Individuals in Time

Linguistik Aktuell/Linguistics Today
Linguistik Aktuell/Linguistics Today (LA) provides a platform for original monograph studies into synchronic and diachronic linguistics. Studies in LA confront empirical and theoretical problems as these are currently discussed in syntax, semantics, morphology, phonology, and systematic pragmatics with the aim to establish robust empirical generalizations within a universalistic perspective.

Series Editors Werner Abraham
University of Vienna

Elly van Gelderen
Arizona State University

Advisory Editorial Board Cedric Boeckx
Harvard University

Ian Roberts
Cambridge University

Guglielmo Cinque
University of Venice

Ken Safir
Rutgers University, New Brunswick NJ

Günther Grewendorf
J.W. Goethe-University, Frankfurt

Lisa deMena Travis
McGill University

Liliane Haegeman
University of Lille, France

Sten Vikner
University of Aarhus

Hubert Haider
University of Salzburg

C. Jan-Wouter Zwart
University of Groningen

Christer Platzack
University of Lund

Volume 94 Individuals in Time: Tense, aspect and the individual/stage distinction by María J. Arche

Individuals in Time
Tense, aspect and the individual/stage distinction

María J. Arche

John Benjamins Publishing Company
Amsterdam / Philadelphia

8

TM

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

Library of Congress Cataloging-in-Publication Data María J. Arche Individuals in time : tense, aspect and the individual/stage distinction / María J. Arche. p. cm. (Linguistik Aktuell/Linguistics Today, issn 0166–0829 ; v. 94) Includes bibliographical references and indexes. 1. Grammar, Comparative and general--Verb phrase. 2. Grammar, Comparative and general--Aspect. 3. Grammar, Comparative and general-Tense. P281 .A675 2006 415/.6--dc22 isbn 90 272 3358 6 (Hb; alk. paper)

2006042929

© 2006 – 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

...........51 3.....................3 Summary of Section 3.....14 2.....39 3.........................38 Chapter 3 Event Classes and Individual-Level Predicates ......2 Agents in Event Structure .......................2...53 3................................................Table of Contents Acknowledgments............2 ..1 ................. Milsark (1974) and Carlson (1977)................... The Structure of Copular Constructions .................................................................................................................................. Chierchia (1995) ........................1.....................1...................1................15 2.....................................1....1............2....................2...................2 . by Tim Stowell ......................................................... Raposo and Uriagereka (1995).....2 Ser and Estar as an Aspectual Differentiation ....................3 Aspectual Differences among Individual-Level Predicates ........ Higginbotham and Ramchand (1996) ..ix Foreword................................ 1995) ..........12 2.........53 3.....................................33 2......................1..... Inner Aspect and Event Types ........ A Brief Stop at “Agentivity”....................2..................40 3...........................3 Summary of Section 3.....................2..................................................3 Summary of Section 2........1........1................. When There Is More Than One Copular Verb: Spanish Ser and Estar.........................5 2..................................................................................1.............1 .......5 2..........61 ........................11 2................3 IL Predicates as Inherently Generic Predicates................59 3..32 2..39 3..............1 Inner Aspect.................1 The IL/SL Distinction as a Semantic Distinction..............5 Summary of Section 2........1 A Cluster of Notions ............................5 2...2......... Kratzer (1988.......... Introduction to the Tests Discerning the IL/SL Status........................2 Event Types and Event Structure................................39 3...........2 The IL/SL Dichotomy as a Syntactic Distinction..............................8 2..........4 The IL/SL Contrast as a Categorical/Thetic Distinction.............2..........61 3.....................2.....................................................................................1 Chapter 2 Individual-Level Predicates .....4 Summary of the Chapter ..........................................................................................xi Chapter 1 Presentation of the Study ..................1 Distribution of Copular Verbs in Spanish ..3.....16 2..1 Proposals about the Individual-Level/Stage-Level Dichotomy ..........................25 2......

....4 ..6......76 3....133 4....3.........................108 4.............90 4.............3 The DP Subject of Relational MPs .........................................1.1 On the Interpretation of the Relational PP ..1 Lexicalist and Logico-Semantic Approaches .....2.....2 On the Optionality of the (Affected) Goal PP .............1 Lexical Semantic Classes of Adjectives ........7 .......96 4...107 4......95 4.....................................4 The Issue: Aspectual Alternation in IL Copular Clauses ..........................................3 Adjectives Predicated of an Implicit Event Argument ..........................84 4....1........................ States and Activities: A Grammatically Relevant Difference?.........3.....7.....105 4............6.........................106 4..7.....................................4..68 3.................3..4....................143 ..1 Aspectual Tests on IL Predicate ........5..4 Summary of Section 3...................5....................3 Summary of Section 4..........62 3.......................6 An Account Based on the Relational PP Complement . The Relational PP Complement.......81 Chapter 4 Aspectual Alternations in Individual-Level Predicates ..2 Syntactic Approaches ..108 4.......................2 The Preposition Introducing the “(Affected) Goal”: An Activity Inductor123 4.........................................................1 The Two-Copulas Hypothesis .......4 Summary of Section 4............. Summary of the Chapter ...................................6......4....................3 ....138 4............................................................................................................3 Relational Mental Properties.....................2 Be as a Copula Shifting a State into an Activity .....3.....80 3..3 .........................................................................2 Differences between Activities and States ............................69 3................................................ The Cruel-type as a Small Clause Integrant............................................4...5......7..8......................................................................................2 Summary of Section 4.......5 .............1 Similarities between Activities and States ............ 69 3...2 Summary of Section 4...2 Adjectival Predicates Showing an Activity-like Behavior ......................5............. Summary of the Chapter ...............85 4..........110 4................1 Prepositions as Aspect Encoders .137 4.........4 Summary of Section 4........................117 4...2 ......4...................................................................83 4........................................................................1........................................................2 Summary of Section 3................................135 4..117 4....................96 4.......................3 The Relational PP with Other APs ................1 Cruel-type Small Clauses Taken by Verbs Other Than the Copula ..................98 4...................143 4..................................................................................6........6 ...116 4........3......84 4..........86 4...72 3.......................................1 Previous Explanations of Nonstative Copular Clauses ............................3....3 Some Confusing Arguments in Defense of a State/Activity Distinction............................................vi Individuals in Time 3......5 Justifying the Approach ........2.....................91 4.............

. 187 5......................................................148 5..............222 ..148 5.........................2........3........2 The Internal Argument of Tense: The External Argument of Aspect...........220 6............................6..1 The Imperfect and Adjectival IL Predicates ........................208 6.............2..............218 6..1 Necessary Conditions for Lifetime Effects .. and Systematicity .........193 6................ Aspect as an Ordering Predicate .............. 1997) .....152 5............................. 176 5............163 5........205 6.....................3.....4 Summary of Section 6.............................1 Habitual Heterogeneous Predications.....................2 Introducing the Determining Role of Contextual Factors......................................1 When the Subject Is a QDP ...................................2 ................2......................................................... 180 5......2...........5 Adjectival Individual-Level Predicates and Viewpoint Aspect .......................5...........147 5.........................................................1 Tense and Aspect as Ordering Predicates ....................210 6..4 The Determination of the TT Content and Lifetime Effects ....................................................4........................................................ 1996) ......................... A Brief Summary of Aspect Notions ....... 179 5..........................3.................1 Temporal Interpretation as a Consequence of Argument Structure...215 6.................. 177 5..3.. 192 Chapter 6 Tense and Individual-Level Predicates .......3 The Progressive and Adjectival IL Predicates ....................................................2.........2......2 Perfective Homogeneous Predications ..........................157 5............ Summary of the Chapter ..................................2 Differentiating Temporal Extensions of IL Predicates ............ Kratzer (1988..................158 5........205 6........................4..4.............5 The Lifetime Reading in Compound Sentences ........................... Proportion.......2 Context Associated to Individuals ...............................................................1 Quantifying over Occasions ........................................................... 186 5.......173 5....................4........205 6...................................2 The Habitual Interpretation: Iteration......3............. 1995) ...2 Aspect as a Quantifier over Occasions ...................5.....................................................2 The Perfective and Adjectival IL Predicates ........3 Articulating the Account....................3 Brief Notes about the Complement of Mental Properties APs ..........................2 Nonpermanent IL Predicates........................1 Permanent IL Predicates ............4 Summary of Section 6.......202 6..........................Table of Contents vii Chapter 5 Outer Aspect and Individual-Level Predicates .................................. Stowell (1993.............................194 6................. 174 5............................................199 6.................4...........5..............212 6.....1 Tense..........1.........................1..............3 The Arising of Lifetime Effects ...........................................199 6..................................2...........3 The Content of the TT in the Arising of Lifetime Readings ...3 On the Relation between the TT and the Habitual Q<occ> ...................4 ...................................197 6.... Musan (1995.....................171 5...........213 6..................3 Inner and Outer Aspect ..

.............. Summary ........................4.....................................................232 6...................................................257 References ....................... The IL/SL Dichotomy Is Not a Permanent/Episodic Distinction ................. Summary of the Conclusions ..........7.....2 Relative Clauses ............................................... The IL/SL Dichotomy and the Ser/Estar Contrast .........6...................235 Chapter 7 Conclusions and Final Remarks ............242 7........................................241 7...................................................279 ........................5..................................................................... Outer Aspect Does Not Affect the IL/SL Distinction ................................2........................ Summary of the Chapter ...........261 Name Index ..........viii Individuals in Time 6...........239 7...249 7..222 6...........................1 Complement Clauses ...............................249 7...............................8................ Some Remaining Questions ...........................................................5...... The IL/SL Distinction Is Not a Matter of Inner Aspect…Completely ...........................................................1.................................................................275 Subject Index............................239 7.............................................................. Recasting the IL/SL Distinction .245 7.........254 7..............................................3.................5...............6................................

Although I have retained the overall structure and the spirit of the proposals defended there. Without them. who was reachable via mail and phone for discussion every time I needed her during the preparation of this book.Acknowledgments This book is a reviewed version of my doctoral dissertation. since my first visit to UCLA until the very last minute of submitting this monograph. Olga Fernández Soriano. Thanks for being always on the other side of the mail. Thanks for your friendship. which have helped me out to substantially improve the original. who helped me obtain financial support for this work. this work would not have been possible. Thanks to them I enjoyed Predoctoral Fellowships at the University Complutense of Madrid and at the University of California. Thanks a lot for your interest in my work and your detailed comments. . for her help in copy editing the last version. among which I must mention Hagit Borer and Daniel Büring. Many thanks also to Myriam Uribe-Etxebarría. This monograph is the result of several years of work at the Instituto Universitario Ortega y Gasset in Madrid and at the Department of Linguistics of the University of California. Remaining errors and loose ends are my own. Los Angeles. Professors María Luisa Hernaz. I want to express my respect for both of you here. Eugenio Bustos. Thanks a lot to Ignacio Bosque. I am especially in debt to my thesis supervisors. and Luis Sáez. Likewise. Thanks for your immense generosity and patience. Myriam Uribe-Etxebarría. Thanks a lot to Ivano Caponigro. Thanks a lot for your encouragement and understanding. punctually commenting on my every thought. Tim Stowell and Violeta Demonte. I have refined some of the ideas that had been merely sketched out in the earlier version. Los Angeles. I owe special thanks to Tim Stowell. I want to express my gratitude to the committee in charge of judging this work at the defense. Thanks to Elizabeth Laurençot. for their encouragement and support at every moment. which made the book look better. defended in the fall of 2004 at the Instituto Universitario Ortega y Gasset in Madrid. The people I have met along the way and the experiences I have had during this time have made it well worth the effort. and Tim Stowell. María Jesús Fernández Leborans. I am truly grateful for their comments and observations. and to Juan Romero for having encouraged me to try America. During various stages of this work I benefited from discussion with many people. for encouraging me to submit the proposal of this book. Many thanks to Olga Fernández Soriano for her advice and support.

You know I miss you all so badly. Mer. understanding. Adriana. You made my life incredibly enjoyable then. Selene. Rafa. For their generosity. encouragement. Eva. To my husband. To our two children. who were born in the process of this work. Javi. and Cristina. Thanks a lot to my parents. Ignacio and Javier. Amàlia. family. Jelena. for having taught me not to give up. Last. and continue to do so now. Stefano. thank you for being such good babies and giving me the time I needed to work. and Ed. but not least. Luca. Thank you for having made this possible.A. Shaee.x Individuals in Time Thanks from the depths of my heart to my friends Ana. this book is dedicated to them. Felipe. and to my “L. I want to express my gratitude to my family.” Heriberto. . and empathy at every moment. Ivano. thank you for your endless support.

depending on the surrounding context. The SL/IL distinction has been the focus of an important body of work in the literature on formal semantic and syntactic theory over the past 30 years. with SL interpretations arising in specific cases through the successive application of syntactic and semantic merger. this claim is contradicted by many direct empirical observations. because they alone among syntactic categories exhibit a robust internal division defined by the SL/IL divide. some are (usually) SL. Maria Arche provides an overview of the literature on this issue and proposes a bold hypothesis that all adjectives are fundamentally IL in their lexical core. others are (usually) IL. but Arche systematically examines the evidence that seems to argue in favor of a lexical distinction between IL and SL adjectives. In this study. the subject of the IL predicate is the topic of the sentence. In contrast.Foreword This book provides a compelling account of the distinction between stage-level (SL) and individual-level (IL) predicates. with greater or lesser degrees of naturalness. and shows that it is ultimately . combining adjectives with various complements and superordinate aspectual and adverbial categories. they serve to report an event or situation. This intuition is often associated with a related intuition about the discourse function of sentences whose main predicates belong to either type. and the sentence serves to provide information about its referent(s). whereas SL predicates involve inessential or transient properties. have been described as thetic. and the status of the subject is confined to that of a participant in the reported event or situation. The picture that emerges is that every adjective begins its life as an IL predicate and has the potential to become (part of) an SL predicate. Adjectival predicates have constituted the central domain of study in the literature. The SL/IL distinction is grounded in a core semantic intuition that IL predicates involve essential. It also provides a comprehensive critical survey of the major theoretical and descriptive accounts in the literature of the SL/IL distinction and the ser/estar alternation. Sentences containing SL predicates. whereas verb phrases and prepositional phrases are (almost) always SL. adjectives are a diverse crowd. At first glance. and a third type (possibly the most numerous of the three) can have either type of interpretation. permanent. Sentences containing IL predicates have been described as categorical. with an empirical focus on the correlation between the SL/IL distinction and the alternation between the two copular verbs ser and estar in Spanish. but no clear consensus has emerged on how it should be formalized. on the other hand. or even immutable properties. Noun phrases are (almost) always IL.

and zeroing in on the essential strengths and weaknesses of opposing theories. and provides a satisfying explanation of the circumstances under which temporal modification of IL predicates is allowed. and sometimes also with respect to the perfective versus imperfective form of the copula in certain languages. reviewing prior accounts of these in the literature. she brings to light important new data that bear on the theoretical debates. She guides the reader on a fascinating tour through a treacherous terrain of complex interactions among several semantic and syntactic phenomena. leading inevitably to this conclusion. Arche shows that most of the prior empirical generalizations in this domain are wrong. Her approach is grounded on a key empirical insight that she adopts from prior accounts of the ser/estar distinction—namely. . however. Systematically. having been based on an insufficiently narrow range of data. She uses this as a potent analytical tool. but Arche shows that many adjectives that alternate between stative IL usages and SL usages behave like activity predicates in their SL usages. and in the process provides an entirely novel syntactic account of how IL predicates are sometimes converted into SL predicates by combining with other syntactic elements. that the choice between ser and estar reflects a true correlation with the SL/IL distinction: predicates are always SL when they occur with estar. She focuses on several key empirical puzzles. she scrutinizes many of the phenomena that have been associated with the SL/IL distinction in the literature. chapter by chapter. including estar and its counterparts in other languages. In each case. The evidence that she uses to motivate this conclusion is diverse. especially those that have been cited in support of the widely (though not universally) accepted idea that IL predicates are necessarily permanent or immutable. Step by step. either providing key evidence in favor (or against) specific existing theories or leading her to develop novel theoretical accounts of her own.xii Individuals in Time untenable. as well as various other aspectual formatives and certain types of PP complements. Adjectival predicates are often assumed to be exclusively stative. Most of these involve interactions with various aspects of the aspectual and temporal properties of sentences in which SL and IL adjectival predicates occur. Nonverbal SL and IL predicates appear to differ from each other with respect to their compatibility with progressive and perfect aspect on the copular verb selecting them. she traces the source of the nonstative aktionsart to the intrinsic semantics of the PP complement of the adjective rather than to an inherent lexical ambiguity residing within the adjective (and its argument structure) itself. It has been claimed in the literature that only SL predicates (or SL usages of predicates) may occur with adjuncts of time or place. Arche shows that this too is wrong. a comprehensive big picture emerges.

Her account of the interaction between discourse structure and the interpretative properties of quantifiers. explains why they often fail to arise. involving paradigms of a sort that have been largely overlooked in previous accounts. The account of the SL/IL distinction that emerges is striking and compelling in its simplicity. the occurrence of the past tense on the copula triggers so-called lifetime effects associated with the referent(s) of the subject of an IL predicate. outer aspect. equally importantly. even if the overall picture that emerges of the interaction between this and other syntactic and semantic phenomena proves to be more complex than what previous researchers had envisaged. providing further support for her account of the SL/IL distinction in the process.Foreword xiii In certain types of contexts. Arche shows how these effects arise and. TIM STOWELL Professor and Chair Linguistics Department University of California. and inner aspect really feels like it is on the right track. tense. Los Angeles . Arche brings to light an enormous body of new data.

