. Advocates of the stronger position include Kim (1966) and Goldman (1971). See e.g. Kim (1991: 643), where it is said that “An event name of the form ‘the exemplification by S of P at T’ names (if it names anything) the event which is the exemplification of the property ‘P’ names by the substance ‘S’ names at the time ‘T’ names.”

2.  Vagueness in event times

will come next. Once the incoming information becomes incompatible with the prediction, one would inspect the environment in greater detail to readjust the prediction. Decreased predictability will then initiate a new perceptual cycle, thereby demarcating the boundaries of events (Shipley 2008: 6). Insofar as predictability decreases gradually, the temporal boundaries cannot be observed as abruptly as one may expect them to be. Nonetheless, to reduce vagueness in event times to vagueness in the predictability of events does not immediately answer the question of what vagueness consists in within the context of temporal vagueness. Prediction is in itself a complex cognitive activity that has both epistemological and linguistic underpinnings. On the one hand, reasonable prediction is a function of one’s current knowledge state. It may or may not turn into genuine knowledge pending further proof or denial. Hence, vagueness in the predictability of events may reflect the epistemological process whereby prediction is gradually turned into knowledge. On the other, provided that prediction is often couched in linguistic terms, vagueness in event predictability may also be due to vagueness in the language we use to make predictions. Our next task is then to select an approach to temporal vagueness that (i) complies with our pre-theoretical intuitions in the event sphere and (ii) implies certain predictability-related mechanisms by means of which it will make manifest why event boundaries appear to be vague. Given these two criteria, in what follows I will examine the applicability of three major theoretic candidates to temporal vagueness, namely supervaluationism, fuzzy logics, and epistemicism.

4. Selecting a theory for temporal vagueness The main aim of this section is to test three major approaches to temporal vagueness against our pre-theoretical intuitions and the requirement of certain prediction-generated mechanisms. I argue that neither supervaluationism nor fuzzy logics seem to honour our pre-theoretical intuitions pertaining to events the way epistemicism does. With respect to the predictability-related mechanisms, supervaluationism simply owes us an account of the gradual change of event predictability. While fuzzy logics could in principle account for the steady shift of event predictability, the ensuing ontological commitment is neither sufficient nor necessary to make manifest what the gradation of predictability consists in. All things considered, epistemicism will emerge as the best theory to model temporal vagueness, as it entails the mechanism of epistemic awareness, the latter of which would elucidate the gradation of event predictability.


46 Minyao Huang



I begin with supervaluationism because it appears to be the dominant approach in the discussion of vagueness in the realm of event.12 For the supervaluationist, the indeterminacy of truth values characteristic of vagueness in general (see Section 2) is taken at face value as indeterminacy of our semantic competence, in the sense that the semantic rules governing the usage of vague locutions fail to specify any precise applications for them. The indeterminacy of vague predicates is construed as the prevailing absence of decisive apparatus within the semantics to specify a non-arbitrary boundary out of an array of equally acceptable (i.e. admissible) boundaries. Accordingly, a borderline proposition is one that is true for some admissible boundaries but false for the others, resulting in a lack of truth value, viz. falling into a truth-value gap. When it comes to temporal vagueness, the gist of a supervaluationist account is that a borderline temporal interval would be one wherein an object neither exemplifies nor not exemplifies a property.13 In particular, for some admissibly precisified temporal boundaries the exemplification takes place, whereas for others it does not. Crucially, the precisifications are alleged to be equally acceptable, with nothing in the semantics to decide among them. My case against the application of a supervaluationist framework to temporal vagueness has several points. First, it is counter-intuitive to say of a temporal interval that neither a property instantiation takes place nor it does not. Apparently something either happened or not happened. Thereby, it is inconceivable that a temporal interval intervenes between the exemplification of a property P and the exemplification of its opposite ¬P, except that the interval in question is an empty one. The inconceivability of truth-value gaps in the context of change is a direct consequence of the general assumption of the discreteness of changes.14 Second, it is part and parcel of a supervaluationist account that admissibly precisified temporal boundaries should be equivalently acceptable. If this is so, it 12. See e.g. Varzi (2002) and Borghini and Varzi (2006), where a supervaluationist approach to the spatial vagueness of event is defended. Here I disagree with these authors particularly with respect to their neglect of the potential of the epistemic view in modelling event vagueness (Varzi 2002: 245; Borghini and Varzi 2006: 313). 13. Here I follow the standard practice in formal semantics of relativising truth conditions to an interval as opposed to an instant. Sentences that express a change of state are true of an interval, “no matter what its size, if the interval is bounded at one hand by one particular state of affairs and at the other by another particular state” (Dowty 1979: 139). In particular, an instant would be construed as a minimal interval that contains no other interval besides an empty interval and itself (Kamp and Reyle 1993: 501). 14. In Kamp (1979: 162), this assumption is referred to as “the Principle of Discrete Change”.

2.  Vagueness in event times

would be pointless to hold any significant debate or genuinely conflicting opinions about the exact starting and/or end points of events. Now imagine that archaeologists are debating on whether a new discovery of a new dinosaur fossil above the K-T boundary (the geological signature of the K-T extinction event) should force them to push the temporal boundary of the extinction event beyond the Mesozoic Era.15 However, if the temporal vagueness of the event denoted by “the extinction of the dinosaur” is a matter of insoluble indeterminacy among some set of semantically warranted precisifications, the imagined debate would not be a substantial one on the actual temporal boundaries of the event, but merely a nominal one on whether one should revise the meaning of “the extinction of the dinosaur” by associating with it a different set of semantically warranted precisifications. A supervaluationist’s indifference to such questions as when the last gasp of the dinosaur took place is of course at odds with the standard practice of taking the relevant debates seriously. To reduce debatable temporal boundaries to admissible ones is to deny the significance of the need to search for more accurate resolutions of the temporal boundaries. Ditto the debate among historians with respect to the temporal boundaries of historical events such as WWII, by evaluating the causal significance of disputable choices. In sum, the common practice of regarding event boundaries as debatable, as opposed to admissible, suggests that a supervaluationist approach to temporal vagueness is killing too many birds with one stone. Third, a supervaluationist approach to temporal vagueness also excludes the possibility of making meaningful predications in borderline temporal intervals. Provided that truth values are missing in these intervals, predictions made in them would be groundless and incredible. Note that a prediction pending proof or denial is not the same as a groundless prediction. A prediction pending proof or denial is a prediction grounded in the to-be-discovered truth or falsity of a proposition.16 Unknown truth status is not the same as lack of truth status. In general, a supervaluationist account gives little explanation on the gradual change of predictability characteristic of the temporal vagueness in events (see section 3). A supervaluationist may contend that one could analyse the event predictability of a borderline interval in accordance with the proportion of admissible precisifications that make true a prediction made in the interval to falsehoodmaking precisifications. However, such a move appears to put the cart before the horse. Even if the predictability could be measured in terms of the proportion, 15. The dinosaur example is due to Peter Ludlow in his comment on my conference presentation. 16. That predictions are taken as either true or false is in tune with our Cognitive Tendency of Bivalence (Burton-Roberts 1999).