.

respectively. This book contributes to this debate in two ways. Likewise. we attribute particular aspectual characteristics to that predicate. On the other. “possessed” by the individual. corresponding to the two kinds of predication. In sum. These two goals are addressed by examining IL predicates in copular clauses. denominated by Carlson as stage-level (SL) predicates. Stative predicates are those that hold but do not take time or have an internal temporal structure. it explores the correlation between the syntactic structure of predicates and their temporal properties. When we say that a certain property is permanent. IL predicates have been traditionally described (Milsark 1974. (2) John is sick. the properties usually associated to the descriptions of IL and SL predicates are notions belonging to the temporal domain. it investigates whether the temporal properties of predicates play any role in the dichotomy traditionally known as individual-level (IL)/stage-level (SL). stages) of an individual. (1) John is blue-eyed. IL and SL. we mean that such a property is true at every temporal segment of an individual’s lifetime. ser and estar. Carlson 1977) as those predicates that apply to individuals and are. among many others). This is the case of Spanish. IL predicates contrast with those referring to spatial and temporal manifestations (i. Fernández Leborans 1999. in this work I use the copula alternation as an analytical diagnostic probe to . especially in those languages that have specific means to distinguish the two types of predication. which differentiates two copular verbs. Demonte 1999. respectively (Bosque 1993.e. when we conceive a particular predicate as stative. On the one hand.. such characterizations have led to associating IL predicates to the properties of stability and stativity. In linguistic research. Sentences (1) and (2) are examples of IL and SL predicates. The opposition IL/SL gains particular interest in copular clauses. in some sense.Chapter 1 Presentation of the Study The relation between the semantic properties of predicates and the syntactic structure where they appear has become one of the central issues of that area of formal linguistics known as syntax/semantics interface. Focusing on IL predicates in copular sentences in Spanish. Stability (or permanency) as well as stativity are two temporal concepts in nature.

remain unexplained when descriptions of IL predicates relying on stability and stativity are applied to them. the adjectival property in all of the examples is restricted to a concrete period of time located in the past (in her youth. as I noted earlier. (5) María fue Maria ser-PAST-PERF-3SG muy very guapa en su juventud pretty in-her-youth rubio de pequeño (6) Pedro era Pedro ser-PAST-IMPERF-3SG blond when-he-was-little (7) El periodista estaba siendo muy cruel con el entrevistado aquella tarde “The journalist was ser-ing very cruel to the interviewee that evening” In the first place. when he was little. (7) is an instance of a nonstative IL predicate. That is. and . respectively). he got tanned. When ser is involved (3). To examine the properties peculiar to ser-clauses. as well as the nonpermanency exhibited by a large number of IL cases. the speaker predicates the properties of the subject on a particular occasion. none of the predicates are understood as “permanent. outer aspect. In the cases with estar (4). stativity and permanency are notions belonging to the temporal realm. or he is in a good mood. In other words.2 Individuals in Time elucidate what lies behind the IL/SL distinction. such as (5)–(7). such alternations are shown in the following examples. that evening). I will take into consideration minimal pairs where only the copula differs. the sentence in (7) appears in the aspectual form of the progressive.” In the second place. dark-skinned. guapo/ moreno/ gracioso (3) Pablo es Pablo ser-PRES-3SG handsome/ dark-skinned/ funny “Pablo is handsome/dark-skinned/funny” guapo/ moreno/ gracioso (4) Pablo está Pablo estar-PRES-3SG handsome/ dark-skinned/ funny “Pablo looks handsome/got tanned/is being funny” Sentences with ser. In particular. I concentrate on those copular clauses where the predicate is an adjective. Since. funny person. The dynamic properties observed. the property is predicated of the individual as such: the speaker claims that the subject is a handsome. only combines with nonstative predicates. will be analyzed in this work. in Spanish. such as those in (5)–(7). which. I will study the temporal properties of IL predicates in the three commonly acknowledged domains—inner aspect. linked to external reasons (maybe because he is wearing a very nice suit.

After showing that the ser/estar dichotomy cannot be recast in terms of tense nor in terms of outer aspect (by alluding to the completion or perfection of the property) nor in terms of inner aspect (mereological properties). the dynamic properties of clauses where adjectives referring to mental properties appear (such as cruel. I study the set of IL predicates patterning with activities and argue that this behavior is rooted in the nature of the syntactic frame intro- . along the lines of Demirdache and Uribe-Etxebarría 2000. Ramchand 2003. The study of tense focuses on the interpretive phenomenon known as ‘Lifetime Effects’ (whereby the individual referred to by an IL predicate in past can be understood as ‘no longer alive’). I will propose a formal constitution of Aspect as a complex functional projection including an ordering predicate (which locates the so-called Topic Time [Klein 1994. This book is organized as follows. Outer aspect and tense are conceived as dyadic ordering predicates that take two time-denoting arguments (Stowell 1993. 1995] with respect to the totality of the Eventuality Time. Ritter & Rosen 2000). I argue that the IL/SL contrast should be rooted in the association to a particular circumstance (SL) or its lack thereof (IL). The discussion of inner aspect is framed in the theoretical debate regarding the effects of syntactic structure on the semantics of predicates. In chapter 4. giving the number of times a certain eventuality holds (along the lines of Verkuyl 1999). In particular. Specifically. In chapter 3. I will then discuss the theoretical and empirical consequences for the IL/ SL dichotomy. Tense is proposed to order the Topic Time with respect to a Reference Time and is argued to work uniformly with all kinds of predicates (SL and IL). my proposal will concern the role of prepositions in event structure. I present some reflections about habituality. I will show that prepositions can contribute an inner aspect shift from states to activities. mean) are proposed to correlate with the presence of the relational complement of these adjectives (cruel to Mary). this being understood in inner aspect terms (Borer 2005. Chapter 2 introduces the notion of IL predicate and presents the distribution of the two copular verbs in Spanish. In this respect. Focusing on copular clauses. the ideas defended in this monograph are sympathetic to the conception of syntactic structure as event structure. which is described as a viewpoint referring to a plural number of occasions.Presentation of the Study 3 tense. kind. Thus. dynamicity is proposed to correlate with the presence of the relational complement and argued to be rooted in the prepositional phrase. 2004) and a quantifier over occasions. I then apply such tests to different kinds of IL copular predicates and show that they can be divided in two groups: states and activities. In this vein. based on hypotheses that attribute aspectual content to prepositions (Hale 1984). 1996). I start the discussion of the temporal domain of inner aspect by first presenting the different event types acknowledged in the literature and the tests to diagnose them.

ordering and quantification over occasions). chapter 7 summarizes the conclusions of the study and offers some considerations about the appropriate description of the IL/SL dichotomy. and second. Chapter 6 centers on the temporal interpretation of IL predicates. the lifetime interpretation depends on the content of the Topic Time. I also discuss the relationship between inner and outer aspect. the influence of contextual factors on the arising of the lifetime reading is derived. Finally. the perfective. I will put forth two claims: first.. I apply all these conclusions to different IL predicates and argue that outer aspect forms do not make an IL predicate an SL one. the imperfect. In this respect.e. Since contextual factors influence the content of the Topic Time. Chapter 5 is devoted to the analysis of some of the outer aspect forms: specifically. focusing on Lifetime Effects. the lifetime reading is a salient reading only with lifetime IL properties.4 Individuals in Time duced by complement these adjectives can have. and the progressive. paying special attention to whether the perfective plays any role in making atelic eventualities telic. I show that inner aspect properties are independent from outer aspect ones (i. .

arguing for an aspectual distinction. PP. the presence of functional projections) and its relationship with the IL/SL distinction. Milsark (1974) and Carlson (1977). 2.1 Proposals about the Individual-Level/Stage-Level Dichotomy 2. since I am chiefly concerned with copular cases.3.1 The IL/SL Distinction as a Semantic Distinction. *There are several policemen insane . Section 2. *There are several policemen intelligent b. I concentrate on the most influential proposals about the contrast between individual-level (IL) and stagelevel (SL) predicates. according to most authors. There are several policemen available b. There are several policemen in the corner a. lexically distinguish the two types of predication. and introduce some of the concrete aspects for which this work makes alternative proposals. the predicates of (1) are allowed. Section 2. but those in (2) are excluded. For example. In section 2. a part of this chapter presents the differences between the two Spanish copulas (ser and estar). Since the data in the work are from Spanish. NP). I present the syntactic analysis I assume for copular cases—namely. This chapter is organized as follows. ser and estar. I will introduce the syntactic analysis that I will assume for copular sentences. Introduction of the Tests Discerning the IL/SL Status Milsark (1974) noted that there is a predicate restriction in the coda of presentational there-sentences. the one offered by Stowell (1978. according to which the copular verb selects for a small clause (SC) containing the nominal argument and the predicate (AP. I present other two aspects basic for my purposes in this work. In section 2. Likewise. IL and SL. After that.2 is devoted to the differences between the two Spanish copulas.1. which. specifically discussing arguments for semantic. 1981). I also introduce proposals by Heycock (1994) and Becker (2000). and pragmatic distinctions. indicating some of the points on which my proposal differs. (1) (2) a. I will first present the main descriptions of this notion found in the literature.4 summarizes the chapter.Chapter 2 Individual-Level Predicates In this chapter I introduce the concept of the individual-level predicate. syntactic.1. who discuss the constituency of the SCs that copular verbs take (in particular. I will critically revise those analyses.

j) & in (Los Angeles) (y)] For Carlson. The former are called individual-level (IL) predicates.1 contains a few remarks about this test. and the latter stage-level (SL) predicates. In Los Angeles is not predicated of the individual John but of a slice of him. (4) means that there is a stage in Los Angeles that realizes the individual John.” He defines properties as those facts about entities which are. if the predicate is IL. (5) (6) Dogs are in the backyard Dogs are mammals (existential or generic) (generic only) According to Carlson. In support of the difference between IL and SL. For example. j for John. possessed by the entity. Whereas.2. This distinction is linked to another previous distinction: the differentiation between two ontological categories. the DP subject is ambiguous between an existential and a generic reading.” Types of properties differ. . stages and individuals. SL predicates need an extra semantic operation (the realization function R). this distinction between IL and SL also draws the restriction between being able to appear as a complement of a perception verb1 or not. Elaborating on these insights. M stands for mammal. as in (5). the DP subject can only be understood as generic.6 Individuals in Time Milsark (1974:211) describes the contrast in the following terms. a mammal (M) is a property that applies to the individual John (j). (3) (4) John is a mammal M (j) John is in Los Angeles ∃y [R (y. is defined as “that whatever-it-is that ties a series of stages together to make them stages of the same thing. He argues that the predicates that are excluded from there-sentences can be characterized as “properties”. However. in what they are predicated of. Carlson argues that subject bare plurals in English are interpreted differently depending on the type of predicate. in some sense.” An individual. If the predicate is SL. Carlson (1977) proposed a distinction between those predicates that apply to individuals and those that apply to stages or happenings of individuals. in turn. the relation between in Los Angeles and John is distinct. However. IL predicates apply to their subject directly. Quoting directly from Carlson (1977:115). Carlson (1977) brings up other scenarios where each kind of predicate behaves differently. Compare these two sentences. 1 Section 2. the type of those allowed in them. as “states. a stage is defined as “a spatially and temporally bounded manifestation of something. whereas states are conditions whose removal does not cause any change in the essential qualities of the entity. then. according to Carlson.

(9) Mary is a mammal (10) Mary is a contestant on Do You Want to Be a Millionaire? Whereas a mammal can be qualified as a permanent set membership. such as NP predicates. (11) *John saw Mary a mammal (12) *John saw Mary a contestant on Do You Want to Be a Millionaire? In sum. although nominal predicates expressing set membership behave as IL (cf. This copula “be2”. they behave alike as complements of perception verbs. how long such set membership actually lasts. The one applying to individuals is an empty verb. the dichotomy “temporary/permanent” does not define the opposition IL/SL correctly. Both are excluded from this context. a contestant on “Do You Want to Be a Millionaire?” cannot.” Note also that. independently of the specifics of a situation. However. . 180). “be1”. seem to behave as SL (cf. Consider the sentences in (9) and (10). Carlson (1977) argues for the existence of two homophonous copular verbs. whether a predicate is permanent or not cannot be used as a test to discern the nature of the predicate. (5) and (7)). Set membership is not predicated of the situation an individual is in. independently of. I quote. However. The other. since IL predicates refer to the individual herself. as is the case of locative PPs. they are typically “temporary” predicates. whereas predicates expressing permanent properties. Carlson mentions predicates like dead.Individual-Level Predicates 7 (7) (8) John saw Mary in the backyard *John saw Mary a mammal Since SL predicates are predicated of temporal and spatial slices of an individual. as IL (cf. those authors who establish a strict correlation between permanency/temporariness and IL-hood/SL-hood have to resort to reinterpretation mechanisms to account for the totality of cases. and as I will emphasize in this work.2). As will be pointed out shortly (section 2. they are typically “permanent” properties. With respect to the copula. “has a translation that maps the adjectives that apply to stages to sets of individuals that have stages the adjective is true of” (p. is taken by those predicates referring to sets of stages. which clearly denotes a state of an individual. At first sight. In turn. as (11) and (12) show. but of an individual. (6) and (8)). predicates denoting temporary properties.1. as Carlson himself observes (1977:122). which proves they two test out as IL predicates. I argue. “be2”. although a significant number of IL predicates can be qualified as permanent and SL predicates can be qualified as temporary. they need not refer to permanent properties. (6) and (8)). but cannot be qualified as “temporary.

whereas SL predicates project this extra argument (which can be understood as a spatiotemporal argument).3 In Kratzer’s proposal. the event variable is existentially quantified).3.4 2 Davidson (1967) took the classical differentiation from Plato between verbs. as Diesing 1992 proposes. authors putting forth the latter line have proposed that there is only one copular English verb (be) and the difference between IL/SL resides in the (nominal) predicate itself (AP. such as Jäger (1999) and Becker (2000). He proposed that action sentences include an event variable as a primitive element in their logical semantics. e) & (in the bathroom. Whereas IL predicates are predicated of the individual directly. denoting nonactions but performing actions and proposed that events could be treated as individuals. which. e)). NP). based on Williams’s (1981) Argument Linking description.2 The IL/SL Dichotomy as a Syntactic Distinction. (In the default case. in the bathroom. PP. at midnight. When it is not projected. . such as Kratzer (1988. The Davidsonian argument is the external argument of the predicate. to mention just a few.2 The eventive variable is considered a primitive element in the logical semantics of events.1. e) & (at midnight. 1995) claims that IL and SL predicates differ with respect to what they are predicated of. with a knife. This gives rise to a difference in their argument structure which is the stem of the difference between the two types of predicates. it is the DP referring to the individual itself which functions as external argument. As will be shown in section 2. as in (i). denoting actions. 1995) and Diesing (1992) but disputed by others. Kratzer (1988. 3 4 I will add a few more words about the eventive variable in chapter 4. 2. SL predicates are predicated of something else: the eventive variable proposed by Davidson (1967). I will not discuss Kratzer’s perspective about where the NP subject should be assumed to generate: either in a specifier of the VP (along the lines of the VP internal subject hypothesis argued by Kitagawa 1986 and Koopman & Sportiche 1991) or in the specifier if the IP. can be modified and quantified over by adjuncts. Schematically.8 Individuals in Time This view has been adopted by some authors. e) & (slowly. 1995) In the spirit of Carlson (1977). and nouns. Kratzer’s proposal looks like (13). the toast. like nominal variables. Kratzer (1988. e) & (with a knife. from Davidson 1967. The trees in (13) just aim to capture Kratzer’s idea on the location of the external argument in the VP predicate. (∃e) (buttered (Jones. (i) Jones buttered the toast slowly. IL predicates do not.

Such an adverb unselectively binds all free variables it has in its scope.5 Let me now introduce another important aspect of Kratzer’s proposal. Heim 1982) (an animal in (20)). potential variables to be bound are those that indefinites introduce (following Kamp 1981. it feeds its son According to Kratzer. as the ungrammaticality of (15) and (17) shows. I argue that they are just an apparent test for SL-hood. In (18)–(20). she looks happy (19) *When(ever) Mary is a mammal. rooted in the argument structure. I discuss the semantics of when(ever)-clauses and. as presented.3. SL predicates (at your disposal) can. If the IL and SL distinction is.Individual-Level Predicates 9 (13) a. This is precisely the case of the IL predicate in (19). it is predicted to be a stable distinction. Other interesting contrasts are accounted for by this spatiotemporal variable: (18) When(ever) Mary is in Paris. when(ever)-clauses of the sort of (18)–(20) involve an implicit quantificational adverb. due to the prohibition against vacuous quantification (Chomsky 1982). (14) Mary is at your disposal in the office (15) *Mary is a mammal in the office (16) Mary will be at your disposal next week (17) *Mary will be a mammal next week Whereas IL predicates (a mammal) cannot be temporally or spatially modified. Kratzer is forced to readjust such a 5 In section 2. differing from Kratzer’s suggestion. This is explained by the possession or lack of the spatiotemporal argument. . Individual-level predicates IP 1 I′ 1 I VP 1 θ-subj VP Kratzer argues that the absence or presence of the eventive argument accounts for contrasts like the following. always. she feeds her son (20) When(ever) an animal is a mammal. the sentence results in ungrammaticality. and the spatiotemporal variable (18). Stage-level predicates IP 1 I′ 1 I VP 1 <e> VP b. However. When there is no variable to be bound. where the absence of indefinites and the absence of a spatiotemporal argument preclude the availability of a variable susceptible to binding.

where to add an argument at Logical Form constitutes a violation of the inclusiveness condition. Throughout this work I will argue that the opposition temporariness/permanency is not what draws the distinction between SL and IL predicates..e. context dependent and vague. although the eventive argument is treated as an argument of the verb. in the framework Kratzer inserts her work (the principles and parameters model of Chomsky 1981). If the IL/SL distinction resides in the argument structure. It is difficult to combine the idea that the difference between IL and SL is rooted in the argument structure with the fact that argument structure can be altered or complemented at the level of Logical Form. Chapter 6 deals with them in detail. it does not play any specified semantic role. This is a problem for her proposal. where a property. Another loose end in this approach is. every configuration has to be built up by the elements present from the initial lexical selection. is understood as “altered” (i. (21) Mary had blond hair (22) Mary was blond when she was little Kratzer (1995:154) argues that the spatiotemporal argument is present just at the level of Logical Form. . nothing can be added in the course of the derivation. in fact. which is considered an IL business. that. where the predicate is clearly restricted to a delimited period of time but indicates set membership. as Rosen (1999) observes. According to this restriction. Kratzer concludes that the IL/SL distinction is. Recall examples like (23) or (24). The same problem persists from the perspective of minimalism (Chomsky 1995 and subsequent work). “temporary”). All these complications arise from the erroneous parallelism between IL/SL and permanent/temporary. and it appears if some predicate takes it as an argument. this is a concomitant differentiation that accompanies a vast number of cases. (23) Juan fue concursante de ¿Quiere Ser Millonario? “Juan was a contestant on Do You Want to Be a Millionaire?” (24) Juan fue profesor hasta que lo contrató una editorial “Juan was a teacher until a publishing company hired him” There are other inadequacies in Kratzer’s proposal explicitly concerning the temporal interpretation of IL predicates also due to the connection she establishes between IL-hood and permanency. In the view of cases such as (21) or (22). argument structure does not vary depending on the interpretation process. At most. how is it that it can be modified contextually? In principle.10 Individuals in Time conclusion for the following reason. As I have pointed out. in principle permanent. Kratzer assumes that the opposition IL versus SL semantically corresponds to permanency versus temporariness.