48 Minyao Huang

the latter does not ipso facto constitute an explanation for the former. It is because predictability varies as an event unfolds that the proportion of truth-making admissible precisifications to falsehood-making ones varies, not vice versa. Given that one cannot envisage the exact unfolding of an event, if the proportion in question reflects a particular stage in its course, it follows that one cannot know a priori the proportion of truth-making admissible precisifications to falsehoodmaking ones. Thus, supervaluationism gives us no explanation of how event predictability varies in borderline intervals, and a fortiori, why vagueness arises in those intervals.17 In a word, supervaluationism is ruled out as inadequate to model temporal vagueness, because it lacks either intuitive plausibility or sufficient mechanisms to explicate the gradation of predictability characteristic of temporal vagueness. 4.2

Fuzzy logics

At first approximation, fuzzy logics maintain that the indeterminacy of truth values in a bivalent system is a result of truth values coming in degrees ranging from 0 (definitely false) to 1 (definitely true). In between these two extremes there exists a cline of truth values, corresponding to the gradual transition from being definitely true to being definitely false. Classical bivalent logics are revised so that borderline instances are assigned degrees of truth between 0 and 1. For instance, if a particular day is a borderline instance of the tree’s being covered in green leaves, the day in question would be associated with a degree of, say, 0.5, with respect to the tree’s luxuriance.18

17. My arguments against semantic indeterminacy in the case of temporal vagueness also apply to contextualism, an approach to predicate vagueness that inherits from supervaluationism the claim about semantic indeterminacy. Contextualism departs from supervaluationism primarily in the addition of “internal contextual factors” (Raffman 1994: 65) – such as the speaker’s classification tendency (Raffman 1994), the relevant standards, the interlocutor’s interests (Graff 2000), cognitive prototypes, etc. – that go into the determination of the extension of vague predicates. Contextualists commonly rely on the mechanism of context shifts to explain why the Sorites Paradox is both untenable and unattractive. I refrain from the application of contextualism to temporal vagueness because it is unclear how the posited contextual factors may function in the judgement of a borderline temporal interval. 18. It is commonly held that the designation of a particular number as the degree of property instantiation is rather arbitrary. That is, it is purely artificial to assign the degree of 0.5, instead of 0.55 or 0.6, to a borderline case. Such arbitrariness is one of the major weaknesses of fuzzy logics. Nonetheless, my argument below against the applicability of fuzzy logics to temporal vagueness does not hinge on this arbitrariness.

2.  Vagueness in event times

My argument against the application of degree theories to temporal vagueness is twofold. First, it makes little sense to say of an event that it happens to a certain degree. While it is intuitively plausible to say of, for instance, a sample of water that it is pure to the degree of 99 per cent, it is odd to say of the water in the headstream of a river that it is flowing to the degree of 99 per cent. Second, fuzzy logics as applied to predicate vagueness are typically committed to the position of ontological vagueness. That is, if degrees of truth are on a par with definite truth and definite falsehood in the metaphysics, an object is mapped onto the extension of a property not in terms of whether the former instantiates the latter, but in terms of the degree to which the former instantiates the latter. Nevertheless, the above metaphysical commitment cannot readily extend to the case of temporal vagueness, if the mapping of an interval onto the extension of a property-exemplification represents first and foremost a cognitive imposition (see Section 3; also Chapter 1 in this volume on event-based conception of temporal intervals). Since event perception plays the role of parsing the spatio-temporal continuum into a series of event units, vagueness in the parsing is vagueness in the cognitive domain. In other words, taking events as intentional objects obviates the need for positing ontological vagueness in event times. Therefore, although at first blush it would sound attractive to correlate the gradual transition of degrees of truth with the gradual transition of event predictability, this move is unpromising if any degrees intelligibly attributed to event times to model the gradation of predictability would be degrees born out of the cognitive realm. I will come back to what these degrees of predictability might be after the discussion of the epistemic view as applied to temporal vagueness. For the present purpose, it suffices to bring home the point that fuzzy logics is neither sufficient in explaining our intuitions pertaining to events, nor necessary in revealing the sort of degrees suitable for an account of temporal vagueness. 4.3


We are left with epistemicism, according to which events have sharp but unknowable temporal boundaries. The boundaries appear to be vague because our knowledge about them in borderline intervals is ill-formed. Given that knowledge is a “safe bet”, the formation of knowledge about the happening of an event requires that the event has been in place for some time. Conversely, there will be time intervals wherein an event has happened but knowledge that it has happened has not fully arisen. The lagging nature of knowledge is generalised as the principle of margin for error (Williamson 1994: 227):



Minyao Huang

Knowledge that a condition obtains is available only if it does obtain (whether knowably or not) in all sufficiently similar cases. If it obtains in a case sufficiently similar to a case in which it does not obtain, then knowledge that it obtains is unavailable in both cases. It cannot be known to obtain within its margin for error.

In particular, if the change from the instantiation of a property P to the instantiation of its opposite ¬P, which marks the onset and/or endpoint of an event, is discrete (Kamp 1979: 162; see Section 4.1), the discrete changing point (viz. a change of truth values of an event-describing proposition) must be flanked by two (minimal) intervals differing in truth values.19 Nonetheless, according to the principle of margin for error, if a state of affairs has obtained in a minimal interval adjacent to an interval wherein it has not obtained, knowledge that it has obtained will not be available. Conversely, for one to know that an event has just taken place, the event must have been taken place for at least some time, which is paradoxical! In accordance with the epistemic view on temporal vagueness, event boundaries appear to be vague because knowledge is lagging behind discrete changes that constitute sharp temporal boundaries. There are at least two reasons why I think epistemicism is the most appropriate theoretic framework for temporal vagueness. First, it is in conformity with our intuition that knowledge about the happening of an event often typically arises after the event has happened, as Examples (2)–(3) show: (2) Our argument turned into shouting before we knew it. (3) After many strokes, I gauge that I have finished the painting. Yet the canvas has been in its final state when I gauge.

As a result of the delay of knowledge in the conception of events, it makes more sense to talk of “knowledge gap” than truth-value gap (see Section 4.1) in the context of temporal vagueness. Second, once we adopt the epistemic view on temporal vagueness, it is easy to see what epistemic mechanism is at work that gives rise to the gradual change of event predictability. Since knowledge implies epistemic certainty, and epistemic modality comes in degrees from mere possibility to absolute necessity in terms of the speaker’s judgement of the likelihood that the event has taken place (Nuyts 2006: 6), it is not implausible to construe the knowledge gap as an augmentation of epistemic certainty as to the occurrence of physical changes. In other words, if one wishes to construe the step-up of event predictability in terms of degrees, the degrees in question are likely to be degrees of epistemic commitment.

19. See Note 12.

2.  Vagueness in event times

To bring home the connection between epistemic vagueness and epistemic commitment, consider again the initial and final phases of an event. At the beginning, one may not be certain that the actual state of affairs has (just) obtained, because other possible states of affairs may have obtained.20 For example, during the first step of a walk, one may have caught sight of a newsflash on a street TV and stood still. During the first sequence of a dance, one may have stumbled and fell down. The point is that at the initial phase, one cannot foresee whether the perceived motion would turn into a walk or a halt, a dance or a stumble. Thereby, one cannot reliably believe that a walk or a dance is taking place; for one cannot ascertain that a stumble or a halt will not be the case. Analogous comments apply to the final stages of an event. Towards the end of a process, one may not be certain that the planned goal will be obtained, because the outcome may turn out to be something else. For instance, one may add some more strokes to finish a painting. During a flight, the pilot may be informed of a volcano eruption and decide to land the plane in a different destination. Again the point is that one may not know whether a course of action would achieve its original goal, until the actual outcome has occurred. Thus one cannot reliably believe that one is moving towards an envisaged end; for one cannot ascertain that a different outcome will not be the case. In general, such epistemic uncertainty reflects the fact that initially, the probabilistic contrast between the actual event and other possible events may not be stark, hence the low predictability of the actual event. Contrarily, diminished probability of other possible events would strengthen the epistemic certainty of the former. For example, during the first few steps of a walk, the probabilistic contrast between the actual walking and other possible events grows. The greater the probabilistic contrast, the higher the predictability of the actual event. Similarly, after the decision to stop painting, the probabilistic contrast between the actual last stroke and other possible outcomes maximises, so that one is most certain about the final product of the painting. In a word, the emergence of a temporal boundary may consist in a build-up of epistemic certainty towards what is taking place in the environment.21

20. I thank Kasia Jaszczolt for suggesting this point to me. 21. In a similar vein, epistemic modality has been invoked in the semantic analysis of the English progressive aspect; see e.g. Dowty (1979), Landman (1992), etc. The difference between the current analysis and other accounts lies in that while others are motivated by the “progressive paradox” in the progressive aspect of telic predicates (e.g. ‘One is eating an apple’ does not entail that ‘one has eaten an apple’), my analysis is primarily driven by the unknowability of the onsets and endpoints of events prescribed by atelic and telic predicates.



Minyao Huang

Lastly, now that we have established the prevailing effects of the knowledge gap overlapping the onsets and endpoints of events, it may not be psychologically real to project the temporal profile of events as bounded by discrete changing points tout court. Rather, we would desire to take into consideration the delay of knowledge in depicting a cognitively real picture of event perception. In this regard, the current analysis lends support to the representation of the concept of time by means of epistemic modality (Jaszczolt 2009), on the grounds that epistemic modality seems to underlie the conception of the temporal boundaries of events.

5. Conclusions To conclude, in this chapter I have defended an epistemic solution to temporal vagueness observed in the conception of event times. My strategy for establishing the case for epistemic vagueness has been to comparatively investigate the applicability of major theories of predicate vagueness to temporal vagueness, given two main criteria requisite to an account of temporal vagueness. One of them pertains to our pre-theoretical intuitions concerning the temporal boundaries of events. The other pertains to the intentional nature of events, and in particular the conceptual role played by events in the parsing of the continuously evolving spacetime. In turn, this would suggest that temporal vagueness consists in the inexact conception of event times in the form of the decreased predictability of events. Assuming these two criteria, I have illustrated the incompatibility of supervaluationism and fuzzy logics with temporal vagueness. The former fails to respect our intuitions about the conceivability and debatability of temporal boundaries. It is also deficient in explaining the gradation of event predictability. The latter’s commitment to metaphysical degrees conflicts with (i) the intuitions about the absoluteness of the happening of events and (ii) the non-metaphysical character of event predictability. I have then vindicated the epistemic view on temporal vagueness by first honouring the intuition that knowledge about the occurrence of a physical change often lags behind the change itself. Moreover, the chief advantage of epistemicism over other approaches in the context of temporal vagueness is that by involving the epistemic notion of knowledge, it is easy to project the gradual change of event predictability in terms of epistemic commitment. That is, vagueness in event times could be construed as the gradation of epistemic awareness towards what is happening in the external world. In sum, all I have argued for is that vagueness in the conception of event times is due to the gradual build-up of epistemic commitment to discrete physical changes.

2.  Vagueness in event times

References Asher, N. 2000. Events, facts, propositions, and evolutive anaphora. In Speaking of Events, eds., J. Higginbotham, F. Pianesi, & A. C. Varzi, 123–150. Oxford: Oxford University Press. Borghini, A., and A. C. Varzi. 2006. Event location and vagueness. Philosophical Studies 128: 313–336. Burton-Roberts, N. 1999. Presupposition-cancellation and metalinguistic negation: A reply to Carston. Journal of Linguistics 35 (2): 347–364. Crane, T. 2001. Intentional objects. Ratio 14: 336–349. Davidson, D. 1967. The logical form of action sentences. In Essays on Actions and Events, 2nd edition (2001), 105–122. Oxford: Clarendon Press. Dowty, D. R. 1979. Word Meaning and Montague Grammar. The Semantics of Verbs and Times in Generative Semantics and in Montague’s PTQ. Dordrecht: D. Reidel. Fine, K. 1975. Vagueness, truth, and logic. Synthese 30: 265–330. Frege, G. 1918–1919. Thought. Trans. by P. Geach & R. H. Stothoff in 1984. In The Frege Reader, ed., M. Beaney (1997), 325–345. Oxford: Blackwell. Goldman, A. I. 1971. The individuation of action. Journal of Philosophy 68: 761-774. Graff, D. 2000. Shifting sands: An interest-relative theory of vagueness. Philosophical Topics 28 (1): 45–81. Jaszczolt, K. M. 2009. Representing Time: An Essay on Temporality as Modality. Oxford: Oxford University Press. Kamp, H. 1979. Some remarks on the logic of change, Part I. In Time, Tense and Quantifiers: Proceedings of the Stuttgart Conference on the Logic of Tense and Quantification, ed., C. Rohrer, 135–179. Tübingen: Niemeyer. Kamp, H. 1981. The paradox of the heap. In Aspects of Philosophical Logic, ed., U. Mönnich. Dordrecht: D. Reidel. Kamp, H., and U. Reyle. 1993. From Discourse to Logic: Introduction to Model-Theoretic Semantics of Natural Language, Formal Logic and Discourse Representation Theory. Dordrecht: Kluwer. Kim, J. 1966. On the psycho-physical identity theory. American Philosophical Quarterly 3: 227–235. Kim, J. 1973. Causation, nomic subsumption and the concept of event. Journal of Philosophy 70: 217–236. Kim, J. 1976. Events as property exemplifications. In Action Theory: Proceedings of the Winnipeg Conference on Human Action, eds., M. Brand & D. N. Walton, 159–178. Dordrecht: D. Reidel. Kim, J. 1991. Events: Their metaphysics and semantics. Philosophy and Phenomenological Research 51: 641–646. Landman, F. 1992. The progressive. Natural Language Semantics 1: 1–32. Lepore, E. 1985. The semantics of action, event, and singular causal sentences. In Actions and Events: Perspective on the Philosophy of Donald Davidson, eds., E. Lepore & B. P. McLaughlin, 151–161. Oxford: Basil Blackwell. Neisser, U. 1976. Cognition and Reality: Principles and Implications of Cognitive Psychology. San Francisco: W. H. Freeman. Nuyts, J. 2006. Modality: Overview and linguistic issues. In The Expression of Modality, ed., W. Frawley, 1–26. Berlin: Mouton de Gruyter.



Minyao Huang

Parsons, T. 1990. Events in the Semantics of English: A Study in Subatomic Semantics. Cambridge, MA: MIT Press. Raffman, D. 1994. Vagueness without paradox. The Philosophical Review 103 (1): 41–74. Sainsbury, R. M. 1990. Concepts without boundaries. Inaugural Lecture, Department of Philosophy, King’s College, London; reprinted in Vagueness: A Reader (1996), eds., R. Keefe & P. Smith, 251–264. Cambridge, MA: MIT Press. Shipley, T. F. 2008. An invitation to an event. In Understanding Events: From Perception to Action, eds., T. F. Shipley & J. M. Zacks, 3–30. Oxford: Oxford University Press. Smith, N. J. J. 2008. Vagueness and Degrees of Truth. Oxford: Oxford University Press. Varzi, A. C. 2002. Event, truth, and indeterminacy. The Dialogue 2: 241–264. Vendler, Z. 1967. Linguistics in Philosophy. Ithaca, NY: Cornell University Press. Williamson, T. 1994. Vagueness. London: Routledge. Wright, C. 1975. On the coherence of vague predicates. Synthese 30 (3/4): 325–365. Wynn, K. 1996. Infant’s individuation and enumeration of actions. Psychological Science 7: 164–169. Zadeh, L. A. 1965. Fuzzy sets. Information and Control 8: 338–353.

chapter 3

Aspectual coercions in content composition Nicholas Asher and Julie Hunter

Université Paul Sabatier, Toulouse / Universite de Pau et des Pays de l’Adour, IJN & EHESS

This chapter investigates the encoding of temporal information in aspect, and how that information may shift based on contextual factors. Analyzing these shifts leads to a more complex view as to how temporal information in a sentence or a discourse results from meaning composition than standardly assumed. Keywords: aspect, contextual meaning, temporal information

1. Introduction Coercions occur when a predicate appears to shift the semantic nature or type of one of its arguments. For instance, the natural interpretation of (1) James enjoyed the book.

is that James enjoyed doing something (presumably reading or writing) with the book. In other words, the argument of enjoy appears to shift from something of type physical object to something of type eventuality. Thanks to the work of Pustejovsky (1995), these coercions have received extensive attention, though their exact formal analysis still remains a matter of debate. This chapter examines a specific kind of coercion known as aspectual coercion, which affects the temporal interpretation of a sentence or text. If one assumes that temporal information is carried by spatio-temporal entities like eventualities (events and states) or by temporal entities like times or temporal intervals, then aspectual coercion involves transforming the argument of some temporal predicate in a way that would be unexpected from a simple compositional calculation of meaning. We look at three types of related aspectual coercions in this chapter.