he submits. Enç 1991b). According to him. In Chierchia’s hypothesis. (28) and (29) are excluded. Taking a neo-Davidsonian approach. is sufficient for ungrammaticality.1. Chierchia explains it as a subcase of the “quantification restriction” (Milsark 6 Chierchia assumes that this variable is projected in the semantic. 1995) hypothesis. the possession or lack of the Davidsonian argument is what draws the line between SL and IL predicates. representation. the Gen operator ranges over situations that are arbitrarily located—that is.. given that tallness (in (29)) is clearly perceivable and (29) is as ungrammatical as (28). (25) ??John is a linguist in his car (26) ??John is intelligent in France (27) ??John knows Latin in his office For similar reasons. .7 Chierchia (1995) argues that the hypothesis that IL predicates are inherent generics can account for the properties they typically display. Rather. (28) *I saw John intelligent (29) *I saw John tall Chierchia suggests that the oddity of (28) is not due to the fact that the property expressed by intelligent is not perceivable. The Davidsonian argument is present only in SL predicates.g. if these properties generally hold of the individual. He defines IL predicates as predicates that ascribe permanent or tendentially stable properties to their subject.3 IL Predicates as Inherently Generic Predicates. the Davidsonian argument has to be locally bound by a generic operator (Gen). Hence Gen is defined as a null quantificational adverb over situations. too. according to which a generic/habitual interpretation is only possible with eventive verbs. which is also the main characteristic of IL predicates. in IL predicates. but not in the syntactic. generics express tendentially stable properties. This.6 The difference between SL and IL predicates is that. Regarding the exclusion of IL predicates from the coda of there-sentences. 7 This proposal is the opposite to the one expressed by other authors (e. Chierchia (1995) argues that all predicates (including states) project a Davidsonian argument ranging over occasions/eventualities. in the line of authors such as Parsons (1990). it is due to the fact that. Chierchia (1995) In Kratzer’s (1988.Individual-Level Predicates 11 2. it makes no sense to link them to a specific situation in which one observes such a property. taken from Chierchia (1995:207). The judgments are his. over situations not restricted to a concrete space or time—which explains the oddity of the locative modifiers (in italics) in sentences (25)–(27).

we would expect that an overt generic operator could appear. it can be classified as belonging to both classes of predicates. which makes IL predicates excluded from existential codas for the same reason that any DP headed by a strong quantifier is excluded in such a context. Specifically. which I have already argued against in the two previous subsections. called “categorical judgments. in some pragmatic sense. are about the individual designated by the subject. Raposo and Uriagereka (1995). Clauses involving an IL predicate are about a prominent argument—a “category” in Kuroda’s (1972) terms. Mary is cultivated” 2. those referring to the event they introduce. therefore. it involves the same limitations affecting Kratzer’s proposal that the IL/SL nature of a predicate could be decided on the basis of its interpretation. As the following sentences show. if a generic operator were actually in the constitution of IL predicates. Chierchia argues that the Gen operator is a strong quantifier. they define IL predicates as those that. if we are referring to a chronic illness or an occasional one). This strict correlation leads this author to propose that when a predicate (like sick). SL predicates are. simply. Regarding Chierchia’s central proposal that IL predicates are inherent generics. (30) *There is [strong Q every computer] (31) There are [weak Q few computers] (32) *There are [students [strong Gen Q contestants] Chierchia assumes that the permanency or stability of the property is what makes IL predicates a natural class. can be understood as stable or transient (for example. IL clauses can be.12 Individuals in Time 1974). In turn. I consider that this conclusion makes the classification IL/SL lose its appeal since it does not have any predictive force. Juan is ignorant” (34) *Típicamente/*característicamente/*generalmente María es María ser-PRES-3SG typically/ characteristically/ generally muy culta very cultivated “Usually.1.4 The IL/SL Contrast as a Categorical/Thetic Distinction. that is not the case—another reason why I will not follow Chierchia’s approach on IL predicates. (33) *Típicamente/*característicamente/*generalmente Juan es ignorante typically/ characteristically/generally Juan ser-PRES-3SG ignorant “Usually. Besides. Higginbotham and Ramchand (1996) Raposo and Uriagereka (1995) argue that the IL/SL distinction is not rooted in the lexicon but is a matter of differences in information structure.” Clauses involving SL predicates .

” “contrast. hosting “topics. the DP is interpreted in the context of that particular event.” “emphasis. in the second part (no está genial). respectively. If geniality holds of the champ regardless of his being in a given event. scopes out to a particular functional position (F). at Logical Form. Thus. its transient character being derived. the DP is not linked to any particular event but “decontextualized” and an IL reading is derived. SL and IL clauses differ in defining the “topic of the predication”. these authors assume that the topic of a sentence is the argument that. Thus. geniality is predicated of an individual under the context of a particular event. genial el campeón . in the first part of (35) (el campeón es genial). Following Uriagereka (1994). However. F 1 el campeón T′ 1 T VP 5 F 1 genial T′ 1 T VP 5 b. genial. they correspond to “thetic judgments” in Kuroda’s classification. Their proposal would look roughly like (36). (36) a. The example they use to put forth their point is that reproduced as (35). the event scopes over the DP. they differ in what the sentence is about. geniality is predicated of a decontextualized champ. When it is the DP. contextualized in a concrete event. in their view. Nevertheless.Individual-Level Predicates 13 report an event. That is.” “focus. which is conceived as a scope-taking process. and an SL reading is borne out. pero no está genial (35) El campeón es the champ ser-PRES-3SG genial but not estar-PRES-3SG genial “The champ is genial (in general) but is not being genial (right now)” These authors note that (35) conjoins two statements in such a way that a contradiction should ensue. Raposo and Uriagereka (1995) assume that the topic of a sentence is established via a process of topicalization. the event is understood in relation to the DP. IL and SL predicates are differentiated by virtue of what gets such an F position: either the individual denoted by the DP subject or the event. Raposo and Uriagereka account for this fact by arguing that. as they claim (and as shown in section 2. When. (35) is not contradictory in Spanish.” and point of view in general. by contrast. then geniality will hold of him at all events he participates in.2).

Finally. the semantic contribution of each copula cannot be ignored. 1995) argue that SL predicates are more complex predicates since they involve something extra: the realization function for Carlson. Higginbotham and Ramchand associate this situational variable with the SL constructions. He proposes that what distinguishes IL from SL predicates is that the latter involve an obligatory generic operator.1 In this section I reviewed the most influential descriptions of IL predicates and their differentiation from SL predicates. Very similarly to Raposo and Uriagereka (1995). these authors consider that SL predicates correspond to predications where an independent external situational variable. according to which all predicates involve a Davidsonian argument. Chierchia (1995). On my view. what are usually called “SL predicates” are cases of “thetic” sentences.” the scoping analysis does not add anything we did not have beforehand (although it may be compatible with it). considers that all predicates involve such a Davidsonian argument. The strong correlation Kratzer establishes between IL-hood and permanency leads her to assume that the spatiotemporal argument can be present or not depending on the temporary or permanent interpretation of the predicate. Higginbotham and Ramchand (1996) also appeal to Kuroda’s (1972) analysis to account for the difference between SL and IL predicates. is the subject of predication. for these authors. Raposo and Uriagereka (1995) and Higginbotham and Ramchand (1996) argue that the distinction between both types of predicates is a matter of information structure. which makes the predicate apply to the subject in all circumstances—that is. Cases where an external situation is the topic of the clause . Carlson (1977) and Kratzer (1988. however.14 Individuals in Time I believe that having estar instead of ser in (35) is not a small detail. In sum. at Logical Form. when interpretation is decided. to look like an inalterable property. which is not found in the argument structure of the predicates per se. that IL predicates are necessarily permanent predicates.5 Summary of Section 2. I have argued that this is an undesirable outcome since it supposes that syntax works in a way contrary to what is commonly upheld in generative grammar. and the Davidsonian argument for Kratzer.1. the theoretical framework of Kratzer’s work. If we end up concluding that estar itself denotes “linking to a particular situation. I have discarded those aspects of Kratzer’s and Chierchia’s approaches based on an assumption I consider to be incorrect—namely.3. Such an assumption leads Kratzer to suppose that the spatiotemporal argument can appear in the last step of the syntactic derivation. Framed in the neo-Davidsonian approach introduced in section 2. 2.1.

the postcopular NP (the predicate) alternates in case (nominative or instrumental) depending on whether the property is permanent (40) or nonpermanent or noninherent (41). I expand the discussion regarding what seems to be an adequate description for the IL/SL dichotomy. see also Rapoport 1987 and Rothstein 1995. found across most of western Africa. In the past-tense cases. Matushansky 2000). hinging on whether the predicate expresses an inherent or definitional property of the subject (38) or not (39) (Greenberg 1994). 9 For further discussion on the overtness of Hebrew copula. the copula is null in the present tense but overt in the past tense (Kondrashova 1995.Individual-Level Predicates 15 are instances of SL predications. In the next section. where the semantics of the two Spanish copulas are introduced. When the predicate is an NP → copula “don” c. Spanish and Portuguese have two lexically different copular verbs. (37) a. those clauses where the topic is an individual are instances of IL predications. 2. In turn. A large part of the population uses Bambara as its mother tongue. The Hebrew copula alternates between overt or null. I take up this issue also in chapter 7.2 When There Is More than One Copular Verb: Spanish Ser and Estar A good number of languages show copular alternations. When the predicate is a PP → copula “bè” (38) Ha-kli ha-ze *(i) patis he tool the this (COP)MASC-3SG hammer “This tool is a hammer” me’od ‘ayef (39) Dani (*hu) Dani (COP)MASC-3SG very tired “Dani is very tired today” byl durak (40) Oleg Oleg-NOM was fool-NOM “Oleg was a fool” ha-yom today 8 One of the closely related languages collectively called Manding. Bambara is one of the officially declared languages of Mali. I have introduced the idea that equating permanency of a property and ILhood is inaccurate. . such as Bambara (Koopman 1992).8 distinguish three types of copular verbs depending on the category the predicate belongs to (37). 1996. and as secondary language it is employed to communicate nationwide. When the predicate is an AP → copula “ka” b. ser and estar. Other languages.9 In Russian.

1.2.” as cited previously (see section 2.1. as . How can the distribution of the two copular verbs be described? As traditionally observed. Roldán 1974) describe the distribution of the two copulas ser and estar on the basis of the Aristotelian dichotomy “essence” versus “accident.11 which.12 10 Note that these descriptions fall in line with the ones given by Milsark (1974) for “properties” and “states. The classic counterexample in this respect is estar muerto ‘be dead’.1 Distribution of Copular Verbs in Spanish Some traditional grammarians (Bello 1847. despite being ungrammatical with ser. and accidental properties as temporary. obviously designates a nontemporary property. if the subject denotes an event (44). insufficient to characterize the copular alternation. Este animal es/*está that animal ser/estar-PRES-3SG a goat “That animal is a goat” professor c. or ser.16 Individuals in Time (41) Oleg byl durakom Oleg-NOM was fool-INSTR “Oleg was a fool (he behaved like a fool)” The question of which factors determine copular alternations is one of the most debated issues in grammar tradition. 12 Although I will not work out this difference here. other authors observed that the distinction between permanent and temporary was. when the predicate is an NP. 2. if not a mistake. Ese artefacto es/*está that artifact ser/estar-PRES-3SG a computer from 1960 “That artifact is a computer from 1960” una cabra b. However. whereas estar is used to refer to those properties considered accidental. if. In the case of Spanish. Pedro es/*está Pedro ser/estar-PRES-3SG teacher “Pedro is a teacher” When the predicate is a PP. I would like to note Demonte’s (1979) observation that this distinction poses an asymmetry in the account of copula distribution.” The copula ser is used to refer to properties that are considered essential of the individual. there is no general agreement in the literature on how the copula alternation should be characterized.1).1. the copula used can be estar. un ordenador de 1960 (42) a. which led to the claim that ser corresponds to permanent properties and estar to temporary ones. 11 See also comments by Milsark (1974) and Carlson (1977) about the same predicate. discussed in section 2. the copular verb must be ser (42). if the subject is a physical entity (43).10 Essential properties are understood as permanent.

such as those referring to origin (46). we can distinguish a group of qualifying adjectives (45) and others that. Although this is discussed more extensively in chapter 7. In the set that only combines with ser. Londres está/*es London estar/ser-PRES-3SG in the United Kingdom “London is in the United Kingdom” en casa de Cristina (44) a. rather than qualify. La conferencia es/*está the conference ser/estar-PRES-3SG in Canada “The conference is in Canada” en Madrid c. La boda es/*está the wedding ser/estar-PRES-3SG in Madrid “The wedding is in Madrid” When the predicate is an AP. it is the nature of the subject. La fiesta es/*está the party ser/estar-PRES-3SG at Cristina’s place “The party is at Cristina’s place’ en Canadá b. (Note that the corresponding English glosses make no distinction regarding the copular verb). Juan está/*es Juan estar/ser-PRES-3SG in Brazil “Juan is in Brazil” en la mesa b. finally. as a first approach. . Juan es/*está Juan ser/estar-PRES-3SG incapable to hurt “Juan is incapable to hurt” reported. El libro está/*es the book estar/ser-PRES-3SG on the table “The book is on the table” en el Reino Unido c. a group that can combine with both. instead of the predicate which seems to count. For details. and. Juan es/*está Juan ser/estar-PRES-3SG very wise “Juan is very wise” incapaz de hacer daño c. let us say that there are some adjectives that only combine with ser. muy ignorante (45) a. classify. the situation becomes more complex. see Demonte 1979. Juan es/*está Juan ser/estar-PRES-3SG very ignorant “Juan is very ignorant” muy sabio b.Individual-Level Predicates 17 en Brasil (43) a. others that just combine with estar.

and those in (51) combine with both copular verbs. For more details about classifying adjectives. When the adjective is referred to the place where one is born. Are origin adjectives qualifying or classifying? I will treat them as classifying because they behave as canonical classifying adjectives in the other three tests that Schmidt and Bache motivate.13. and (c) impossible participation in binary correlations (49). and Demonte 1999.6). also. the copular clause sounds degraded: (i) the Cuban guy → the guy is Cuban (ii) the Cuban fight against the embargo → ??the fight is Cuban The reading where African is understood as a particular way of behaving is not relevant here. one of the properties that Schmidt (1972) and Bache (1978) provide to distinguish classifying adjectives from qualifying ones is. that classifying adjectives are excluded as copulative predicates (The presidential trip vs. Bosque and Picallo 1996. However. These tests are (a) impossible participation in comparatives (47). Further discussion is presented in chapter 7 (see section 7. 14 Note. (b) impossible modification by a degree adverb (48). precisely. that their acceptability as predicates in copular constructions seems to be confined to those cases where the subject is a physical entity. 15 13 .18 Individuals in Time alemana (46) Jelena es/*está Jelena ser/estar-PRES-3SG German “Jelena is German” Origin adjectives do not evaluate the DP they refer to. see Bosque 1993. it is not gradable.14 (47) *Este chico es más africano que el otro15 “This guy is more African than the other one” (48) *Este chico es muy africano “This guy is very African” (49) *Africano/inafricano African/un-African The adjectives in (50) combine just with estar. *The trip was presidential). If the subject is a resultative nominal.

ser cases can be described as cases where the adjectives are predicated of the subject as an individual. Pablo es/está gracioso Pablo ser/estar-PRES-3SG funny “Pablo is being funny” What is the difference between a copular clause with ser and another with estar? One sound way to approach such a question is to focus on cases where we find minimal pairs. In the estar examples. Pablo may be saying funny things that evening because he is in a very good mood. or dark-skinned person. which allow either copular verb. All the estar cases are perfectly compatible with Pablo being unkind and bitter as a person. or light-skinned. This can be proved by their compatibility with the following assertions. Let me focus on examples such as those in (51). Pablo es/está Pablo ser/estar-PRES-3SG handsome “Pablo is/looks handsome” b. the speaker claims that Pablo is a funny. Thus. estar sentences can be described as cases where the property is not only predicated of the individual. but of the individual and an occasion. or unattractive. Pablo está/*es Pablo estar/ser-PRES-3SG barefoot “Pablo is barefoot” guapo (51) a.Individual-Level Predicates 19 contento (50) a. and Pablo may look dark-skinned because he got tanned. Pablo es/está moreno Pablo ser/estar-PRES-3SG dark-skinned “Pablo is dark-skinned/is tanned” c. Pablo está/*es Pablo estar/ser-PRES-3SG naked “Pablo is naked” descalzo c. . Pablo may look handsome because he is wearing a nice suit. Pablo está/*es Pablo estar/ser-PRES-3SG glad “Pablo is glad” desnudo b. and the native intuitions about this are very clear and sharp. which do not lead to contradiction in any sense. which may happen very rarely. When the copula ser is involved. handsome. the speaker predicates the property of the subject in a particular occasion. In turn. These yield markedly different interpretations. depending on which copular verb is used.

pero está muy guapo (53) Pablo no es Pablo not ser-PRES-3SG handsome but estar-PRES-3SG very handsome “Pablo is not handsome. all of the inflected versions in (55)–(58) that we can construe as paraphrases have to be done by using estar. Consider the next group of examples. pero está muy gracioso (52) Pablo no es Pablo not ser-PRES-3SG at-all funny but estar-PRES-3SG very funny “Pablo is not funny but he is being funny” guapo. I would like to bring up some discussion about the contribution of perception verbs in distinguishing an IL predicate from an SL one (a test used by Carlson 1977. (59) Noté que Juan estaba/#era muy guapo noted that Juan estar/(#ser)-PRET-IMPF-3SG very handsome “I noted that Juan was looking/was very handsome” . not ser— that is. where a perception verb (note) has a SC (NP + AP) as a complement. whenever the copular verb is ser. I am dealing with an IL predicate.1. throughout this work. Consider (59)–(62). see section 2. pero está moreno (54) Pablo es Pablo ser-PRES-3SG very light-skinned but estar-PRES-3SG dark-skinned “Pablo is very light-skinned. With the copular distinction in Spanish in mind. the copula estar will be considered a lexical sign of an SL predicate. Correspondingly. Interestingly.1). the copula designing SL-hood. as we already know from (51). but he looks very handsome” muy pálido. (55) Noté a Juan muy guapo I-noted Juan very handsome (56) Noté a Juan muy pálido I-noted Juan very pale (57) Noté a Juan muy moreno I-noted Juan very dark-skinned (58) Noté a Juan muy gracioso I-noted Juan very funny All of the adjectives in (55)–(58) can combine with either ser or estar. Thus. but he is tanned” The terms in which the contrast between ser and estar has been described fall in line with the IL/SL distinction introduced in section 2. I will consider that.20 Individuals in Time nada gracioso.1. where the pound symbol indicates when the paraphrases are not appropriate.