Nicholas Asher and Julie Hunter

In contrast to many of the chapters in this volume that adopt a cognitive approach to the interpretation of temporal or spatial expressions in language, our approach is couched within the framework of formal compositional semantics and uses tools like the lambda calculus that are standard in this area (ter Meulen’s contribution to HCP 36 is similar in this respect to ours, though it deals with a different parameter of context sensitivity than we do). While formal approaches have considerable predictive power precisely because of the demands of formalisation, coercions pose a problem because the standard homomorphic mapping from syntax to semantics and the simple use of function application does not suffice for an adequate analysis. We show that nevertheless a precise account of lexical meaning can do justice to such phenomena and has the expressive power to provide analyses for the rich set of empirical data discussed in this volume. 1.1

Coercions with progressive aspect

One of the truisms about progressive aspect is that it does not apply to stative constructions – (2a), for instance, supports this generalisation. Nevertheless, (2b–d), which are progressivisations of the stative John is silly, John is John, John is an asshole, are perfectly unproblematic. (2)

a. b. c. d.

#John is knowing French John is being silly. John is just being John. John’s being an asshole.

(2b–d) no longer describe a state but an activity that results in the state of being silly. The coercions in (2b–d) are a language-specific phenomenon. Languages like French that lexicalise progressive aspect do not seem to support this meaning shift: (3) a. Jean est idiot. b. #Jean est en train d’être idiot. c. Jean est en train de faire l’idiot.

This type of coercion is therefore not the result of a general cognitive principle of strengthening or weakening due to Gricean or Neo-Gricean constraints on communication.

3.  Aspectual coercions in content composition


Aspectual coercion and modality

Another sort of aspectual coercion results when a perfective aspectual operator applies to VPs containing an ability modal. Consider the following French examples: (4) a. Jeanne a dû prendre le train. → Jeanne a pris le train. / Jeanne a pris le train. b. Jeanne devait prendre le train. → (5) a. Jeanne a pu prendre le train. → Jeanne a pris le train. / Jeanne a pris le train. b. Jeanne pouvait prendre le train. →

The ‘→’ signifies what is known in the literature as an actuality entailment. Why are the actuality entailments interesting? Treating ability modals as true modals – i.e. as symbolised by K and L – provides the following formalisation of the actuality entailment: (6) a. Kϕ → ϕ b. Lϕ → ϕ or ϕ → Kϕ.

which implies a collapse of the modality (Bhatt 1999). With the imperfective aspect, however, these inferences vanish. There is no collapse. The puzzle is, why does the use of the aspectual perfect collapse the modality? This is unpredicted and indeed bizarre on a Montagovian view of composition. Actuality entailments with the perfective aspect are also a language-specific phenomenon. In English, for instance, the actuality entailment is at least not a hard entailment, as the following sentences are coherent: (7) John was able to take the train, but he ended up taking the bus instead. (8) John refused to go to the meeting because he had to take the train to get to it. (9) ?John has been able to take the train for years, but he never does.

None of these have the actuality entailment, though the ability modal clauses might have what one could call an actuality implicature. Thus, the actuality entailment cannot be the result of some general cognitive but non-linguistic principle of strengthening. 1.3

Temporal adverbials and aspectual coercion

Matters are still more complex in the interaction of aspect because of the way temporal adverbials interact with modality and aspect to produce or to block actuality entailments. In (10) the actuality entailment seems to hold, even though it



Nicholas Asher and Julie Hunter

is generally agreed that the imperfective does not license the actuality entailment. Thus, the actuality entailment’s existence must be due to the interaction between the adverbial and aspect. Similarly, it appears that certain adverbials can block the actuality entailment as in (11), as argued by Mari & Martin (2007). (10) #Soudain, Jean pouvait ouvrir la porte, mais il ne l’a pas fait. (11) Le robot a même pu repasser les chemises à un stade précis de son développement mais il ne l’a pas fait.

We believe the coercions we have discussed should receive as uniform a treatment as possible within a semantic/pragmatic framework of lexical meanings and how they combine together to form meanings for larger semantic constituents like propositions or discourses. But we can only address this issue adequately within a larger view of how context affects interpretation. This gives us the tools with which to understand contextual effects at the level of clausal content composition and apparent meaning shifts. We then apply the machinery to aspectual coercions.

2. Background Almost all words that function as predicates introduce constraints on their arguments in the form of selectional restrictions. For instance, the verb try requires that its subject be an intentional agent; a verb like hit imposes the restriction that its object or internal argument must be a physical object. Thus, (12) Mary hit John’s idea.

is predicted to be difficult to interpret unless the context allows us to interpret John’s idea as some sort of physical object. Selectional restrictions of an expression ∈, unlike syntactic constraints of number and gender agreement, pertain to the type of object denoted by the expression with which ∈ must combine. However, this information about the type of argument is not of a piece with the asserted content of the predication but is rather a type of presupposed content. Selectional restrictions resemble presuppositions because their satisfaction seems to be a prerequisite for any expression containing them to have a well-defined semantic value. Their demands for

. These observations are due to Vincent Homer. . That selectional restrictions are type presuppositions is a fundamental principle of Asher (2011).

3.  Aspectual coercions in content composition

satisfaction or justification percolate up the semantic construction very much in the way that ordinary presuppositions do. In dynamic semantics, presuppositions constitute a particular sort of test on the input context. Consider a sentence like (13) Jack’s son is bald.

The presupposition generated by a definite noun phrase like Jack’s son (namely, that Jack has a son) must be satisfied by the input context, if the interpretation of the rest of the sentence containing the definite is to proceed. One way of satisfying the presupposition in (13) is for it to be already established in the context of utterance of (13) that Jack has a son. This can occur, for instance, when (13) is preceded by an assertion of Jack has a son. Presuppositions can also be satisfied by contexts within the scope of certain operators, as in (14), even though it has not been established in the discourse context that Jack has a son: (14) If Jack had a son, then Jack’s son would be bald.

In dynamic semantics the content of the antecedent of a conditional is added hypothetically to the input context, and it is this hypothetical context that serves to interpret the consequent and the presupposition it generates. Thus, the satisfaction of the presupposition by the antecedent of the conditional in (14) means that the presupposition places no requirement on the actual input context to the whole conditional, the context of utterance. The satisfaction of the presupposition by elements of the discourse context entails that the presupposition does not “project out” as a requirement on the context of utterance. Thus (14) is consistent with the assertion that in fact Jack has no son. On the other hand, dynamic semantics predicts that if we change (14) just slightly so that the antecedent does not provide a content that satisfies the presupposition, the presupposition projects out as a requirement on the input context to the whole conditional: (15) If Jack were bald, then Jack’s son would be bald too.

Selectional restrictions act in the same way in similar contexts. For instance, to say something like (16) The number two is blue.

. One of the great successes of dynamic semantics has been to show that the behaviour of presuppositions introduced by material within the consequent of a conditional follows straightforwardly from the conception of the conditional as a complex test on the input context and thus offers a solution to the so-called projection problem.


60 Nicholas Asher and Julie Hunter

is to invite at the very least quizzical looks from one’s audience, unless the context makes clear that the number two refers to some sort of physical object and not the only even prime. However, a counterfactual with an admittedly bizarre antecedent can satisfy the type presupposition projected from (16) in much the same way as the antecedent of (14) satisfies the presupposition of the consequent: (17) If numbers were physical objects, then the number two would be blue.

What happens when a presupposition cannot be satisfied by the discourse context? It depends on what sort of presupposition is at issue. Some presuppositions, such as many introduced by definite noun phrases, are easily “accommodated”. In dynamic semantic terms this means that the input context is altered in such a way that the presupposition is satisfied, as long as the result is consistent. Other presuppositions, such as that generated by the adverbial too, are much less easily accommodated. Given that operators like conditionals can add “intermediate” contexts between the context of utterance and the site where the presupposition is generated, we need a theory of where and how to accommodate in case the input context does not satisfy the presupposition. In some theories of presupposition that operate on semantic representations, like that of van der Sandt (1992), accommodation simply involves adding a formula for the presupposition to an appropriate part of the representation for the discourse. Van der Sandt stipulates a particular procedure for handling the binding and accommodation of presuppositions: one tries to bind a presupposition first, and one tries first to bind it locally and then in a more long distance way. Then, if binding fails, one tries to accommodate at the outermost context first and if that fails then one tries to accommodate at the next outermost context. The constraint on accommodation is that the addition of the presuppositional material be consistent with the discourse context. So, for instance, one cannot add the presupposition that Jack has a son to a context where it is established that Jack does not have a son. Something similar happens with selectional restrictions. In effect selectional restrictions are type presuppositions. Type presuppositions in normal circumstances are bound or justified in that the type of the argument expression matches the type presupposition of its predicate. But they can sometimes be accommodated. Consider the noun water. It can combine with determiners that require either a noun that denotes something of type mass (18a) or with determiners that are intuitively count determiners:

. In other theories like Heim (1983) the accommodation procedure is not really well developed; see Beaver (2001) for a detailed account of accommodation in a Heimian approach to presupposition.

3.  Aspectual coercions in content composition

(18) a. some water b. a water

One way of accounting for this is that water itself does not determine its denotation to be either of a subtype of type mass or of type count. If that is the case, then we can accommodate the requirements of the determiner simply by applying the type count or mass to the type of the expression water – in simplified terms, a water ends up denoting a property of properties that have a non-empty intersection with the collections of portions of water.

3. From type presupposition to coercion Sometimes the argument does not satisfy the type presupposed by its argument and cannot be accommodated or bound in the given context. In that case, semantic composition crashes and there is no well-defined value for the semantic composition, as in (12). What is not so obvious is what the principles are for accommodating type presuppositions that sometimes permit the rescue of a predication in those cases where the argument’s type does not satisfy the predicate’s type presuppositions. Examining this issue takes us to the heart of coercion. Consider the following example, discussed at length in the literature, in particular by Pustejovsky (1995). (19) Julie enjoyed the book.

The intuition of many who work on lexical semantics is that (19) is equivalent in meaning to: (20) Julie enjoyed doing something (e.g. reading, writing, …) with the book.

The idea is that enjoy requires an event as its direct object as in enjoy the spectacle, enjoy the view. When the direct object of a transitive use of enjoy does not denote an event, it is “coerced” to denote some sort of eventuality. It is difficult, however, to understand what coercion is. Is it, for instance, just the transformation of the denotation of the noun phrase, such as the book, into some sort of eventualitydenoting expression? Asher (2011) shows that neither the meaning of the predicate nor the argument shifts. What shifts is how these two combine. In effect the predicational relation changes, and this change is governed by lexical content and the composition process. In other words, many coercions result from a semantic rather than pragmatic process. That is the approach we adopt here. . Not all mass nouns work so well.



Nicholas Asher and Julie Hunter

Our view requires the background assumptions that (a) verbs, and predicates more generally, distinguish between arguments that denote eventualities and those that denote, say, physical objects and (b) there is a clear distinction between physical objects and eventualities. But both of these assumptions are cogent and well supported linguistically across a wide spectrum of languages (Asher 2011). In the next section, we sketch a formal system that works out our view technically.

4. A sketch of a formal theory of lexical meaning We now turn to implement talk of type presuppositions in a formal framework. Predicates impose two sorts of type constraints on the context to come and the types of their arguments: one is an absolute requirement; the other, which doesn’t always hold, licenses a modification of the predicational environment between the predicate and its argument. We code these type constraints in a store π, so that presupposition justification mechanisms can operate on this store independently of the basic rules of the λ calculus. This means complicating the lexical entry of all words to include a context parameter in which these type presuppositions can be encoded. This parameter functions as a locus of constraints on the context to come within the framework of continuation semantics, and it reflects our hypothesis that type presuppositions are a constraint on the context to come. The type parameter in continuation semantics can be an arbitrary data structure on which various operations may be performed separately from the operations of β reduction in the λ calculus. This is the sort of flexibility we need in order to handle various operations of presupposition justification at the lexical level. There are three operations on type presuppositions. Binding occurs when the type presupposition is a super type of the type of the argument and thus satisfied by subtyping; in this case the type presupposition is met and the more specific type information is simply carried forward in the context. Simple Accommodation occurs when the two types are compatible in that they have a meet that is not ⊥; the result is to take the meet as the result of the operation. The third sort of operation consists of those that allow the readjustment of the predicational relation, one of which we consider in detail below. These operations together with the standard lambda calculus comprise the relevant fragment of type composition logic or TCL, the system of Asher (2011), that we need to look at aspectual coercions. While Asher (2011) details a general architecture for the passing of type presuppositions from predicate to argument, we look here only at how verbs and their projections make use of type presuppositions. The lexical entries chosen determine how presuppositions percolate through the derivation tree, and they predict that presuppositions are typically justified locally to the argument’s typing

3.  Aspectual coercions in content composition

context (the π that determines the typing of the argument). In general, predicates must pass their typing presuppositions onto their arguments. More specifically, a VP should pass typing presuppositions to its subject, and a transitive verb should pass its typing requirements to its object. For example, in the book is heavy, the type requirements made by the verb phrase on its subject argument should be satisfied or accommodated in order for the predication to succeed. In keeping with standard assumptions about the syntax-semantics interface, this means we must rethink our lexical entries to track these presuppositions and to put them in the right place. Consider an intransitive verb like fall. We want its type presuppositions to percolate to the subject. We can do this by assuming that an intransitive verb, or more generally a VP, is a function from DP denotations to propositions. This function feeds the appropriate type presuppositions to the subject DP in the same way that nouns feed their presuppositions to their modifiers. In order to do this properly, we must make our verbal predicates be relatively catholic as to the type of DP they accept. The specific type requirements that verbs place on their arguments are not placed on the DP itself, but be placed rather in the presupposition slot so that it may be propagated to a local justification site and justified there – i.e. on the individual argument term furnished by the DP, if that is possible. Thus, for an intransitive verb like fall, we have the lexical entry in (21). Recall that π is the parameter that encodes type information. π*arg1fall : p specifies the type assignment for fall on its first argument place to be of type p or of physical type. This first argument place is ultimately filled by the object level variable or constant introduced by the translation of the DP in subject position but may be filled by other variables as well. In particular, the variable introduced by the DP Φ is one and y is one as well in (21). (22) shows the result of applying fall to the DP John; we assume that John introduces a constant of type person. (21) λΦλπ Φ(π * arg1fall : p) (λyλπ' fall(y, π')). (22) λπ λPλπ'' P(j, π'' * person) (π * arg1fall : p) (λy : pλπ' fall(y, π')).