(63) Noté a Juan cansado I-noted Juan tired (64) Juan está/*es cansado Juan estar/ser-PRES-3SG tired (65) *Noté a Juan ignorante I-noted Juan ignorant (66) Juan es/*está ignorante Juan ser/*estar-PRES-3SG ignorant However. since they combine with estar. desnudo and descalzo16 do not. which only combine with estar. since the participial forms corresponding to the verbs they are related to are desnudado and descalzado respectively. whereas those combining with ser are excluded in such a position. 16 . some adjectives. as in (65) and (66).17 Adjectives such as desnudo or descalzo are described as “cut-short adjectives” (Bosque 1990).Individual-Level Predicates 21 (60) Noté que Juan estaba/#era muy pálido noted that Juan estar/(#ser)-PRET-IMPF-3SG very pale “I noted that Juan was looking/was very pale” (61) Noté que Juan estaba/#era moreno noted that Juan estar/(#ser)-PRET-IMPF-3SG dark-skinned “I noted that Juan was tanned/was dark-skinned” (62) Noté que Juan estaba/(#era) muy gracioso noted that Juan estar/(#ser)-PRET-IMPF-3SG very funny “I noted that Juan was being/was funny” This is consistent with the fact that adjectives that combine with estar can appear as complements of a perception verb. Observe the following contrasts: Juan estaba/*era {desnudo/descalzo/alterado/preocupado/cansado} Juan estar/ser-PRET-IMPF-3SG {naked/agitated/worried/tired} “Juan was naked/agitated/worried/tired” (68) *Noté a Juan desnudo/descalzo I-noted Juan naked/barefoot (69) Noté a Juan {alterado/preocupado/cansado} I-noted Juan agitated/worried/tired (67) Although alterado/preocupado/cansado (strict participial adjectives) work as expected. as in (63) and (64). and would therefore be expected to be grammatical as complements of notar. are not so. Chapter 7 contains more discussion on these adjectives.

e. adjectives compatible with proportional degree modifiers are those gradable predicates with a closed structure scale (with minimal/maximal elements).. In turn. One existing difference between the adjectives of (68) and (69) is that only the latter can be modified by muy ‘very’: Noté a Juan muy {alterado/preocupado/cansado} noted Juan very {agitated/worried/tired} (ii) *Noté a Juan muy {desnudo/descalzo} noted Juan very {naked/barefoot} (i) However. it is not the case that any perception verb allows for any type of adjectives as its complement.e. interesting. correlates with an “absolute” (i.. (vii) and (viii)). completamente ‘completely’. argue that notar is sensitive to the type of scale structure of the adjective since it is sensitive to the possible subjective evaluation of the property and the context (in)dependency of its standard of comparison. inexpensive} We could. Notar seems sensitive to other properties of the predicates appearing in the small clause selected by it.) notes. whereas it allows open-scale adjectives. short. context dependent) standard of comparison: (v) partially/half/completely {empty. alterado ‘agitated’ sounds quite natural when combined with a proportional degree modifier. open. it rejects closed-scale ones. they argue. full. gradable adjectives incompatible with such modifiers involve an open scale structure (without minimal/maximal elements). this does not mean that predicates such as desnudo ‘naked’ or descalzo ‘barefoot’ are not gradable. closed} (vi) ??partially/half/completely {long. This and other issues concerning the semantics of the predicates allowed by perception verbs need further examination. The following sentences where they are used in comparatives and are modified by proportional degree modifiers such as parcialmente ‘partially’. (vii) ??{parcialmente/medio/completamente} cansado {partially/half/completely} tired (viii) ??{parcialmente/medio/completamente} preocupado worried {partially/half/completely} (ix) ?{parcialmente/medio/completamente} alterado {partially/half/completely} agitated 18 As Violeta Demonte (p. verbs such as percibir ‘perceive’ or advertir . which. then. However. For example.22 Individuals in Time Other perception verbs. although cansado ‘tired’ and preocupado ‘worried’ seem to behave as proper closed-scale adjectives (cf. non-context-dependent) standard of comparison. correspondingly.18 such as ver ‘see’ seem to discriminate IL versus SL at first sight.19 since the adjectives yielding grammatical sentences in contexts 17 The semantics of the adjectival predicates admitted with notar deserves a much deeper exploration than what I can offer here. or medio ‘half’ suggest that they can be considered as such: (iii) En esta escena el actor aparece más desnudo que en la anterior “In this scene the actor appears more naked than in the previous one” (iv) Iba {parcialmente/completamente/medio} desnuda por la calle “She was {partially/completely/half} naked along the street” According to Kennedy and McNally (2005).c. correlates with a “relative” (i. which.

moreno. rather than IL/SL-hood. physical. or both types of perception) and the role of modifiers like muy ‘very’. where a state such as know languages is excluded. Among the issues to be explored would be the specifics of the meaning of these verbs when combined with small clauses of the type under discussion (whether they refer to mental. subject pronouns need not be overt in Spanish. 19 With infinitive complements. Other adjectives. Consider the following contrast. but you stay’). or gracioso. bastante ‘quite’. the sense of the verb is that of sub- ‘notice’ are odd with predicates such as guapo. pálido. improve the sentence.Individual-Level Predicates 23 such as (70) and (71) are those that combine with estar (cf. (70) Te he visto desnudo “I have seen you naked” (71) *Te he visto ignorante “I have seen you ignorant” (72) Te he visto muy pálido → He visto que estabas/#eras muy pálido you have-seen very pale → have-seen that estar/ser-IMPF-PRET-2SG… “I have seen you looking/being very pale” However. and disambiguate between verbal forms (cantaba ‘sing-IMPF’ can refer to the first and the third person). emphasize (Quiero que vengas tú ‘I want you [and not other] to come’). pero tú te quedas ‘Your brother can come to the theater. the semantics of perception verbs deserves a deeper investigation than what I can offer here. Juan {guapo/pálido/moreno/gracioso} (i) *Percibí/advertí a I-perceived/noticed Juan {handsome/pale/tanned/funny} al contrincante {muy tranquilo/nervioso/apagado} (ii) Percibí I-perceived the opponent {very quiet/nervous/spiritless} As mentioned in footnote 16. they appear to contrast (Tu hermano puede venir al cine. In this regard. as in (ii). rather than ser (72). the verb see distinguishes between states and events. 20 .20 When they are overt. but an activity (paint on the desk) is allowed: idiomas (i) *Te he visto saber you have seen know-INF languages “I have seen you know languages” (ii) Te he visto pintar en el pupitre you have seen paint on the desk As is known. it is interesting to note the role that overt subjects (pronouns) play in the meaning of this perception verb in Spanish. the corresponding paraphrases contain estar. however. and so on that contribute to the acceptability of the sentence. 50 above) and. when the adjective is combinable with the two copulas. Roughly speaking. not all the results are so clear cut.

the past tense (ii) favors an SL interpretation. Whereas. Veo que estás muy guapa I-see that you-estar-PRES-3SG very pretty “I see that you look very pretty” b.22 (74) a. Yo creo/a mí me parece que eres muy guapa I think/to me it seems that are very pretty “I think/it-seems-to-me that you-are very pretty” c. (i) Mario te ve muy guapa Mario you sees very pretty “Mario sees you very pretty” (‘Mario thinks that you are very pretty’) (ii) Mario te vio muy guapa Mario you saw very pretty “Mario saw you very pretty” muy guapa (iii) A Mario le pareció que estabas to Mario it seemed that estar-PAST-IMPF-2SG very pretty “Mario thinks/thought you looked very pretty” 22 I owe such an interesting observation to Olga Fernández Soriano. you look very good in those pants)’ b. when subjects are overt. Te veo muy guapa (hoy. Yo lo veo muy ignorante I him see very ignorant “I think/for me he is very ignorant” 21 When the subject is a third-person DP. you look very good in those pants) ‘I see you very pretty (today. Yo te veo muy guapa (no sé por qué tienes tantos complejos) I you-CL see very pretty (not know why have so-many complexes) ‘I see you very pretty (I don’t know why you have so many complexes)’ Interestingly. . the present tense (i) favors an IL reading (the opinion about the beauty of the person is understood not to be tied to a particular occasion).21 Consistent with this fact. other factors can disambiguate between the two readings of ver ‘see’. with estar. te sientan muy bien esos pantalones) you-CL I-see very pretty (today. maybe something along the lines of Mario te vio muy guapa en la fiesta ‘Mario saw you very pretty at the party’ meaning (iii). those adjectives that go with ser are allowed (74c) . Consider the following contrast: (73) a.24 Individuals in Time jective perception and the predicate of the SC is understood as IL. whereas without the overt subject (73a) the sentence should be paraphrased by using the copula estar (74a). for example. with an overt subject the sentence is paraphrased by using ser (74b).

is already found in Gili Gaya 1961. .” As I understand Luján’s proposal. 2. She establishes a parallelism between estar and predicates like write a letter. they select estar. the two copulas differ in the nature of the temporal reference they make. or at least one of them is. the work is from 1945. in general terms. and Fernández Leborans (1999). whose beginning and end are assumed (75). Schmitt (1992). they must bear the feature “–[–perfective]. In turn. and the [±perfective] feature in the copula.Individual-Level Predicates 25 In sum. the results emerging from them are consistent with the distribution of the copulas in Spanish. Following Querido (1976). states where no beginning or end is assumed—they select ser.2. Ser is conceived as a predicate parallel to write or admire. but rather that it makes a different contribution to temporal reference. Luján (1981) claims that all adjectives are stative. (75) Perfective predicate A(x) at time tj 23 The cited date corresponds to the ninth edition. When they refer to perfective states. tests based on perception verbs should not be taken as totally safe auxiliary tests to distinguish between IL and SL predicates. Among those who have worked along these lines. Thus. both undelimited predicates. 2. Although. but can be either [+perfective] or [–perfective] (Luján 1981:175). it is not that the copula ser is atemporal. ser expresses that a predicate applies to an individual during a stretch of time with no beginning or end assumed (76). where a beginning or end is assumed.1 Luján (1981). Estar expresses that a predicate is true of an individual for a delimited time period. there are many other factors intervening in the semantics of these constructions. if they combine with both. as contento ‘glad’. basically in the sense of ‘durative’ versus ‘punctual’. When the predicates are used to refer to imperfective states—that is.2. Examples (75) and (76) are Luján’s (1981:177). How can we explain those cases where an adjective can combine with both copulas? Luján argues that. her account of the distribution of copular verbs (concretely in combination with adjectives) relies on two factors: the [±perfective] feature in the adjective.2.23 This author argued that the property distinguishing ser and estar is imperfective versus perfective. a delimited process.2 Ser and Estar as an Aspectual Differentiation The idea that the distinction between the two copulas ser and estar should be described in aspectual terms rather than as a permanent/transitory dichotomy. According to Luján. I will discuss work by Luján (1981).

2. which suggests. Predicates like be in the garden are not covered by Luján’s generalization. Schmitt provides several scenarios where estar and ser contrast in their possibility of appearance and argues that the explanation rests on the aspectual properties of estar.1]. APs (human. Estar corresponds to a result state of an accomplishment verb. PPs (when the subject is an event (party) or an immobile object) APs (tired. In contrast. scenarios incompatible with a result state are excluded for estar. female.2. . PPs (when the subject is a mobile and immobile object) Estar + Table 2. APs (human. Among the facts that Schmitt cites to demonstrate the internal structure of estar and the underspecification of ser are those enumerated in (a)–(d).2). old). female. round. since there is no reason to consider in the garden as a delimited predicate. That is. However.) Copula Ser + Portuguese NPs. old). its wider flexibility in combining with adjectives indicates that it does not involve internal temporal characterization. swim). the description of SL predicates as perfective (in the sense of “delimited”) does not capture the intuition about the whole class of SL predicates. happy).1. nor is it an event or a process. Schmitt (1992) argues that the distinction between ser and estar is aspectual in nature. ser has no inherent temporal structure. her reasoning is as follows. since the distribution of Portuguese copulas nearly parallels those of Spanish [see Table 2. Since estar refers to a result state. happy). Portuguese and Spanish copular verbs The concrete claim of Schmitt (1992) is that estar involves aspectual properties. whereas ser is underspecified with respect to aspect. 24 All these concepts will be adequately presented in chapter 3 (see section 3. In the spirit of Luján. Accomplishments are those predicates that refer to a process with a inherent endpoint (write a letter or build a house).2 Schmitt (1992). States are eventualities with no duration or endpoint (be tall). PPs (when the subject is a mobile object) Spanish NPs.1.) 2.26 Individuals in Time (76) Imperfective predicate A(x) at times tj… tj+k However. but. its underspecification in aspect. and processes or activities are those events with duration but without an inherent culminating point (walk.24 In a nutshell. PPs (when the subject is an event (party)) APs (tired. I take it that her conclusions about the topic can be applied to Spanish. for this author. (More on this in chapter 7. round. (Schmitt works on Portuguese. It is not a state. ser manifests a wider flexibility.

since it would be the aspectual head that licensed the aspectual form.Individual-Level Predicates 27 (a) Possibility of appearing in pseudo-clefts and equative constructions Only ser can appear in pseudo-clefts and equative constructions. if the reason for the grammaticality of (80) is the incorporation of ser into the . Schmitt continues. where the copula does not play any role in terms of selection and does not add any semantic aspectual content to the construction. estar predicates cannot. a house has been built but in John was building a house. With the progressive. I will make two brief remarks. kind. then. (80) Juan estaba siendo muy cruel con el entrevistador Juan was ser-PROG very cruel to the interviewer “Juan was being very cruel to the interviewer” To account for cases such as (80). In this respect. In sentences like John built a house. there is no result reading available. Second. However. el chico del que te hablé (77) Juan es/*está Juan ser/estar-PRES-3SG the guy about whom you I-told “Juan is the guy I told you about” es culto (78) Lo que Juan es/*está what Juan ser/estar-PRES-3SG is cultivated “What Juan is is cultivated” (b) Possibility of appearing in the progressive Whereas ser predicates can appear in the progressive. by virtue of the aspectual underspecification of ser. it is not surprising that just predicates of the kind like cruel are the ones that are grammatical in progressive contexts. the house has not been built yet. according to Schmitt. ser + some APs (such as cruel. it seems that the incorporation of ser into the aspectual head (progressive) is what licenses the progressive. which looks like a circular explanation. (79) María está *estando/siendo simpática Maria is estar/ser-PROG nice “Maria is being nice” Schmitt argues that this contrast is explained by the fact that the progressive cannot range over result states. First. mean. nice) can appear in the progressive. Since the progressive favors subjects that have some control over the predicate. Schmitt (1992) argues that ser incorporates into an aspectual head (the progressive in (80)) and becomes a predicate with internal temporal structure.

28 Individuals in Time aspectual head. (82) *Juan vio a María estar descalza Juan saw Maria estar-INF barefoot (83) *Juan vio a María haber construido una casa Juan saw Maria have-INF built a house Although Schmitt does not discuss this point in detail. In other words. (See section 2. (81) *Juan estaba siendo esquimal Juan was ser-PROG Eskimo “Juan was being Eskimo” One of the central goals of this book is to establish a link between the type of adjectives and the peculiar aspectual behavior just mentioned. it is not clear that tests along the lines of (82) and (83) show that estar designates a result state. Schmitt also notes that estar. they may just show that estar predicates behave as states. the fact (also noted by Schmitt) that not all adjectival predicates combine with the progressive (cf. 1995). (c) Inability of estar to appear as a perception verb complement In relation to its inability to appear in the progressive. Chapter 4 develops a proposal about the strict correlation between the type of APs and this syntactic behavior.1. However. is not allowed as a perception verb complement. the reason why estar and the perfect are excluded as complements of perception verbs may be because both behave as states. she seems to suggest that the fact that estar predicates behave like perfect forms as complements of perception verbs provides further evidence that estar predicates are result predicates. like all verbs in the perfect. (81)) remains unaccounted for.” (I will introduce these notions properly in chapter 3). as I already suggested. such as when(ever)-clauses: . I will argue that a set of APs (to be delimited) involves particular properties that make the constructions they are in pattern as “activities” rather than “states.) (d) Acceptability of ser in whenever-clauses Schmitt (1992) raises an important issue regarding the acceptability of ser in scenarios considered typical for SL predicates by Kratzer (1988.2. and (infinitive) states are not accepted as perception verb complements.

the situational variable linked to SL predicates is just one of the possible variables susceptible of “referential variability. See section 2. among others). I do not share this judgment in Spanish. the fact that it is allowed in when(ever)-contexts is completely unexpected under Kratzer’s hypothesis. however.1. if ser is the copula designating IL-hood. he regrets it right after” In effect. Some specific examples will help explicate this point. that different adjectival predicates need different conditions to license their appearance in when(ever)clauses. when(ever) can refer to predicates that can be iterated in time.25 Schmitt offers an explanation for the acceptability of cases like (84) based on two points. ha estado allí para escucharme “Whenever I have needed him. As the glosses indicate. That when(ever)-clauses select for properties that can be iterated in time seems clear. no matter what type of predicate is involved. and find (85) and (86) grammatical. I attended him” (86) Siempre que lo he necesitado. rather. as Schmitt seems to intimate. I also agree that this does not have to be equated with SL-hood.” Specifically. predicates in perfect aspect are excluded in when(ever)-clauses.” which is the crucial factor licensing distributivity (Beghelli 1995. That is. Furthermore. since the copula ser perfectly fits in such contexts. Consider first the contrast in (87) and (88). First. se arrepiente después (84) Siempre que Juan es whenever Juan ser-PRES-3SG cruel he-regrets right-after “Whenever Juan is cruel. On my view. from Kratzer’s perspective. I would not say that only some adjectival predicates are allowed in these contexts. whenever-clauses are allowed only in the presence of an eventive argument. (85) is a simple past-perfective form and (86) a compound perfect form.” According to her judgment. Recall that. yo le atendí whenever he come-PAST-PERF-3SG to see-me I him attended “Whenever he came to see me. she proposes that when(ever)-clauses do not select for SL-hood but for “certain aspectual properties. (85) Siempre que vino a verme. he has been there to listen to me” Schmitt claims that when(ever) selects for predicates that can be distributed over time. Szabolcsi 1996.2. 25 . but. Beghelli & Stowell 1996. that is. she argues that when(ever)-clauses select for “some sort of durative aspect. which does not mean that it selects for predicates denoting properties of stages.Individual-Level Predicates 29 cruel. only with SL predicates.

30 Individuals in Time (87) Siempre que Juan es cruel. contrary to fact: (90) Siempre que Juan está descontento con su trabajo. . Chierchia (1992). If it finds a variable susceptible of referential variability. Although. according to Schmitt (1992). This is what licenses. expands on this account about iteration. Schmitt argues that ser is allowed in when(ever)-clauses by virtue of its underspecification with respect to aspect. in all his reincarnations whenever Juan ser-PRES-3SG Eskimo lleva una vida llena de dificultades y penurias he-bears a life full of difficulties and shortages “(In all his reincarnations).) As a last remark regarding when(ever)-clauses. kind. ser can be ruled out neither in when(ever)-scenarios nor in progressive ones. which Schmitt treats in 26 De Hoop and de Swart (1989).26 (For the moment this is sufficient to introduce my point. he regrets right-after “Whenever Juan is cruel. Durative aspect would be opposed to “result state” in her proposal. of a tale). even adjectival predicates of the type of Eskimo fit. note that. Consider (89). he regrets it right after” (88) *Siempre que Juan es esquimal… whenever Juan ser-PRES-3SG Eskimo… “Whenever Juan is Eskimo…” Schmitt does not give any specific explanation for contrasts like that in (87)– (88) but says that only a subset of adjectival predicates (cruel. For this reason. under whenever. mean. siempre que Juan es esquimal. which would lead to the conclusion that estar is excluded in such contexts. for example. he gets angry” In sum. se arrepiente después whenever Juan ser-PRES-3SG cruel. the presence of Eskimo. he bears a life full of difficulties and shortages” Here is a context where the existences of Juan can be iterated. (89) (En todas sus reencarnaciones).” Instead. whenever Juan is Eskimo. Chapter 5. nice) are possible in these contexts. I have formulated the distribution of whenever (which I consider a temporal distributor quantifier) as a subcase of distributivity in general. they select for some sort of durative aspect. in this case. at first sight. I argue that if an appropriate context is built up (think. which is dedicated to Aspect. se enfada whenever Juan estar-PRES-3SG dissatisfied with his work gets-angry “Whenever Juan is dissatisfied with his work. it is licensed. a permanent predicate all through someone’s life. this may seem to be the case. and Jäger (1999) argue that whenever is correct with predicates that are “naturally iterable.

Fernández Leborans considers that in contexts such as the progressive (91) or temporal forms suggesting that the property does not hold any longer (92). to say that ser is allowed because it does not involve internal temporal structure is. I will not have to propose that when the copula ser combines with them it becomes an SL predicate. are also perfectly grammatical in such scenarios. I argue that the oddity (or not) of such aspectual forms is not related to the IL/SL distinction. alludes to the lexical content of the predicates. as she explicitly points out (Luján 1981:206). she argues that the ser/estar distinction can be described as an IL/SL dichotomy. I argue that the availability of the progressive is related to the internal temporal properties of the construction.2. I show that the perfective preterit can be licensed under the appropriate contextual conditions. in her view.4 Summary of Section 2. where no . Luján conceives of ser as a predicate that applies to an individual during a stretch of time with no beginning or end assumed.3 Fernández Leborans (1999). which.Individual-Level Predicates 31 parallel to when(ever)-cases. I consider any appearance of ser to denote an instance of an IL predicate. However. Whereas estar possesses internal temporal structure. 2. is founded on aspectual properties. 2.” As I understand her proposal. ser lacks any inherent temporality and is described as “inert with respect to aspect. For this reason. Likewise. In her study on copulative clauses of Spanish.2. and estar as a predicate that is true of an individual for a delimited period of time. since other predicates. at best.2. which are conveyed by a type of adjectives in a specific way.2. (91) Pedro está siendo muy cruel Pedro is ser-ing very cruel (92) Maria fue muy guapa en su juventud Maria ser-PRET-PERF-3SG very pretty in her youth “Maria was very pretty in her youth” As discussed throughout this work. by “inert with respect to aspect. In sum. just a partial answer.2. Luján (1981) and Schmitt (1992) defend the idea that the distinction between ser and estar should be understood in aspectual terms.” she means that serpredicates are not tied to temporal limits—the reason why she argues such predicates can be described as “durative.2. Although Luján refers to this differentiation as an imperfective/perfective contrast. the opposition. Ser predicates are considered parallel to write or admire. ser predicates work as SL predicates. Another author arguing in similar terms as Luján (1981) and Schmitt (1992) is Fernández Leborans (1999).” As chapter 7 discusses in greater detail. which do involve internal temporal structure (estar in her own proposal and many others).

Therefore. lining myself up with Bosque (1990). as I advanced. since the features that appear to be at stake are not the same as the ones that decide the differences among different inner-aspect types. Fernández Leborans (1999) argues that the IL/SL distinction is a primary parameter of lexical aspect.). after having introduced a formal description of inner aspect and other temporal domains. the subject is categorized as belonging to the class denoted by the predicate. I will assume that with estar (SL) predicates. ser is underspecified with respect to aspect. I have critically reviewed her arguments. I have considered the native intuitions about the different interpretation of such pairs.32 Individuals in Time endpoint is involved. and others that are able to combine with both copular verbs. is the topic of chapter 4. for the time being. in the rest of the work I will consider that any predicate in combination with ser is IL and any predicate in combination with estar is SL. yielding minimal pairs. known as “inner aspect. paying special attention to her discussion on the behavior of ser in combination with certain adjectives (cruel. when a predicate appears in combination with ser. mean. which is that the IL/SL distinction is not a matter of inner aspect. which. a process where a delimited point is involved. based on Escandell-Vidal and Leonetti 2002. Demonte (1999). It is premature to introduce and develop a discussion about whether the real differentiation between IL and SL concerns inner aspect. More concretely.3 Summary of Section 2. that there are some that combine just with estar. That is. I have concluded that the opposition ser/estar is adequately described in terms of the IL/SL split. What I can advance here is the conclusion I will draw. others that combine just with ser.” Along similar lines.” or “aktionsart. APs present a more complicated paradigm. it can be said. etc. Whereas NP predicates obligatorily combine with ser. whereas estar involves aspectual properties (it corresponds to a result state of an accomplishment verb). Although I will add some remarks in chapter 7. Since the descriptions of the readings fall in line with Carlson’s (1977) distinction IL/SL.2 In this section I introduced the distribution of copular verbs in Spanish.” “lexical aspect. The terms in which Luján treats the dichotomy IL/SL refer to the internal temporal properties of the predicates. rather than states. 2. I will describe ser-predicates as classificatory. I have shown that her explanation does not account for the fact that only a concrete type of adjective leads ser predicates to pattern with activities. depending on which verb is used as the leading hint to define the opposition ser/estar. Fernández Leborans (1999). and Escandell-Vidal and Leonetti (2002). whereas estar predicates are taken as parallel to write a letter. the property is associated to a concrete circumstance (also along the lines of Higginbotham & .2. Schmitt (1992) considers that the difference between ser and estar is aspectual in the sense that. I undertake this point in chapter 7.

3 The Structure of Copular Constructions Because this work concerns IL predicates in copular constructions. In this way. which be takes. from where it later raises until reaching the specifier of Inflection Phrase (Tense). In the aforementioned examples. (95) IP 2 Johni I′ 2 is AP 2 ti A′ g blond . Stowell proposed to analyze the subject and the predicate of copular sentences as integrants of an SC. 1981) proposal that the copula takes (as its complement) a small clause (SC) containing the NP subject and the predicate. I will just debate whether SL-hood is a matter reducible to properties of (outer or inner) aspect or tense. Stowell drew a parallelism between the copula and verbs like consider: (93) be 1 be SC 1 John blond (94) consider 2 consider SC 2 John intelligent More specifically. Stowell proposed that the SCs were in fact projections of the predicate. In essence. the SCs are APs. I will introduce here the syntactic analysis I assume for copular sentences. 2.Individual-Level Predicates 33 Ramchand 1996). I adopt Stowell’s (1978. How this is brought about is a complex issue that I discuss only partially in chapter 7. leaving a trace. the NP acting as their argument is generated in the specifier of such APs.

Heycock’s original motivation for proposing an extra position between the SC and certain verbs was to provide an extra slot. and remain but not with others like seem or be considered. *The worst problem seems his attitude (100) a. *The worst problem was considered his attitude The different structures proposed for the two types of verbs are in (101) and (102). which in turn selects for the lexical SC. whereas others select for a functional projection (Aspect). whether an SC has functional projections depends on the selectional properties of the verb taking the SC. In particular. which the predicate can pass through.34 Individuals in Time As depicted in (95). More recently. Our real problem becomes John b. the copular verb takes a lexical predication structure. According to Heycock. John is the culprit b. What to do next remains the real problem b. The following examples show the asymmetry between the two types of verbs in allowing or disallowing inversion. predicate inversion is allowed with verbs such as be. become. The culprit is John (97) a. Stowell argued that. Heycock (1994. to account for the phenomenon of predicate inversion. she proposes that some verbs select for a purely lexical SC. (96) a. (101) VP 2 spec V′ 2 seem AP be considered 2 DP AP . John becomes our real problem (98) a. SCs lack functional projections. The real problem remains what to do next (99) a. unlike matrix clauses. His attitude seems the worst problem b. 1995) argued that. His attitude was considered the worst problem b.

there is no Aspect node mediating between the SC and be. IL predicates do not. In turn. AspP 2 remain (event) Asp′ ∃ 2 Asp AP 2 DP AP According to Heycock. it obtains existential quantification. become. They only involve a lexical SC. the thematic DP subject generates inside the lexical SC. on the semantic properties of the predicate in the SC. as in (101)). 27 The possibility that the event argument may be present or not is presented as an assumption. assumed to be associated with existential quantification. in Heycock’s proposal. and remain select for an Aspect projection. the selection of a lexical SC or an Aspect node depends. whereas SL predicates project an Aspect projection. (The latter is the case for both.Individual-Level Predicates 35 (102) VP 2 spec V′ 2 be. such as Becker (2000). for other authors. precisely. with no allusion to the distinction IL/SL. (103) Firemen are/become/remain available (existential or generic reading) (104) Firemen seem/are considered available (generic reading only) As mentioned. The essence of her proposal is in (105) and (106). When the event argument is not present and there is no existential quantification. be. if any variables (indefinites. be selects for Aspect. the projection of Aspect (and the event argument) depends on the selectional properties of the raising verbs. become. In other words. if the predicate in the SC is SL. Becker argues that if the predicate in the SC is IL. is present. bare plurals) are introduced. Due to the existential quantification associated with it. However. . when the event argument is not present and when Aspect is not projected. the interpretation for such variables is generic. although not necessarily for an event argument.27 When the event argument. Restricting her attention to copular cases with be.

. The evidence adduced to by Felser for the presence of an eventive argument is that perception verbs complements have to be “eventive. in the presence of a functional projection (Aspect) and an Event Phrase (in the spirit of Kratzer 1988.” rather than stative. *I saw John know the answer As for the existence of an extra projection. 1995) or its lack thereof. I saw John draw a circle b. as in (107). Becker’s argumentation about the connection between the projection of Aspect and Event Phrase and SL-hood is an extension of Felser’s (1999) proposal for perception-verb complements. 28 See section 2. Felser argues for the existence of an eventive argument and for the existence of a functional projection that. besides hosting the event argument. can host an argument (a subject) in its specifier.28 (107) a. Felser presents evidence of a filled specifier position.1 for a discussion in which opposition (eventive/stative or IL/SL) is at stake.36 Individuals in Time (105) IP 2 I′ 2 be AspP 2 Asp′ 2 Asp Event Phrase 2 SC 2 DP PP (106) IP 2 I′ 2 be SC 2 DP NP Becker considers that SL and IL copular clauses differ structurally—namely.

perception verbs take Aspect Phrases. assuming that to occurs in the T head (Emonds 1976. Furthermore. (109) *We saw [that John draw/ drawing a circle] (110) *We saw [John to draw a circle] The fact that an overt complementizer (that) is excluded suggests that the complement is not a CP. 1995) in that SL predicates do not involve the projection of just an event argument (here an Event Phrase). Heycock (1994). unlike Schmitt. Whereas SL predicates are linked to aspectual content. In particular. As can be appreciated from (105). Becker’s (2000) idea differs with Kratzer’s (1988.1.1. suggests that the perception-verb complement is not a TP either. Summarizing. who. argued for the existence of two homophonous verbs be. Thus. but the projection of two nodes. Becker considers that an aspectual distinction does not exclude in and of itself a split conceived as IL versus SL. see (105). Because there is not a thematic subject. Regarding the internal constitution of the SCs taken by the copular verb. Becker establishes a correspondence between the presence of Aspect and SL-hood. for Becker. Event Phrase and Aspect Phrase. She just needs to assume one copular verb. whereas Stowell (1981) originally conceived SCs as lexical projections.Individual-Level Predicates 37 (108) I wouldn’t like to see [there be so many mistakes] Her argumentation goes as follows. Becker (2000) takes this correlation between the projection of an Eventive Phrase and an Aspect Phrase and extends it to the representation of SL predicates in main clauses. such an aspectual content is absent in IL predicates. I argue that IL-hood 29 See also Guéron and Hoekstra 1995. which are presented as codependent. the difference between SL and IL copular constructions resides in the complement of the copula. . as pointed out in section 2. Felser concludes. In the first place. since both establish a relationship between copular distinction and aspectual content. Felser proposes the extra layer to be an Aspect projection. the exclusion of a to-infinitive. differing in this respect from Carlson (1977). The reasons to discard a CP or TP projections come from the ungrammaticality of sentences like (109) and (110). in particular to copular sentences. Pollock 1989). other authors have argued for the existence of functional projections inside them too. authors such as Schmitt (1992). and Becker (2000)29 have put forth the presence of a functional node (Aspect). Likewise. Chomsky 1986. the contribution of this book will be twofold. However. it cannot be said to have been generated inside the VP but inserted at a higher level within the perception-verb complement. The proposal defended by Becker (2000) shares the spirit of Schmitt’s (1992). Thus.

I start by dealing with their internal temporal characteristics (discussing whether they are stative. I do not submit the idea that only estar-predicates involve aspectual content. then. I consider IL predicates to be classificatory predicates that apply directly to the individual. as the minimal pairs of “copula + AP” in Spanish suggest. that is. 1981). As a result. dynamic. dynamicity. an important part of this book is dedicated to investigating the syntactic constituency of the SCs the copular verb ser takes and their temporal properties. as I will argue. I will develop a hypothesis about Aspect in predicates combining with ser and. and Fernández Leborans (1999). etc. In particular. from Kratzer 1988. 1995. I develop this hypothesis in chapter 4.. I continue with an analysis of their outer aspectual properties (their combination with imperfect. Schmitt (1992). subject of the predication. and Chierchia 1995. such as the distribution of copular verbs in Spanish.3 introduced the syntactic analysis of copular constructions I will assume. 2.. strictly speaking.) in chapter 5 and finish by examining their tense properties (i.e.e. NP). I argued that such a distinction is operative in the grammar of natural languages since it captures important contrasts. and the predicate (AP. I will not argue for a plain functional head but. In the second place. I therefore discarded those proposals establishing a strict correlation between IL-hood and the permanency of the property in the subject (differing. temporal interpretation) in chapter 6. Following Stowell (1978.38 Individuals in Time and SCs with aspectual content are not incompatible. which contributes activity properties) comes from a lexical item—namely. I defend IL SCs with aspectual content. I propose that the aspectual import I will be dealing with (i. PP. In essence. Chapter 7 takes up the debate about the characterization of the SL/IL distinction. aspectual content does not amount to SL-hood. I will not consider that Aspect is. In this work I defend a finer grained distinction among IL constructions with ser. much in the line of Milsark (1974) and Carlson (1977). Therefore. I do not conceive of the IL/SL contrast as a permanent/temporary contrast. for a lexical item whose semantic import has a correlation in the aspectual realm. taking into account the analysis of the temporal domains presented. I consider that the copular verb takes an SC containing the nominal argument. whereas SL predicates refer to properties that are presented linked to a situation. . Centered on ser + AP cases. Because my primary concern is copular IL predicates.4 Summary of the Chapter This chapter introduced the dichotomy between IL and SL predicates. a preposition. based on Hale 1984 and Demirdache and Uribe-Etxebarría 2000. etc. section 2. perfective. unlike Luján (1981).) in the next two chapters. based on the different aspectual properties they show. the difference between IL and SL copular clauses. among others).

whereas another can be justified as dynamic predicates (i.1 Inner Aspect and Event Types In this section I deal with two points. I described individual-level (IL) predicates as predicates that are said of their subject independently from a given situation. 3. One group tests out as states.e. First. the criteria to distinguish them. Specifically. also called “aktionsart” or “situation aspect” (Smith 1991). Inner aspect has to do with properties such as duration. I first introduce the notion of “inner aspect” and the different types of events that can be distinguished according to it. Because permanency and stativity are temporal concepts in nature.4—namely. Differing from most previous literature. or whether there is more than one (IL) copular verb. 1993). or delimitation. Once we have become familiar with the different event types. I present one of them in section 3. culmination.. the extent up to which the dichotomy between states and activities is real and grammatically relevant. To accomplish this task. I will examine the temporal properties of these predicates.1 Inner Aspect “Inner aspect” (Verkuyl 1989. refers to the internal temporal properties of events in sentences. their internal temporal properties (inner-aspect properties). and the tests to diagnose them. I will show that not all cases in which the copula ser participates share the same inner-aspect properties. I will demonstrate that copular adjectival constructions can be classified in two groups. as activities). Such properties tell us whether the event inherently tends . I then present a survey of the distinct types of predicates traditionally distinguished in the literature. I also emphasized that this should not be taken to mean that IL predicates are stative and permanent predicates. I defend a finer grained distinction among IL predicates based on the different inner-aspect properties they show. According to classical and well-established aspectual tests. I will bring the proofs onto IL predicates. Other more specific issues. 3. are relegated to the next chapter. Such a contrasting behavior raises a number of questions and invites reflection on some general issues. as has been widely assumed across the literature.Chapter 3 Event Classes and Individual-Level Predicates In chapter 2. according to a set of tests. in this chapter.1. such as where the stem of the contrast resides (in the copular verb or in the adjective itself). I start by investigating. I give a few introductory notes about inner aspect and its relevance in the description of the properties of sentences.