(22) reduces to: (23) λπ λyλπ' fall(y, π') (j, π * arg1fall : p * person)

and then to: (24) λπ fall(j, π * arg1fall : p * person).

. This applies the same trick used to handle transitive verbs taking generalised quantifier arguments – which makes the latter functions from DP denotations to VP denotations.


64 Nicholas Asher and Julie Hunter

We can now justify the presupposition of fall via binding if we assume that persons are physical objects or via accommodation if not. The final result is: (25) λπ fall(j, π * arg1fall : person).

The type of an intransitive verb or a VP, which in Montague Grammar has the type e ⇒ t, now has the more complex type, dp ⇒ (∏ ⇒ prop), or the functional type from general DP types to propositional types. We shall come back to propositional types in a minute. Transitive verbs look very similar. The presuppositions of the predicate must percolate to the internal (direct object) arguments as well as the external (subject) arguments. Let’s look at an example of a transitive coercion verb, enjoy. It has the following entry: (26) λΦλΨλπ Ψ(π * arg1enjoy : ag) {λx : λπ'' Φ(π'' * arg2enjoy : evt – ∈(hd(Φ), hd(Ψ))) (λyλπ' enjoy(x, y, π'))}.

Coercion predicates like enjoy introduce a complex type restriction – an absolute type presupposition that its second argument be an event (indicated by the type evt), and a type licensing a modification of the predicational environment. This modification, indicated by the polymorphic type ∈(hd(Φ), hd(Ψ)) allows a functor to introduce an eventuality of a type to be determined by the “head” type of the objectual variables or constants introduced by the DPs Φ and Ψ. Polymorphic types are functions from types to types; their output value changes depending on the types that are their arguments. A familiar example comes from considering adjectives like flat. Its type is naturally construed to be a function from types to types and whose value changes given the input type. This reflects the intuition that flat changes its meaning rather dramatically depending on the noun it combines with: (27)

a. b. c. d.

flat country flat curvature flat tire flat beer.

With these preliminaries in place, here is a derivation of the meaning for a standard coercion. Let’s look at an example where enjoy applies to a DP like Anna Karenina as in (28). For book we use the complex type book. Though artificial, this allows us to bypass the complexities of so-called dual aspect types that have generated considerable discussion (Pustejovsky 1995; Asher & Pustejovsky 2006; Asher 2011; Luo 2010). (28) George enjoyed Anna Karenina.

3.  Aspectual coercions in content composition

Constructing a logical form for the DP and applying it to the entry for enjoy gives us: (29) λΦλπ Φ(π * ag)λv λQ Q(π * arg2enjoy : evt – ∈(hd(Φ) * book) (ak)) (λy1λπ1(enjoy(v, y1, π1) ∧ ag(y1) = v(π1))).

Let us assume that polymorphic type presuppositions prefer a local justification, near the verb. Continuing the reduction and abbreviating our type constraints on x and y1, we get: (30) λΦλπ Φ(π * ag) (λv enjoy(v, ak, π * evt – ∈(hd(Φ), book)) ∧ ag(x) = y1(π * evt – ∈(hd(Φ), book))))).

The presuppositions in the nuclear scope of the quantifier cannot be satisfied as they stand. But this particular verb licenses a transformation and the introduction of a polymorphic type functor with a polymorphic type that serve to justify the basic type presupposition. The polymorphic type functor applies to the λ abstract in the consequent given by the verb, λy1λπ1(enjoy(v, y1, π1) ∧ ag(y1) = v(π1)). For type presuppositions, this is a general procedure for presupposition justification. The functor introduces a predicate related to the polymorphic type. For example if the polymorphic type maps a cigarette to an event of type smoke(agent, cigarette), then we need to use the predicate smoke(e, x, y). When the polymorphic type is underspecified and of the form ∈(α, β) we take the predicate ϕ∈(α, β)(e, x, y). The functor instantiated for this example looks like this: (31) λPλuλπ'' (∃z : ∈(evt, book - ct) ∃z1 : ag(P(π'') (z) ∧ ϕ∈(ag, book - ct)(z, z1, u, π''))).

Applying the functor on the designated λ term within (30) and using Substitution, Binding, Application, and Substitution, we get: (32) λΦλπ Φ(π * ag) [λv∃z∃z1 (enjoy(v, z, π) ∧ ag(z) = v ∧ ϕ∈(ag, book - ct)(z, z1, ak, π)))].

We can now integrate the subject into (32) and exploit the fact that ag is a function to get the finished result: (33) λπ∃y(y = g(π) ∧ ∃z : ag(enjoy(y, z, π) ∧ ag(z) = y ∧ ϕ∈(ag, book - ct)(z, y, ak, π))).

The type of functor in (31) suffices to handle all cases of event coercion with verbs whose type presuppositions are sensitive to the type of both the subject and object. We’ll call this the functor. Polymorphic types in event coercion describe a morphism from types of objects to polymorphic types of eventualities involving


66 Nicholas Asher and Julie Hunter

those objects, and the reflects that morphism from objects to eventualities in logical form. Why should this transfer principle and type shift from objects to eventualities be sound? The answer has to do with the presuppositions of the particular words that allow for this morphism, like e.g. the aspectual verbs and enjoy. Enjoying a thing, for instance, presupposes having interacted in some way with the thing, and that interaction is an event. The verb enjoy, however, doesn’t specify what that event is. The event could be just looking at the object as in enjoy the garden or perhaps some other activity. Similarly, one can’t finish an object unless one is involved in some activity with that object, whether it be creating it or engaging in some other activity towards it. That is why such transformations are lexically based; it is the lexical semantics of the words that license the coercion and that make the rules sound. Semantics gets us only so far. We now need to specify the underspecified type ϕ∈(human, book). The type specification logic allows us to write appropriate axioms for defeasibly specifying underspecified types. (α 5 human ∧ β 5 book) > ∈(α, β) = read(α, β) (α 5 author ∧ β 5 book) > ∈(α, β) = write(α, β) (α 5 goat ∧ β 5 book) > ∈(α, β) = eat(α, β)

One of the advantages of the higher-order system is that we can reflect on the type of entries that we want for various words. It has been standard practice since Davidson (1968/1969) to introduce additional eventuality arguments to predicates derived from verbs as we have done for the verbal predicates that characterise coerced eventualities. But one need not do this. Instead we could have characterised an event object more abstractly via a realisation function. This permits a much more precise characterisation of eventualities and requires the theory to be more precise about what requirements lexical meaning and composition make on the characterisation of various types of entities, in particular eventualities. We come back to this point in the next section.

5. Modality, aspect, and the verbal complex The interpretation of the verbal complex, which includes the verb and its arguments as well as the projections that include tense, modality, and aspect, is a rich area of study for linguists and has many philosophical implications. Many linguists, Vendler (1967), Dowty (1979), Smith (1991), inter alia, have observed that . The > connective is a weak conditional connective used to encode defeasible generalisations in Asher (2011).

3.  Aspectual coercions in content composition

certain meaning shifts occur on the type of object denoted by the verbal complex when aspect and tense are applied. Since Vendler’s work in the 1950s, it has been customary to distinguish between different types of denotations of verbs and their projections. Vendler and most of those following him (with the notable exception of Dowty) have talked of these denotations as types of eventualities, which include events of different types and states. Combined with Davidson (1968/1969)’s treatment of action sentences, this has led to the received view of verbal modification by various projections, according to which verbal modification involves predication of an eventuality introduced by the verb (and bound by tense). If the modification applies at a node above Tense, for example with an adverb like allegedly or probably, then the eventuality is no longer available as an argument and so such modifications are customarily treated as modifications of some sort of abstract entity like an intension. On the other hand, nominal modification is much more heterogeneous, depending on whether the modifier is intersective, subsective, or non-subsective. TCL treats verbal and nominal modification on a par. Some modifiers, though not all, take the VP as an argument, as in Montague Grammar. Temporal and locative modifiers seem to fall into this class. Such modifiers behave both syntactically and semantically like adjuncts. For instance, we expect to be able to add adjuncts ad libidem to the verbal complex and this seems to be true for temporal and spatial modifiers. We can have several temporal modifiers that simply narrow down the time at which the event described by the verbal complex took place. Locative modifiers work similarly. (34) a. On Monday Isabel talked for two hours in the afternoon between 2 and 4. b. In Paris John smoked a cigarette on the train in the last second class compartment in seat number 27.