I will use the term “eventuality.1. Taking into account the presence or absence of the cited features (duration. I leave the arguments of this last issue for later.” where it is.2 Event Types and Event Structure The first known study of event classification goes back to Aristotle. semantics. My intention in doing so is to present the basic features describing the nature of events and the properties that more than one type share. and “events. whose main insights’ spirit has been shared until today. making no further distinction. Pustejovsky (1988. 2000. . culmination. 2003) have shown that adverbial modification.g. many researchers have attempted to identify a number of types into which events can be classified.” The term “inner aspect” contrasts with what I refer to as “outer aspect” (also following Verkuyl 1989. over. “telic. Aristotle distinguished between those with an inherent terminal or culminating point. Egg (1995). Investigations by Tenny (1987. Grimshaw (1990). named “kinesis-verbs” (e. or De Swart (1998). 3. In his work Metaphysics. Borer (1994. or syntax) or which grammatical phenomena can be directly derived from it..40 Individuals in Time toward an endpoint. He distinguished between “states. The events that lack an inherent endpoint are named “atelic”. at this point I will briefly summarize the classifying work. A large number of event classifications are offered in the literature. although in and of itself does not explain event structure. event classification proves helpful in describing the basic characteristics of events that need to be explained. 1989.1) Among events. 1 Other authors. 1994). van Voorst (1988). As Rosen (1999) points out. and Ritter and Rosen (1996. can be derived from inner aspectual properties. I will focus on those that offered insights that are generally adopted in some way or other. or before its beginning. This taxonomic line of study does not address questions as to how and where events are represented in the grammar (within the lexicon. 1993) in future chapters. (To refer to all event types including states. build the house). 2005). or delimitation) and their combination in a predicate. outer aspect does not have to do with the internal structure of events but with the number of instances a certain eventuality takes place and whether it is presented as ongoing. and I reserve the term “event” to allude to nonstate types. whether it does not.” following Bach [1986]. 1998. reserve the term “event” to telic predicates.” where dynamicity is not involved at all (be red). 1991). As will become clear. Dowty (1991). Aristotle proposed a classification based on the notions of dynamicity and terminus. as well as thematic relations and the syntactic projection of arguments. This will be an important point when discussing the event types of the predicates I am interested in here. those that involve culmination. or whether it involves no duration at all. such as Herweg (1991).

(5) is a resultative construction: Pablo walked until he got dizzy. To make the characterization of each type clear. Vendler divided eventualities into “states.g. The action is delimited and its duration gets also delimited as a consequence.” Vendler considered predicate classes as a strictly lexical matter of verbs. This behavior holds in (2) as well. let us consider the meaning of the following examples: (1) (2) (3) (4) (5) (6) (7) (8) Pablo walked Pablo walked around the park Pablo walked to the park Pablo walked a mile Pablo walked himself dizzy Pablo fell asleep Pablo is tall Pablo sneezed In (1) the NP subject was involved in a process of walking where no specific endpoint is logically entailed or needed. 1994). It is the same with (4). where the verb appears with an internal argument (a cognate object). In (3). Although we lean toward thinking that the subject will not be walking forever. It has been observed that whereas a sentence . nothing in the predicate (in the sentence) forces an endpoint.” Sentence (3) does not behave like this. Predicates that behave this way are known as “activities. 1991) proved wrong. Finally.” “achievements.” This division has been kept in its basic terms by most of the subsequent literature.” It is worthwhile to add something here about the difference between activities and accomplishments. can be considered the most influential work on classification system. the locative phrase delimits the process. which also bounds the process. 1989. Kenny (1963). where the adjunct around the park refers to the space where such an event took place. This taxonomy of predicates spread in the Anglo-Saxon tradition mainly through the work of Ryle (1949). and Pustejovsky (1988. Smith (1991) added an additional class: “semelfactives.Event Classes and Individual-Level Predicates 41 and those that are ongoing without any definite terminus. That is. together with Dowty’s (1979) set of diagnostics. The predicates that fill the template described for (3)–(5) are called “accomplishments.. Vendler’s work. and Vendler (1967). The park can be conceived as the goal of the walking process undertaken by the subject of the sentence. These authors got deeper in the characterization of event types and provided linguistic tests to diagnose the membership of a predicate to a class. Tenny (1987.” and “accomplishments. named “energeiaverbs” (e. walk).” “activities. 1993). When the state of dizziness is reached. the event of walking ends. a point that authors such as Verkuyl (1972. the event of “walking” denotes a process that is not internally delimited.

Instantaneous and nonculminating events swim. Event types Sentences (1)–(8) illustrate the event classification as proposed by Vendler (1957. sneeze. build houses and read novels are conceived as different building-houses and reading-novels events distributed over time. 1967) (in addition to the class proposed by Smith). fall sleep. walk around the park be green. that is why. know.” In (8). arrive. That is. Whereas the former two clearly refer to processes that need not ever end. knock Table 3.1. This predicate does not take time or involve any kind of “process”. I return to the meaning of habituality in chapter 5. be tall.1 briefly characterizes each event type and provides some more examples. Events with duration but no endpoint die. and it is the habit that is not understood as delimited in time. love. the subject is involved in an action that is instantaneous and lacks an intrinsic endpoint. However. since they can be . awaken.42 Individuals in Time like (4) works as an accomplishment. we also see that such templates cannot be a pure lexical matter. the latter two refer to a habit that does not have to end. be born. They lack any kind of internal structure. belong. In (6). 1994 and Pustejovsky 1988). explode walk to the beach. Achievements Instantaneous events. with an endpoint but no duration Actions with a culminating point that take duration to be completed. there is a slight difference between activity-typed events like walk or push a cart and others like build houses or read novels. Verbs involving such a property are called “achievements. each building-a-house or reading-a-novel event is conceivable to reach an endpoint. so I will not discuss them further. build a house. be sick. Table 3. fall asleep denotes an endpoint. hate cough. However. collapse. push a cart. it lacks duration. Both are conceived as habitual happenings. importantly. write novels. but. read a chapter Accomplishments Activities States Semelfactives No-actions that hold in time but do not take time. Semelfactives are not of great relevance to this work. recognize. They can be considered as complex events with internal steps toward an end. such events are called “semelfactives” (Smith 1991). Eventualities like this are named “states.” The predicate in (7). in itself. has neither duration nor endpoint. As noted earlier. the same with a bare plural as an internal argument instead (build houses) behaves as an activity (see especially Tenny 1987. 1989. it is not an action in any sense.

has been taken (since Lakoff 1966. 1993). The other (more radical) option is to think that there are no such event type primitives. the possible or impossible occurrence in the progressive form. (3)). Juan está paseando Juan is walking d.Event Classes and Individual-Level Predicates 43 affected by other elements present in the sentence (cf.2. Many of the proofs consist of testing the compatibility of the predicate with an adverb or a verbal form that explicitly express the property under testing. Only the latter can appear in such a form. The test in (9). (The judgments I give are based on native intuitions from Spanish. aspect is a crucially compositional issue. Others make visible the features that two or more event types have in common. Tenny (1987. 1991). and Dowty (1979). 1989. I present a summary of the tests to diagnose event-type membership from Kenny (1963). I then proceed with the tests that differentiate among different types of events.. Verkuyl (1972. proposes. Identifying which type an eventuality behaves like may be helpful just to pinpoint the determining inner features of a predicate. I will follow this second option throughout this work. and Ritter and Rosen (1996). made clear that internal arguments and certain types of locative phrases affect the category a predicate can be classified into before entering into syntax. Each test establishes a distinction of a different kind.1. (9) Occurrence in the progressive form a. I will be adding some discussion about them. Pustejovsky (1988. others test out not aspectual properties strictly speaking. Some distinguish between states versus nonstates. among others. for example. like agency. As I present them.) 3. ??Juan está dándose cuenta de que su madre tiene razón Juan is realizing that his mother is right c. Vendler (1967). 1994). Finally. There still exists the option to believe that verbs have a sort of default aspectual type at the lexical level that is “modified” throughout the syntactic derivation. I start out by discussing some tests that distinguish statives versus nonstatives in (9)–(11).1 Events versus States. Next. among others. rather than a lexical one. Juan está trazando un círculo Juan is drawing a circle . (2) vs. That is. *Juan está siendo alto Juan is being tall b. Ryle (1949). put it. but properties that seem to depend on aspectual ones. as Pustejovsky (1988). as Verkuyl (1993) and Borer (2005). Put in other words. 1970) as a proof distinguishing statives versus nonstatives. “event templates” do not have a grammatical reality.

there is a process preceding it. but. with an achievement it gets an inchoative sense. although the specific semantic weight of achievements is carried by the culminating point. (10) Pseudo-cleft with happen (‘take place’) a. The good or bad combination with adverbs such as “usually” or “regularly” is generally taken as a proof of the interpretation of present tense as 2 Not all the achievements seem to allow for the progressive at the same level. where a previous process might be more difficult to justify. we will see that they share a part of their semantic component with states (which is why they do not give perfect results). as well. present in the structure. it is easy to note that the progressive form does not have the same meaning as all the predicates it can be compatible with. It is this process. Those authors who conceive achievements as events lacking duration also judge the progressive form as excluded for such a type. Whereas with an activity and an accomplishment. on the other hand. Roughly said. (i) *Juan está encontrando la aguja Juan is finding the needle (ii) *Juan está reconociendo la foto Juan is recognizing the picture . the partial acceptability of cases like (9b) shows a property of the nature of achievements. Lo que sucedió fue que Juan trazó un círculo What happened was that Juan drew a circle Kenny (1963) noted that only nonstatives get a habitual interpretation in the present tense. the progressive means that the eventuality is in its beginning. the progressive looks degraded. However.2 As Pustejovsky (1988) observed. it means that the eventuality is in progress. On the one hand. Lo que sucedió fue que Juan paseó What happened was that Juan walked d. Lo que sucedió fue que Juan se dio cuenta de que su madre tenía razón What happened was that Juan realized that his mother was right c. *Lo que sucedió fue que Juan era alto What happened was that Juan was tall b. Only verbs referring to actions or processes—that is. However. they are not totally excluded. The test in (10) separates eventualities in two. that can be conceived in progress and accept a progressive form.44 Individuals in Time Whereas activities (9c) and accomplishments (9d) give grammatical results in the progressive form. The status of achievements (9b) is a bit trickier. states (9a) do not. things that “happen”—can be complements of verbs whose meaning precisely denotes that or can appear in pseudo-clefts constructions with them. With predicates like (i) or (ii).

Cinque 1999). Juan realizes that his mother is right c. A durative adverbial. The eventualities whose interpretation in the present is not habitual are states. Normalmente. I argue that the crucial condition for a predicate to appear in habitual form is its possibility of restarting. Juan pasea Usually. I start out with the tests in (13) and (14). States are argued to be the only type that cannot be interpreted as “habitual. it might seem that this test is a diagnostic for IL-hood rather than for stativity.2.” as their reaction against combination with habitual adverbs proves. Juan draws a circle Nevertheless. on the one hand. *Normalmente. . Juan se da cuenta de que su madre tiene razón Usually. with stative though SL predicates. Juan sabe matemáticas Usually. are completely grammatical. Juan knows mathematics b. (11) Interpretation as habitual in present tense a. In this subsection I go through a number of tests that separate activities and states from achievements and accomplishments.3.Event Classes and Individual-Level Predicates 45 habitual (Comrie 1976.1.2 Activities and States versus Achievements and Accomplishments. Juan traza un círculo Usually. 3. I show that this is a possibility for every type of predicate (including IL predicates) if a suitable context is created. according to the temporal adverbials that refer to the time an event lasts they admit. and those that do not. which divide activities and states. However. sentences like Juan is usually sick or Juan is usually at his best friend’s place. whereas Juan walks does. Normalmente. involving an IL stative predicate. on the other.2).4 Whereas (11a). Normalmente. is excluded in combination with a habitual adverb (which proves it cannot be interpreted as habitual). Such compatibility cuts the pie of eventualities in two: those that have a habitual interpretation. it is worthwhile to notice that when no (habitual) adverb is present. Juan walks d. and achievements and accomplishments. It seems clear that Juan is sick or Juan is at his best friend’s place does not have the habitual reading as the only (and preferred) reading. whereas (SL) statives can be said to be ambiguous. 4 I thank Tim Stowell for interesting discussion on this topic. Bertinetto 1986. such as “for + x time” goes well with those predicates that refer to eventualities that extend over time with no involvement of an 3 I make a few more remarks on this in chapter 5 (section 5. nonstatives are interpreted as habitual.

5 (12) For + x time a. For a recent account about the different readings available in such sentences. which is the reason why they are compatible just with eventualities entailing a delimit point. the entailment of a delimit 5 Note that this test does not distinguish between IL and SL predicates.46 Individuals in Time ending point. (13) In + x time a. They express for how long someone has been engaged in a particular activity or for how long a particular state has held of an individual. As the ungrammaticality of (14) shows. as (13) shows. *Pablo estuvo enfermo en tres semanas Pablo was ill in three weeks b. Pablo se dio cuenta de que su madre tenía razón en tres semanas Pablo realized that his mother was right in three weeks The following contexts also group achievements and accomplishments together. see Piñón 1999. Pablo viajó durante tres semanas Pablo traveled for three weeks c. *Pablo construyó una casa durante tres semanas 6 Pablo built a house for three weeks d. Pablo estuvo enfermo durante tres semanas Pablo was ill for three weeks b. IL predicates can also combine with for-adverbials: (i) Juan fue camarero durante tres semanas Juan was a waiter for three weeks 6 Not all accomplishments are excluded with a for-adverbial: (i) The police jailed John for four weeks The for-adverbial refers to the period of time that the result state from jailing lasts. *Pablo se dio cuenta de que su madre tenía razón durante tres semanas Pablo realized that his mother was right for three weeks In turn. *Pablo viajó en tres semanas Pablo traveled in three weeks c. in-adverbials refer to the period of time it takes to complete a certain event. . Pablo construyó una casa en tres semanas Pablo built a house in three weeks d.

and he is still walking b.3 Activities and Accomplishments versus States and Achievements. Whereas it is licit to establish (16b) as an entailment from (16a). Juan estaba enfermo y todavía lo está Juan was sick and he still is The next test (Kenny 1963). in (16). it is legitimate to assert that the event has taken place at any point of its process. Juan salió a pasear por el parque y todavía sigue paseando Juan went for a walk. . *Juan arregló la lámpara. it is not licit to draw (16d) from (16c). rather than between IL and SL predicates. (17) separates predicates that 7 Note that an IL predicate gives the same results as a SL (stative) one like (15b): (i) Juan era camarero y todavía lo es Juan was a waiter and he still is As discussed in chapter 6. Pablo ha construido una casa Pablo has built a house 3. since there is no endpoint privileged. Pablo ha paseado Pablo has walked c.1. Pablo está construyendo una casa Pablo is building a house d. accomplishments can be asserted to have taken place just when the process toward a culminating endpoint is over. which is that they can be interrupted in their process and their subjects may have the control over the stopping of the event. The test in (17) reveals one property that activities and accomplishments share. With activities. and he is still arriving (15) a.2. unlike activities and states. (16) Perfect entailment from the progressive form a. this test and the for-adverbial one distinguish between permanent and nonpermanent properties. constitutes another diagnostic to differentiate accomplishments from activities. Pablo está paseando Pablo is walking b. (14) a. but not when it is ongoing. *Juan llegó.Event Classes and Individual-Level Predicates 47 point in accomplishments and achievements makes them be incompatible with an assertion of noncompletion. However. and he is still fixing it b. y todavía sigue arreglándola John fixed the lamp.7 as (15) proves. y todavía sigue llegando Juan arrived.

Pablo ha dejado de pasear Pablo has given up walking d. those event types lacking dynamicity.. testing the possibility of combination of predicates with dejar de ‘give up’. *Pablo paró de darse cuenta de que su madre tenía razón Pablo stopped realizing that his mother was right It is interesting to consider (17) together with (18).8 (e. I judge its combination with activities and accomplishments as correct when they are interpreted as a habit. with an activity. they show different preferences regarding the aspectual class of their complements. there is no difference between SL and (nonpermanent) IL statives. Its meaning is extremely close to parar de. (18a) means that the subject does not love Maria anymore). are predicted to be uncombinable with parar de. but. (18) As a complement of dejar de ‘give up’ a.48 Individuals in Time can serve as a complement for parar de ‘stop’ and those that cannot. Logically. *Pablo ha dejado de construir la casa Pablo has given up building the house c. the preferred reading is that the subject stopped being involved in an 8 In this respect. Pablo paró de construir la casa Pablo stopped building the house c. like achievements and states. Pablo paró de pasear Pablo stopped walking d. interestingly. (i) Juan ha dejado/*parado de estar enfermo Juan has given up/stopped being sick (ii) Juan ha dejado/*parado de ser camarero Juan has given up/stopped being a waiter . *Pablo paró de amar a María Pablo stopped loving María b. (17) As a complement of parar de ‘stop’ a. De Miguel considers that dejar de is just combinable with states.g. Both are ungrammatical after parar de and grammatical after dejar de. as de Miguel (1999) notices. ?Pablo ha dejado de darse cuenta de que su madre tiene razón Pablo has given up realizing that his mother is right Whereas with a state. as in (18c). dejar de means that the state stopped holding. however. Pablo ha dejado de amar a María Pablo has given up loving María b.

Pablo dejó de pasear por una temporada/??/?por unos minutos Pablo gave up walking for a period/for a few minutes Whereas the durative adverbial for a few minutes goes well with parar de. where an adverbial signaling an extended period of time fits perfectly. (19) a. ‘Juan finished walking for that evening’). suggests. 3. However. the verb finish does not accept anything but accomplishments as complements. Pablo paró de pasear por unos minutos/*por una temporada Pablo stopped walking for a few minutes/for a period of time b.1. among others. As Pustejovsky (1988). but he does not anymore’. In this case what is interpreted that stops happening are different occurrences of ‘realizing’. it is grammatical if we think of a habitual interpretation like ‘(Usually) Juan used to realize about when his mother was right. consider the judgment of (18d).4 Accomplishments versus All the Rest. The fact that both states and habits can appear after dejar de suggests that states and habits have some properties in common. a durative signaling an extended span of time does not. rather than a concrete instance of it. The contrary is observed with dejar de. what gets interpreted as interrupted is the habit of walking. The following tests (complements of finish and their interpretation in combination with the adverbial almost) draw a line between accomplishments and the rest of event types. the habit of undertaking such an activity. *Pablo ha terminado de darse cuenta de que su madre tenía razón Pablo has finished realizing that his mother was right 9 Activity verbs can appear after finish if we give the event an arbitrary endpoint (Juan terminó de pasear por esa tarde. an achievement. for a period of time. (20) As a complement of finish a. There is a contrast with (17c) above. In a similar vein.Event Classes and Individual-Level Predicates 49 event of going walking as a habit. ??Pablo terminó de pasear9 Pablo finished walking d.2. . which makes them close to accomplishments (‘John finished his [daily] walk for that evening’). That is. which is understood as the interruption of a particular instance of the event. *Pablo ha terminado de estar enfermo Pablo has finished being sick b. This difference can be observed in the pair of (19). in principle expected to not combine with dejar de. Pablo ha terminado de construir la casa Pablo has finished building the house c.

the subject did not even start out the process of building. the subject was involved in the process of building up a house but did not finish it. ¡Pasea! Take a walk! d. that is. Pablo casi caminaba con ocho meses Pablo almost walked when he was eight months c. yielding ungrammatical results. ??¡Date cuenta de que estás confundido! Realize that you are wrong! . In one of them. the tests in (22)–(25) can be used to diagnose stativity.1.5 Agentivity Tests. (21) Interpretative entailments from the adverb almost a. an achievement. and (21c). (Section 3. too.5 in this regard). I want to introduce another group detecting the presence of an agentive subject (Ryle 1949. Pablo casi construyó la casa Pablo almost built the house b.2. an activity.2 gives a more precise notion of “agent. get differentiated from activities and accomplishments. Pablo casi se dio cuenta de que su madre tenía razón Pablo almost realized that his mother was right Sentence (21a). (21b). 1970). However. Lakoff 1966. as van Voorst (1988) and others point out (see also section 3. *¡Estate enfermo! Be sick! b. states and achievements. can only have the second interpretation. an accomplishment. in (22)–(25). (22) Occurrence in command imperative form a. In the other. the subject did not get to walk or realize. As can be appreciated. which are grammatical. To conclude this survey of aspectual tests.50 Individuals in Time The possible interpretations of almost put in contrast accomplishments with other event types. 3. To the extent that an agent is conceivable only if a proper event (a nonstate) is at hand. ¡Construye la casa! Build the house! c. can have two interpretations.”) The roles that DPs play in the event and the analysis of the event itself usually overlap.

Event Classes and Individual-Level Predicates 51 (23) Pseudocleft with hacer ‘do’ a. *Pablo se dio cuenta de que estaba confundido deliberadamente Pablo realized that he was wrong deliberately (25) As a complement of force a. A “+” . the good behavior of activities and accomplishments in agency contexts is due to the property of dynamicity the two of them involve.1. *Lo que hizo fue darse cuenta de que estaba confundido What he did was realizing that he was wrong (24) Combination with volition adverbials (deliberately) a. Pablo forzó a Juan a pasear Pablo forced John to walk d. *Pablo forzó a Juan a darse cuenta de que estaba confundido Pablo forced Juan to realize that he was wrong As discussed further in section 3. In the second part. Pablo forzó a Juan a construir la casa Pablo forced John to build the house c. where other elements like the object play a determining role.3 Summary of Section 3. Lo que hizo fue pasear What he did was walking d. but it is a compositional matter. Pablo paseó deliberadamente Pablo walked deliberately d.2 summarizes the tests and the behavior of event types. *Lo que hizo fue estar enfermo What he did was be sick b. *Pablo forzó a Juan a estar enfermo Pablo forced John to be sick b. Pablo construyó la casa deliberadamente Pablo built the house deliberately c. I presented different tests usually employed to diagnose which event type a concrete predicate belongs to. 3. *Pablo estuvo enfermo deliberadamente Pablo was sick deliberately b.2.1 The first part of this section presented the notion of inner aspect. This accounts for the ungrammaticality of both states and achievements. Lo que hizo fue construir la casa What he did was build the house c. I also introduced the idea that inner aspect does not depend on the inner properties of the verb itself. Table 3. which pattern together because of the stative feature they share.

2.52 Individuals in Time indicates when the event type can have the interpretation described or can appear in the context expressed on the left of the table. (17). the results of the tests (9)–(11). and (22)–(25) reveal that activities and accomplishments share an important part of their . Tests for event types Among other things. as pointed out above in the description of each test. The bracketed numbers in the leftmost column refers to the number of the test as they appear in the text above. and a “%” if the event type shows a positive behavior in the test but just under a certain interpretation. (20). States – – – + + / – + – / Activities + + + + + + + % + – Accomplishments + + + – + – – + – + + Achievements % + + – + – / – % – – (9) Progressive form (10) Complement of ‘happen’ (11)Interpretation as habitual in present tense (12) For + x time (13) In x time (14) and (15) Assertion of noncompletion (16) Perfect entailment from the progressive (17) Complement of parar de (18) Complement of dejar de (20) Complement of ‘finish’ (21) Ambiguous interpretation with the adverb ‘almost’ (22) Command imperative form (23) Cleft with ‘do’ (24) Combination with ‘deliberately’-like adverbs (25) Complement of ‘force’ – – – + + + + + + – – – – + + – Table 3. The symbol “/” means that the test does not apply to the event type for some reason. a “–” when it cannot.

To the extent that just some types of events can be undertaken and commanded.” The cleft with “do” and the test of command imperative detect eventualities that can be “done. three notions involved: causation. To begin this investigation. I consider two aspects related to this.” Also. agency tests work as event-type tests. activities and accomplishments may be the most persistent co-patterning of the table.1 A Cluster of Notions Although.” Then. As will become clear. First. Further discussion about the tests and about some concrete event types will come in section 3. when I discuss the aspectual behavior of copular IL predicates. volitionality. As mentioned before. agency tests detect eventualities that denote stuff that can be “brought about. agency tests separate states and achievements from activities and accomplishments.2. In fact. although volition usually entails control. and (22)–(25) also suggest that states and achievements have characteristics in common. (26) (27) El portazo rompió el cristal The bang broke the glass Juan rompió el cristal Juan broke the glass . I present a brief discussion about the agency tests above and about the notion of “agent” that I will observe in this work. it is important to distinguish the semantic features it contains and to establish whether there is some feature more basic than the other. which diagnose agency. cannot hold with any willfulness on the subject’s behalf. consider the classical minimal pair in (26) and (27). stuff that is not performed. The results of (17). Likewise. but just holds. In the following section.2 A Brief Stop at “Agentivity” The set of tests in (22)–(25). there are. I will briefly introduce the relation between agents and event structure. Actually. 3. the term “agent” clearly denotes ‘the argument that brings about an event’. while volitionality and control are properties that just animate causers can embrace. In agency. not all controllers involve volition. and control. which gives bad results in contexts with adverbials such a deliberately. causation is the most basic notion involved in agency contexts. (20). may be used to discern properties of eventualities as well. In particular.3. Vendler (1967) himself considered states and achievements a natural class.Event Classes and Individual-Level Predicates 53 properties. at least.” and therefore “commanded. In this section. in principle. 3. I will deal with the different notions that get blended in the term “agent.

Both the bang and Juan are the arguments “responsible” of the event. El frío congeló el agua → ??Lo que hizo el frío fue congelar el agua The cold froze the water → What the cold did was freeze the water b. a scenario where past tense was possible might be created. and John who causes it in (27). tense seems to play a role in this regard.10 Although this is the traditional view. yielding a general statement interpretation. in this respect. given that they behave differently in certain contexts: (28) a. the pseudo-cleft with “do” is possible only when the cause is animate (29b). (i) El pepino elimina las arrugas Cucumber eliminates wrinkles (ii) Lo que hace el pepino es eliminar las arrugas What cucumber does is eliminate wrinkles The same sentence in past tense sounds degraded: (iii) ??El pepino eliminó las arrugas Cucumber eliminated wrinkles (iv) ??Lo que hizo el pepino fue eliminar las arrugas What cucumber did was eliminate wrinkles Still. it is easy to note that they are not completely the same. they can be considered on a par. Interestingly. can appear just with certain causers. Juan congeló el agua → Lo que hizo Juan fue congelar el agua Juan froze the water → What John did was freeze the water Adverbs such as deliberately or intentionally. Juan rompió el cristal deliberadamente Juan broke the glass deliberately (29) a. which mark volition. *El portazo rompió el cristal deliberadamente The bang broke the glass deliberately b. inanimate DPs are perfectly natural in do-pseudo-clefts. the feature that differentiates between bang and Juan seems to be animacy. Observe (i). and the acceptability of the do-pseudo cleft construction in (ii). However. which suggests that such adverbs prove the presence of animate causers. this is not totally true. where the DP cucumber cannot be understood as a volitional agent. In particular. A: Wasn’t he being treated with insulin? 10 .54 Individuals in Time In sentence (26) it is the bang that causes the breaking of the glass. When present tense is involved. Consider the following: (v) A: Why did he die? B: He had a problem with his sugar and acetone blood levels. but just as a cause. Likewise.

Collins 1997. given that not all causers are understood the same way. 11 See Chomsky 1995.12 (30) vP 3 deliberately vP → volition causer 2 DP[+animate] v (vi) B: Sí.” However. As a result. languages exist that show agent/patient splits. lo que hizo la insulina fue restablecer los niveles de azúcar. A possible alternative is the following. and references therein about this issue. multiplying elements in the lexicon is an uneconomical move. 2000. but.Event Classes and Individual-Level Predicates 55 In the framework of the generative Minimalist Program (Chomsky 1995 and subsequent work). One conveying something like “animate cause” and another one involving “inanimate cause. thus far. such as Lakhota (Mithun 1991). which makes a morphosyntactic distinction between subjects that perform or instigate the action (those of verbs like walk) and subjects that do not (those of fall.11 any uneconomical proposal becomes automatically undesirable. It can very well be the case that it is the whole set “specifier + head” that licenses intentional adverbs (or not). or be tired). such as Algonquian languages (Berardo 1999) and Southern Tiwa (Rosen 1984). pero no pudo controlar los de acetona. it is usually considered that a functional projection (socalled small v) contributes the content of ‘cause’. . die. I have not found any clear example showing that an inanimate causer is treated morphosyntactically different from an animate one. a more economical answer can be that the interpretation of a cause with or without volition results from the conjoint interpretation of the head plus the specifier. Now. rather than the head (small v) by itself. Assuming the generative premises that economy principles are observed by the system of language. where agreement markers switch according to the relative animacy of the arguments. Since volition seems to be tied to properties of the argument in the specifier. The first guess can be that there are actually two types of small v. 2001a. There are languages with animacy splits. As is known. but it could not control the acetone levels. What the insulin did was reestablish sugar levels. a possible question might be how volitional cause is distinguished from nonvolitional cause. the DP in its specifier gets interpreted as the ‘causer’. I am grateful to Olga Fernández Soriano for discussion on this point. There are many other kinds of splits regarding subjects (Ritter & Rosen 2003). 12 It would be crucial to investigate whether there are languages that distinguish overtly between volition or animate causers and nonvolition ones.

but not all causers are on a par. the kind of adverb. Consider (ii). The evidence mostly studied regarding the subject/agent orientation of the cited adverbs comes from passive sentences. or at the I′ level. I wanted to show that “cause” is the most basic feature involved in the concept of “agency. Whereas in the active counterpart (i). or the intentionality of the agent. From now on.” An agent is the cause of the event. Adverbials of the type of deliberately can be adjuncts at the V′ level. volition—that is linked to the properties of the noun heading the DP subject (animacy). However. if it (assumedly) bears the role of “patient” (given the fact the sentence is in passive voice)? McConnell-Ginet (1982) proposes that the passive auxiliary be is interpreted as ‘to act’. where they assign an adjunct role to the surface subject. the willfulness. Zubizarreta (1987) argues that adjuncts such as adverbials assign “adjunct theta roles” to arguments. If deliberately is an agency marker. proving its agentive properties. With this brief discussion. the adverb clearly refers to the DP subject. Cinque (1999). (i) The doctors examined Juan deliberately Los médicos examinaron a Juan deliberadamente (ii) Juan was examined by the doctors deliberately Juan fue examinado por los médicos deliberadamente The interpretation of the adverb in cases like (ii) has been considered ambiguous in English. which explains the agentive property of its subject. I will reserve the term “agent” to causers with volition. where they assign a role to the underlying subject. reporting an example from Jackendoff 1972. and the position of the adverb matter. the passive form of (i).” Adverbials such as deliberately refer to the volition. Some involve an additional characteristic— namely. This is a very complex issue where the kind of verb.56 Individuals in Time (31) vP 3 *deliberately vP → nonvolition causer 2 DP[–animate] v More accurately. in the passive (ii). (I give the Spanish counterparts. it has also been extensively discussed whether such adverbs refer to the “subject” rather than to the “agent” in a sentence. such an ambiguity poses a puzzle: how is it that it can also refer to the DP subject. also gives (iii) and (iv) as cases showing that some adverbs retain their orientation on the subject. it is not an obligatory interpretation (see below for discussion of sentences like John hit him with no intention). I turn next to some discussion about the property of “volition.13 Actually.) (iii) Joe intentionally has seduced Mary Joe intencionalmente ha seducido a Mary 13 . agents have been defined as ‘causers Tests based on the suitability of adverbs such as deliberately or intentionally were proposed by Ryle (1949) to diagnose whether the DP subject of a predicate plays an agent role. since my judgments are based on them. we should say that when the DP is [+animate] the interpretation as a volition causer is possible. it seems to be able to refer to the subject as well as to the DP in the agentive by-phrase complement.

with the intention that Mary reaches the state of being seduced. since their meaning refers to the by-phrase. which is what allows for the presence of volition. Since John is interpreted as responsible for Bill’s action. As Martin puts it. they differ. (32) John made Bill eat The interpretation of this sentence is one where John performs some action (such as issuing a command) to affect Bill’s performance of an independent action. I would like to add something else in a similar vein. However.” since the latter is not met in some cases. (33) Juan golpeó a Pedro deliberadamente/intencionadamente Juan hit Pedro deliberately/intentionally (34) ??Juan pegó a Pedro deliberadamente/intencionadamente Juan struck Pedro deliberately/intentionally (iv) Mary intentionally has been seduced by Joe Mary ha sido seducida intencionalmente por Joe My judgment on this in Spanish differs from the one given for English. In terms of volition.Event Classes and Individual-Level Predicates 57 involving volition’ (Dowty 1975. . we would be unable to tell the crucial property shared by the DPs John and Bill in (32). Thus. If we relied solely on “volition. with the intention. On my view. There is another property even more basic.” just taking into account tests based on deliberately-like adverbials. From (iii) I understand that Joe has undertaken a set of actions (of heterogeneous nature) with the goal. From (iv) I understand that Mary has been the object of a seduction and that seduction was caused by an individual who brought it about with a clear intention in mind. concurring with Martin (1991). The affected state of having been seduced is by no means intentional. since it is a state in itself. such an interpretation is maintained in the passive. bearing the agent role in passives. that the notion of “controllability” gives a much more restricted definition of agent. of getting Mary seduced—that is. Such a property is controllability. it is not always the case that volition or willfulness is involved in animate causers. Bill can be seen as “less volitional than John. Consider (33) and (34). both John and Bill can be considered “agents” in terms of controllability. Both have the control of the action. among others).” Sentence (32) shows that volition and controllability do not always go in hand in hand. however. in this work I will make use of deliberately and intentionally as true agent-oriented adverbs. This leads us to conclude. while Bill controls or mediates the degree to which the eating is successful. John controls or mediates the extent to which the causation is successful. Consider a sentence like (32) from Martin 1991. rather than “volition. Thus.

as (33) and (34) show.” 14 I do not investigate here how this difference between hit and strike should be represented.14 The adverbials sound odd when they mean precisely the opposite of doing something on purpose (cf. many predicates can be interpreted as taken place with intention on the subject’s behalf. adverbs such as deliberately and intentionally. tests based on the suitability of adverbs like deliberately are testing whether a volition feature accompanies the control feature. However. This makes (35)–(36) a stronger test than deliberately in (24). There are some predicates whose subjects involve the feature of volition and others where such a characteristic is optional. Other adverbials yield similar contrasts. sound a bit funny with (34). In particular.58 Individuals in Time The verbs golpear and pegar are very close in meaning. Consider also (35) and (36). That is. In conclusion. . (35) Juan golpeó a Pedro sin querer Juan hit Pedro without intention (36) ??/*Juan pegó a Pedro sin querer Juan struck Pedro without intention Just the predicate hit sounds natural with an adverbial like without intention. which seems more basic in essence. as by chance. The same adverbial complement with strike yields an odd sentence. (37) Juan golpeó a Pedro a propósito Juan hit Pedro on purpose (38) ??Juan pegó a Pedro a propósito Juan struck Pedro on purpose That is. with an adverbial complement meaning that the action was brought about without intention. Adverbials marking volition (on purpose) and lack of volition (without intention. these adverbs sound funny and redundant when the fact to bring about the action inherently involves doing it on purpose. though acceptable in both predicates. but only a subset involves the subject’s intention as a necessary condition. it sounds somewhat redundant and therefore funny in (38). so often used across the literature. One possibility (pointed out to me by Tim Stowell) would be that strike incorporated an abstract element analogous to the adverb deliberately. the adverbials sound redundant in (34) (there is no other way to strike someone but intentionally) but not in (33). Whereas on purpose makes sense in (37). with no enthusiasm. to learn about the properties of the predicates. because an action like strike can only be understood as performed intentionally. above without intention). reluctantly) can be labeled “volition oriented.