The assumption that temporal modifiers take the VPs as arguments makes predictions in TCL. Predicates pass their type presuppositions onto their arguments in TCL, not the other way around. So TCL predicts that temporal adverbials can affect the type of the verbal complex, as seen in (35) below. The temporal modifiers can change the aspect of the verbal complex from an achievement or accomplishment in (35a, c) to an activity in (35b, d). (35)

a. b. c. d.

John wrote a letter in an hour. John wrote a letter for an hour. John kissed Mary at 10 in the morning. John kissed Mary for an hour.


68 Nicholas Asher and Julie Hunter

For an hour takes a VP like wrote a letter as an argument and imposes the type presupposition that the variable of complex and polymorphic type that it modifies must be of type activity. We now have a case of coercion since write(p, p) is a subtype of achievement. A version of licenses an accommodation of the activity type presupposition by inserting an iteration operator over the VP, yielding an appropriate interpretation of wrote a letter for an hour. TCL also predicts that temporal modifiers may lead to fine-grained shifts in meaning in the verbal complex. For example, consider: (36) a. She left her husband at the airport 5 minutes ago. b. She left her husband at the airport two years ago.

In (36b), we have a very different sense of leave than in (36a). (36a) interprets her husband in a physical object or location sense whereas (36b) interprets her husband in a more institutional sense; that is, (36b) means that the subject has left her marriage. Conditional type constraints in the type specification logic can model the effects of the temporal adverbials on the predication. The next step to figuring out how the construction and modification of the verbal complex works is to investigate the nature of the verbal complex itself. With Neo-Davidsonians, we could stipulate that the verb projects an event variable to the tense and aspect projections. However, we need not do this. Some sentences intuitively don’t denote anything like an eventuality. Whatever is bound by the tense projection of the verbal complex is not an event or a state, at least if we take states and events to have some sort of spatio-temporal location (and if we do not, it’s unclear why we should call such entities states or events in the first place). Consider (37) Two and two make four.

What (37) describes is not a state of the concrete physical world but rather a fact or a true proposition, a collection of possible worlds that contains the world of evaluation. TCL countenances fine-grained types not only for words but also, thanks to the use of polymorphic types, for more complex expressions, even clauses; and these types for the verbal complex need not all be eventualities. In fact, a more uniform approach is to take the type of the verbal complex when saturated with its syntactically given arguments to be a subtype of prop. In effect there is no need . It is important to note that such facts also have temporal modifications. (i) Two and two make four, and two and two will always make four. But we do not go into this here.

3.  Aspectual coercions in content composition

to introduce a Davidsonian event argument (and in fact neo-Davidsonian proposals are easily translated into the higher type framework); stative sentences, or rather the verbal complexes they express, simply have propositional denotations (subtypes of prop) that are true at times and worlds. When needed, as in nominalisation, we can isolate this spatio-temporal region via a realiser. This minimises event promiscuity in one’s ontology and also takes care of the nasty problem of eventuality projection across quantifiers and operators like negation or modals as well as eventualities of such timeless sentences as 2 + 2 = 4. Since verbs don’t introduce events, there isn’t any problem about projecting them through the logical structure of the asserted content. In TCL the verbal complex is a subtype of prop; the specific type is determined by the appropriate instance of the polymorphic type with its type parameters specified by the verb’s arguments. Thus, the type of an intransitive verb IV in (38) is a generalised function from subtypes of a DP type and a presuppositional context type to a subtype of prop. This is a refinement of the type in (38b) that a more Montagovian version of TCL would assign. (38) a dp → π → iv(hd(dp)) b. dp → π → t

A similar story holds for transitive and ditransitive verbs. The “tail” of the type or value of the verb type in (38a) may take different finegrained values for different values of its parameters; it is also subject to modification by operators that take the verbal complex in its scope. Temporal adverbials, tense, and aspect modify this proposition, for instance; perhaps other modifiers do too. Some of these modifiers may, if we wish, force the introduction of an eventuality or a fact that realises the propositional content.10 For simple action sentences, modification by, for example, manner adverbials produces a realising eventuality for the content given by the verbal complex that the manner adverbial then modifies. But if a verbal complex is modified by negation, then tense or the presence of a locating adverbial like that in (39) may introduce a realiser that is a fact. (39) No one danced at the party.

Similarly, if the type of the verbal complex contains no action verbs but expresses simply a relation between informational objects as in (37), temporal adverbs or . The type of the realiser is polymorphic upon the fine-grained type of its argument as well as on parameters of evaluation. Thus, realiser is another example of a polymorphic type. 10. For a discussion of facts versus eventualities, see Asher (1993).


70 Nicholas Asher and Julie Hunter

tense force a coercion that introduces a realiser of the content that must be a fact. This would predict that the temporal modification of (i) in Note 8 means something like it will always be true that 2 and 2 makes 4. Thus, most modifications involve a coercion from the verbal complex’s internal semantic value, which is a subtype of prop, to an event or fact realiser. This allows us to use TCL to predict that certain eventuality types may be derived from others. For instance, walk is an activity but walk to the store with a goal PP is an accomplishment. The type system predicts that accomplishments should consist of an activity together with a natural endpoint or telos (given by the goal PP). TCL supposes that goal PPs are in fact part of the verbal complex, attaching directly to the verb or verb plus direct and indirect object arguments, and so occupy a position prior to any coercion to a realiser. Thus, such goal PPs contribute to the overall type of the verbal complex, and as realisers are a function of this type, it is no surprise that walk has a realiser that is an activity but walk to the store has a realiser that is an accomplishment. 5.1

Coercions with progressive aspect revisited

With this sketch of TCL’s view of verbal modification, let us return to aspectual coercion in the Examples (2), one of which we repeat below. (2) b. John is being silly.

Because the types of the verbal complexes can be quite fine grained, we can distinguish between verbal complex types whose realisers are facts, those whose realisers are events (here we have in mind paradigmatic event sentences like John kissed Mary), and verbal complex types whose realisers are facts or states but have a close connection with eventualities. Included in this class are examples that have been described as statives that accept progressivisation – for instance, copular sentences that involve some stage-level predicate like is silly, is stupid, is an asshole, …. (40) illustrates that these have a natural link to certain activities: (40) a. John was ridiculous to insist on fighting that guy. b. John was stupid/insane/silly to give his money to that woman. c. John was an asshole in being so rude to that student.

This construction doesn’t work with other stative predications like (41) #John knew French to give that speech/in making that speech.

These constructions indicate that copular predications with stative adjectives form a particular subtype of prop. While these predications are usually classified

3.  Aspectual coercions in content composition

as statives by the usual tests of adverbial modification, they are special in that they have a very tight connection with activities of which they describe the result. This subtype is an argument to the progressive and then produces a particular kind of realiser after tense is applied. The progressive introduces a functor that describes some process that leads at least in the normal instances to the appropriate realiser of the proposition given by the verbal complex (Dowty 1979; Asher 1992, inter alia). While the progressive does not apply to what one might call “pure statives” like John knows French, the progressivisation of this special subclass of statives introduces an eventuality realiser of the verbal complex given by a VP produced from a copula with adjective complement. When it combines with an adjective, the copula passes the presuppositions of the predicate to its DP argument. We’ll assume John is silly has a perfective or completed aspect that introduces a realiser. This realiser is of type state because of the presence of the copula which affects the fine-grained type of the verbal complex. We give the aspectual operator’s contribution first and then the end result. Below  is the type of the adjectival VP and Φ as usual is a variable of type dp. (42) a. λλΦ λπ∃z : state realises(z, ∧ {Φ(Pre(arg1) (π)) ((π))}) b. λπ∃z : state(realises(z, ∧ {silly(j, π)}).

Now let us turn our attention to (2b). After constructing the logical form of the VP, we apply the progressive operator in Aspect. The progressive aspect also introduces a realiser but it must be an event type that is non-stative. So it demands a realiser that is an activity. At this point local justification is attempted by introducing a realising eventuality for the verbal complex. A coercion takes place when the aspectual information combines with the verbal complex, prior to Tense, but here the coercion is more complex because the verbal complex still requires that any realiser be stative (it is a type presupposition of the verbal complex itself). So we need Aspect together with the fine-grained type of the verbal complex, which reflects the copula + adjective construction, to license a polymorphic type of the form activity(σ, α) whose parameters are σ 5 state and the bearer of the state. The output or value of the polymorphic type isa type of activity or process involving an object of type α that results in a state of type σ. The associated functor for this polymorphic type is: (43) λPλeλxλπ∃s (ϕ(activity(hd(P), hd(x))(x, e, π) ∧ (result(e, s, π) ∧ P(π)(s)(x))).

We now use a version of event coercion to justify the progressive’s type presuppositions, and we get the following meaning for (2b): (44) λπ∃e : activity(e º now ∧ ∃s (ϕ(j, e) ∧ result(e, s) ∧ realizes(s, ∧ silly(j, π)))).



Nicholas Asher and Julie Hunter

In words this says that John is doing some activity whose result state is s and s includes the temporal span of some aspect of John in which he is silly. The assumption that there is no aspect in John is silly leads to essentially the same logical form, but this time we have a direct coercion to the result state interpretation from the propositional content ∧silly(j, π). These are the intuitively right truth conditions for such a sentence. Our discussion has shown how aspectual coercion falls within the TCL approach to coercion.11 This discussion also shows us what is behind the polymorphic types that we used for simple event coercion; they are introducers of event realisers for an underspecified verbal complex. 5.2

Aspectual coercion and modality revisited

Let us now turn to the interaction of modals and aspect for another illustration of meaning shifts. Once again we are interested in the contributions of aspect, but in order to understand how these contributions interact with the semantics of modals, we have to use a more expressive framework, that of TY2, in which world and time evaluation variables become explicit parameters in the logical form and, accordingly, the type system countenances atomic types for worlds and times. What, first, is the position of aspect with respect to modality? Which takes which as an argument? The answer to these questions seems to depend, as Hacquard (2009) and others have suggested, on which modality we are interested in. Consider first epistemic modals like might in English. Let’s assume that the verbal complexes in the first sentences of (45a, b) license an event realiser. Using the mechanisms that discourse linguists like Roberts (1989) have used to analyze modal subordination, we predict a substantive difference between (45a) and (45b), which is born out. (45) a. John might run the Marathon tomorrow. It would take him at most 3 hours. b. John might run the Marathon tomorrow. It will take him at most 3 hours.

(45b) is worse than (45a), which is a classic case of modal subordination with events. This difference also shows that the eventuality given by the verb is within the scope of the modal. But what about the might claims themselves? Do they

11. See de Swart (1998) and Bary (2009) for a more extensive discussion of uses of coercion to describe different uses of the passé simple and imparfait in discourse. As far as we can tell, all of these coercions are of a piece with the story for aspectual coercion that we have spelled out here.

3.  Aspectual coercions in content composition

describe or introduce some sort of eventuality? If we suppose that a state is introduced by the epistemic modal, we should be able to temporally modify it anaphorically. But that isn’t possible: (46) #John might run the Marathon. That will last for a couple of years. (where that should pick up the possibility of John’s running the Marathon)

When we turn to adverbial modification, we have another indication that modals don’t introduce states. For instance, for adverbials, which provide one test for statehood of (the realization of) the verbal complex, are infelicitous with epistemic modals. (47)

a. b. c. d.

John was sick for two weeks. John was sick at 2 pm. #John might finish his dissertation for two years. John might finish his dissertation at 2.

(47c) sounds really bad to us, but conceptually it should make sense. (47d) sounds fine but the adverbial modifies the VP under the scope of the modal, not a state introduced by the modal itself. The story about spatial and temporal modifiers that hold of states seems much the same when we look at epistemic modals: (48) a. John was sick at work. b. John might finish his dissertation at Jean Nicod.

The last example is fine but the spatial modifier doesn’t modify a state given by the epistemic modal but rather the event described by finish. The facts are quite different for ability modals. (The following examples are due to Vincent Homer p.c.) (49) a. Hier, Jean devait rendre son devoir demain, mais les choses ont changé : il doit maintenant rendre son devoir aujourd’hui. b. Pendant des semaines, Jean a dû rendre son devoir demain, mais les choses ont changé : il doit rendre son devoir aujourd’hui. c. Il n’y a qu’en France que les gens peuvent aller adopter un enfant au Mali, c’est interdit partout ailleurs.

Davidson’s theory regarding the introduction of eventualities to account for adverbial modification doesn’t hold up for epistemic modals. The fact that it’s also difficult to pick up these eventualities anaphorically suggests that perhaps they aren’t there. It looks then as though epistemic modals take very wide scope; all temporal and aspectual modification takes place within the scope of the modality. Complicating this picture, however, is the observation that epistemic possibilities can shift with time, and thus are in some sense temporally located:



Nicholas Asher and Julie Hunter

(50) Two years ago, we might have taken that option, but not now. (51) Suddenly might we not need Google for much of our web browsing?  (Benjamin Cohen on Technology, April 22, 2010) (52) Kendrick Meek suddenly might have a shot.  (Atlantic Wire, May 11, 2010) (53) And in his palm he might hold this flower, examining the golden dainty cup, and in him suddenly might come a sweetness keen as pain.  (Carson McCullers, Ballad of the Sad Cafe)

These examples show that epistemic modals can get situated in space or time, but not by VP-adjoining adverbs, only IP adverbs. Also they are difficult to modify with ordinary tense in English. This shows, we think, that like Homer we can take epistemic modalities to have wide scope over tense and aspect, and over some higher projection of VP, and that the realiser of a modal statement is a fact. Temporal and spatial modifiers of epistemic modals follow the same analysis as the temporal and spatial modifications of other facts: they contribute parameters of realisation to facts. In contrast, the data show that spatial and temporal VP modifiers can clearly modify an ability modal claim. So this would suggest that ability modals are much closer to the root verb position and so would fall within the scope of tense and aspect. Aspect is traditionally understood to bind the event variable introduced by a verb phrase and Tense is supposed to locate the eventuality in time. But perfective aspect in many languages takes on an evidential function, which has not to do with events but with propositions (Faller 2006). We provide a framework in which this is natural. Aspect can bind a parameter in the modality; it can also bind an eventuality or a fact. It is the binding of a parameter in the modality that gets us the actuality entailment. Note that in TY2, temporal modifiers automatically attach via a Davidson-like rule to the time parameter, thus not requiring the introduction of any eventualities on that score. Perhaps one might also countenance a space-time parameter directly in TY2 to account for spatial modifiers as well. Given the type of worlds s and the type of times ti as basic types, we also have the type s → s, which is the type of a modal transition. This allows us to rewrite basic possibility and necessity modalities as: λwLϕ(w) 7 λ < : s " s λw ϕ(< (w)) λwKϕ(w) 7 λ < : s " s λw ϕ(< (w)) ∧ ¬∃