When the subject is [–animate] and. *Pablo paró de darse cuenta de que su madre tenía razón Pablo stopped realizing that his mother was right Only subjects that can have control over the event (agents) can be subjects of parar de. thus cannot be understood as agentive. I will assume that the ability of an eventive verb to combine with the progressive is evidence that the event involves dynamicity. the infinitive is grammatical after parar de only when the subject can be agentive. As the following contrasts show.2 Agents in Event Structure In this subsection I offer a few introductory words about the possible role of agents in event structure.” yet nevertheless is grammatical. used as tests to identify the different event types. the sentences are ungrammatical. Given the special status of the subject of meteorological verbs (cf. At this point I just want to notice the consonance of the results of the tests diagnosing dynamicity and of those diagnosing agency. (17) As a complement of parar de ‘stop’ a.1. I will not consider this kind of case as a severe counterexample to my suggestion. that is. Assuming that progression in time defines dynamic events. the verb parar de implies dynamic control by its subject. In the first place. *Pablo paró de amar a María Pablo stopped loving María b. which I take up later in the work. where the verb parar de does not have a subject that can be said to involve “control.Event Classes and Individual-Level Predicates 59 3. the syntactic contexts reviewed in section 3. which suggests that only with dynamic events the subject has control over its ending.15 15 Tim Stowell (p. repeated here). dynamic events yield grammatical sentences when they appear as complements of parar de ‘stop’ (see (17). show a certain correspondence between events that are agentive and those that can appear in the progressive. As is traditional.) points out the interesting case of Paró de llover ‘It stopped raining’. Pablo paró de pasear Pablo stopped walking d. these events are typically activities and accomplishments. we can conclude that only dynamic events can have agentive subjects. We can prove that an event involves “internal movement” if it can express its progression in time. where their subject was a [+animate] DP such as Jupiter).c. In the second place. Pablo paró de construir la casa Pablo stopped building the house c. .2. Latin.

both of which give completely grammatical results.60 Individuals in Time (39) a. Juan empezó a pasear a las tres de la tarde Juan started walking at three in the afternoon Example (40b) shows that achievements are ungrammatical as complements of start. the beginning is understood as the moment at which the subject starts undertaking the process. With activities and achievements. ??Juan empezó a estar enfermo con 50 años Juan started being sick in his fifties b. with states. where there is no agency or responsibility over the event involved. the beginning is interpreted as alluding to the moment at which the state starts holding of the subject. when he saw he did not trust him A crucial difference between states and activities and accomplishments is that. States (40a) are not completely excluded. Juan empezó a escribir una carta a las tres de la tarde Juan started writing a letter at three in the afternoon d. however. but they contrast with activities and accomplishments. *Juan empezó a estar en Madrid cuando consiguió trabajo Juan started being in Madrid when he got a job b. *Juan empezó a llegar a la meta a las tres de la tarde Juan started arriving at the goal at three in the afternoon c. Some stative predicates yield sentences more acceptable than others: (41) a. Only in dynamic events is the subject is responsible for the beginning of the event. which is expected since the weight of their semantic import is on the resultant state (Pustejovsky 1988). too. . Consider the following examples: (40) As a complement of empezar ‘start’ a. Juan empezó a estar a disgusto con su jefe cuando vio que desconfiaba de él Juan started being uncomfortable with his boss. Empezar can have both [+animate] as well as [–animate] subjects. whose subjects can only be [+animate] and agentive. Juan/*la rueda paró de girar Juan/the wheel stopped rotating The interpretation of the subjects of activities and achievements in combination with start is interesting. Juan/*la rueda paró de moverse Juan/the wheel stopped moving b. In this respect. there is an interesting difference with parar de.

given the correspondence between certain types of events and certain types of subjects. as mentioned earlier. Throughout the tests. it seems that three properties can be connected: dynamicity. is the lexical expression of IL-hood). 1989. control over the ending. In the case of accomplishments. it is natural to wonder whether the agentive argument plays any role in aspectual properties. as Ritter and Rosen (2003) put it.Event Classes and Individual-Level Predicates 61 (42) La rueda/Juan empezó a girar The wheel/Juan started rotating In sum. (which. Now. We saw that not all arguments that could be labeled “agents” embrace all such properties. although all volition agents involve control.2. This will constitute the first step in the account of such predicates offered here.3 Aspectual Differences among Individual-Level Predicates This section explores a group of IL predicates from an aspectual perspective. Tenny (1987. 3.3 Summary of Section 3. the end is inherent. but I leave the discussion about their interaction with aktionsart to chapter 5. I introduced the issue of the role of agents in the eventive structure itself as “initiators” of the event. Specifically. the ending is arbitrary and subject to the subject’s willingness. not all controller agents involve volition. which overlaps with aspectual notions. In the last part of the section. in the event structure. I looked briefly at some of the notions involved with the concept “agent”: causation. 1994) and Pustejovsky (1988) observed the role that objects have in the structure of events. which has been very often linked to conscious volition only. Objects grammaticize telicity. The fact that agents only appear with events that may have an end and whose subjects have the control over its beginning16 leads us to think that agentive subjects are responsible for the initiation and durativity (the sustaining) of the event. all must involve causation. I will ignore the differences between imperfect and perfective aspect that exist in Spanish. the initiator of the event. 3. I examine the behavior of copular cases with ser. Agentive subjects can grammaticize an aspectual role—namely. In particular. 16 . In turn. through the battery of tests introduced above to diagnose event types.2 This section includes remarks about the tests regarding the thematic property of agency. as argued in chapter 2. This supposes a remark on the notion of agency. volition. in the case of activities. They are relevant to some points. and control. but only animates can involve volition and control. and (with the considerations just made) control over the beginning of the event.

*Lo que sucedió fue que Juan era rubio What happened was that Juan was blond c. 3. Juan estaba siendo muy cruel con Pedro Juan was being very cruel with Pedro (44) Pseudocleft with happen (‘take place’) a. *Normalmente. The predicates I will test are a classificatory AP (Eskimo) and two qualifying APs (blond and cruel).” They are considered to be eventualities that lack any bounding and are inherently durative.1. ?Lo que sucedió fue que Juan fue muy cruel con su adversario What happened was that Juan was cruel with Pablo (45) Interpretation as habitual in present tense a. As described above. (43) Occurrence in the progressive form a. The eventualities yielding ungrammatical results are taken as stative.1 regarding the present-tense interpretation as habitual (see also chapter 5 for more important . IL predicates are. states are eventualities that do not “happen. Juan es cruel con Pedro Usually Juan is cruel with Pedro Despite all the observations introduced in section 3. simply. Juan es rubio Usually Juan is blond c. as has been widely agreed on in the literature.3. I will make use of the tests introduced above in the same order.” do not “take place. *Juan estaba siendo rubio Juan was being blond c. *Juan estaba siendo esquimal Juan was being an Eskimo b.3.1 Aspectual Tests on IL Predicates The aim of this subsection is. To do that.62 Individuals in Time 3. *Lo que sucedió fue que Juan fue esquimal What happened was that Juan was an Eskimo b.” but just “hold. Juan es esquimal Usually Juan is Eskimo b. *Normalmente. to identify the aspectual properties of IL predicates. taken to belong to the group of states. In (43)–(45) I reproduce the tests marking for stativity versus eventivity.1 Events versus States.2. Normalmente. I will concentrate on copular clauses and test out whether all IL predicates are appropriately grouped in the aspectual set of states. in general.1.

Consider a scenario like the following: (i) Lo que sucede/pasa es que su primo es esquimal. .18 states and activities from accomplishments and achievements. hold for that person’s entire lifetime. 17 Much like do-pseudoclefts (see footnote 10). Actually. that is why she can speak Inuit” 18 As mentioned in section 3. which is why the limitation of its duration with a for + x time complement sounds weird. although it gives ungrammaticality for one of the predicates that above tested out as a state (Eskimo). It is interesting to note. As I introduced in chapter 2 (and will discuss further in chapter 6). The test in (46) distinguishes. Rather. The main point of (46) and (47) is to show that the AP manifesting eventive-like properties (cruel) pattern with activities rather than with accomplishments or achievements. por eso sabe hablar Inuit “What happens is that her cousin is Eskimo.b)17 and the other comprises one kind of AP manifesting eventive properties (c). on the other. but the oddness in (46a) calls for further explanation. Even in the absence of the habitual adverbial. If we compare these results with the ones obtained in the previous section. the next step is to find out exactly what kind of event it is.1.2. its acceptability in (46b) is expected. predicates like Eskimo refer to properties that. and cruel. once again.Event Classes and Individual-Level Predicates 63 remarks). Since it appears that the AP in (c) behaves like an event. pseudoclefts with happen are acceptable with states when present tense is involved. states are durative and for-adverbials are durative adverbials. a clear contrast emerges among different types of APs in combination with the copula ser.2. as the contrasts between (46c) and (47c) show. (45c) gets a habitual interpretation. we see that one group comprises APs showing a state nature (a. It does so as well with IL predicates. as shown earlier. in-adverbials refer to the period of time it takes to complete a certain event. it seems that these tests consistently separate the APs in two groups: Eskimo and blond.2 Activities and States versus Achievements and Accomplishments. on the one hand. rather than a state. In turn.1. which is why they are compatible only with eventualities entailing a delimit point. for-adverbials express how long someone has been engaged in a particular activity or a state has held of an individual. 3. when they hold of an individual. If. the reasons for its ungrammaticality are independent from the test itself. as reasoned in the literature. These three tests show that not all APs in combination with ser behave the same way.3. that the period that a stative IL adjectival predicate such as blond holds of the subject can be restricted with an adverbial (46b).

since this is only possible with eventive predicates.2.3 Activities and Accomplishments versus States and Achievements.2. Once again. (48) Perfect entailment from the progressive form a. *Juan fue esquimal durante tres semanas Juan was an Eskimo for three weeks b.3.1. Juan está siendo cruel con Pedro Juan is being cruel to Pedro b. Juan fue rubio durante tres años Juan was blond for three years c. which argues in favor of considering the predicate (ser cruel) as an activity.3. *Juan fue rubio en dos horas Juan was blond in two hours c. As mentioned in section 3.64 Individuals in Time (46) For + x time a. (48b) is a licit entailment from (48a). *Juan fue cruel con su adversario en dos horas Juan was cruel with his opponent in two hours The test in (48) corroborates the results of (46) and (47). a predicate like ser + AP would be unexpected as a complement of parar de ‘stop’. such a predicate shows a nonstate patterning. only activities. the combination of ser and an AP such as cruel ‘cruel’ looks good. However. *Juan fue esquimal en dos horas Juan was an Eskimo in two hours b. tests like (49) divide activities and accomplishments from states and achievements. Juan fue cruel con su adversario durante diez minutos de la entrevista Juan was cruel with his opponent for ten minutes in the interview (47) In + x time a. where there is no endpoint. As shown in section 3. . Juan ha sido cruel con Pedro Juan has been cruel to Pedro 3.1. In principle. and not accomplishments.2. yield a perfect entailment from the progressive form.1.

(51) a. . One can perfectly say something like (51a).Event Classes and Individual-Level Predicates 65 (49) As a complement of parar de ‘stop’ a.” be blond need not be so (see chapter 6 for further discussion). ?Juan paró de ser cruel con su adversario Juan stopped being cruel to his opponent As a complement of dejar de. as we saw was the case with activities above in (18). First. *Juan dejó de ser esquimal cuando llegó a la adolescencia Juan stopped being Eskimo when he became an adolescent Second. As I mentioned before. Juan dejó de ser rubio cuando llegó a la adolescencia Juan stopped being blond when he became an adolescent b.19 (I am circumscribing myself to these brief notes about predicates like cruel here. this has to do with the type of state at hand. whereas one canot say (51b) under any circumstance. This test breaks up the states in two: one type can stop holding. Whereas be Eskimo is considered a “lifetime property. (50) As a complement of dejar de ‘give up’ a. Juan dejó de ser rubio Juan gave up being blond c. *Juan dejó de ser esquimal Juan gave up being Eskimo b. however. Chapter 4 devotes more attention to their complexity). observe that not all stative AP predicates are fine as a complement of dejar de. the reference of his opponent is subject to reference variability (‘whoever he was’). as the ungrammaticality of (50a) suggests. it is interesting that (50c) give up being cruel to his opponent can get an interpretation like ‘the habit of being cruel to his opponent stopped holding’. Juan dejó de ser cruel con su adversario Juan gave up being cruel to his opponent Two facts are worth noticing here. according to my earlier arguments. ser + cruel proved to be possible. whereas the other cannot. 19 When it is understood as a habit. *Juan paró de ser rubio Juan stopped being blond c. which is unsurprising since both states and activities can. *Juan paró de ser esquimal Juan stopped being an Eskimo b.

Although be cruel predicates do not give perfect results after finish. the adverbial almost can have two interpretations: (a) that the subject was involved in the process denoted by the predicate but did not finish it. whose combination with almost is impossible. except for cruel.4. (52) As a complement of finish a.2.1. ?Juan casi fue cruel con su adversario Juan almost was cruel to his opponent b.20 Although (53a) does not sound extremely natural. Juan no empezó a ser cruel con su adversario Juan did not start being cruel to his opponent c.1. None of the AP predicates with ser. .4 Accomplishments versus All the Rest.1.66 Individuals in Time 3. *Juan casi fue esquimal Juan almost was an Eskimo 3. and (b) that the subject did not even start out the process. ?Juan terminó de ser cruel con su adversario (en el minuto 10 del debate) Juan finished being cruel to his opponent (in the 10th minute of the debate) The following sentences with almost. Only ser cruel is possible under a command imperative form (54) or in combination with adverbs like deliberately (55). like (53c). 20 Recall that. since the unique licit entailment from (53a) is (53b). Juan finished being an Eskimo b. they clearly contrast with the other two. fit in canonical agentive contexts. *Juan terminó de ser rubio Juan finished being blond c. as mentioned in section 3. which suggests that dynamic properties are somehow involved in these predicates. (53) Interpretative entailments from the adverb almost a. The former reading is available with accomplishments. *Juan terminó de ser esquimal.5 Agentivity Tests. it contrasts with other APs with ser. which also test accomplishments versus all the other eventualities.3. All tests give the same results. Tests (52) and (53) show some more contrasts consistent with the results from (49).3. the latter is with activities. confirm be cruel as an activity. The next set of contrasts aim at testing whether the process that be cruel can be considered is agentive or not.

” besides being the instigator (controller) of the process. Sé cruel con tu adversario Be cruel to your opponent! (55) Combination with volition adverbials (deliberately) a.1. *Juan fue esquimal deliberadamente Juan was an Eskimo deliberately b. all of which need. because of their inherent semantic reasons. Juan fue cruel con su adversario deliberadamente Juan was cruel to his opponent deliberately Note also that be cruel cannot combine with adverbials such as without intention. or regret (60). a sharp contrast with the other APs is observed. be cruel is a grammatical complement for verbs like persuade. (56) ??Juan fue cruel con su adversario sin querer Juan was cruel to his opponent without intention/will (57) ??Juan fue cruel con su adversario sin darse cuenta Juan was cruel to his opponent without noticing (58) Juan fue cruel con su adversario a propósito Juan was cruel to his opponent on purpose Finally. *Sé esquimal Be Eskimo! b.2. . but. *Sé rubio Be blond! c. I take this to show that the subject of this predicate unequivocally involves “volition. Be cruel may not be absolutely natural in do-cleft constructions (61).Event Classes and Individual-Level Predicates 67 (54) Occurrence in the command imperative form a. an agentive subject for their infinitival complements. still. force (59). As argued in section 3. *Juan fue rubio deliberadamente Juan was blond deliberately c.

b) can appear as a complement of regret but with a counterfactual interpretation: I regretted to be blond (because he said he would only fall in love with dark-haired people). ?Lo que Juan hizo fue ser cruel con su adversario What Juan did was be cruel to his opponent 3. First. we have learned that. There seems to be a group of APs that can be taken as Examples (60a. 22 Recall that although not all processes are necessary agentive (cf. taking into account the properties of this group introduced in section 3.68 Individuals in Time (59) As a complement of persuade or force a. there seems to be a group whose behavior patterns with canonical activities. The leaf was falling down from the tree). among state be-predicates. I will attribute such an eventive-like behavior to the properties of the AP itself. Juan lamentó haber sido cruel con su adversario Juan regretted to have been cruel to his opponent (61) Pseudocleft with do a. *Lo que Juan hizo fue ser esquimal What Juan did was be Eskimo b.3 In this section we have learned two things.e.1. Juan forzó a Pedro a ser cruel con su adversario Juan forced Pedro to be cruel to his opponent (60) As a complement of regret21 a. as shown by a number of tests.22 I dedicate detailed attention to these predicates that do not behave as states in chapter 4. *Juan lamentó haber sido esquimal Juan regretted to have been an Eskimo b. Among them. agency is notable.2 Summary of Section 3. *Lo que Juan hizo fue ser rubio What Juan did was be blond c. activities and accomplishments) can be agentive. where it is attributed to the copular verb be instead. Rather. not all of them can be considered alike. as shown in the set of aspectual tests.2.3. *Juan lamentó haber sido rubio Juan regretted to have been blond c. only processes (i. This amounts to claiming that there is a group of IL cases that involve typical eventive properties. *Juan forzó a Pedro a ser esquimal Juan forced Pedro to be Eskimo b. differing from other authors’ account. *Juan forzó a Pedro a ser rubio Juan forced Pedro to be blond c. not all IL predicates belong to the category of states.. 21 . Second.

swim. As noticed earlier. (62) For + x time a.. belong) lack an inherent endpoint. (62) and (63) reproduce these tests.4 States and Activities: A Grammatically Relevant Difference? One of the important points observed thus far is the aspectual difference between states and activities found among IL predicates. both are atelic). be sick.4.. In the next section. there is no real distinction between them. They do not advance toward a culminating point that delimits the temporal contour of the eventuality (i. I devote more attention to their interpretation in chapter 6. and another group that cannot. I consider their similarities and differences in this section. 3.g. as shown in the aspectual tests.g. Such a property makes them pattern together in contexts involving for-adverbials and in-adverbials. push a cart.e.. Pablo estuvo enfermo durante tres semanas Pablo was ill for three weeks b. as a consequence. and. states and activities share properties that make them pattern together. Since part of this work deals with the opposition between states and activities. This fact has led a large number of logicians and semanticists to claim that activities and states constitute a natural class. They are distinguishable in contexts that make an endpoint explicit (51).1 Similarities between Activities and States Section 3. walk) and states (e.1 pointed out some similarities between activities and states. 3. *Pablo construyó una casa durante tres semanas Pablo built a house for three weeks d. Both activities (e. Pablo viajó durante tres semanas Pablo traveled for three weeks c. know. For convenience. *Pablo se dio cuenta de que su madre tenía razón durante tres semanas Pablo realized that his mother was right for three weeks . I consider the aspectual difference observed here (those between states and activities) from a broader point of view.Event Classes and Individual-Level Predicates 69 lifetime properties.

Likewise.. Mourelatos (1978). an activity) from 2 P. (65) Pablo is talking Pablo has talked Another important semantic property of activities and states is “additivity” (based on the principle of cumulative reference of Quine 1960). if we assert of John that he pushed a cart (i. to use a more precise term. it is entailed that at every subinterval between 2 P.. defined in (66). If we take a state. as pointed out by Vendler (1967). makes both eventualities homogeneous. *Pablo estuvo enfermo en tres semanas Pablo was ill in three weeks b. Pablo construyó una casa en tres semanas Pablo built a house in three weeks d.M.e. At each moment of (65) it is correct to say both Pablo is talking and Pablo has talked. “homoemerous. including every moment of time I. toward which to tend. it is entailed that John was sick at every subinterval between Monday and Friday. Contexts like (65) also prove the subinterval property. be sick. both activities and states share the so-called subinterval property.M. Bennet and Partee (1972). and assert that it was true of John from Monday to Friday. then the sentence is true at every subinterval of I. to 3 P. As notably put by Bennet and Partee (1972).70 Individuals in Time (63) In + x time a. any part of an activity or a state has the same properties as the whole. and Dowty (1986). Carlson (1981).M. to 3 P.” That is. Pablo se dio cuenta de que su madre tenía razón en tres semanas Pablo realized that his mother was right in three weeks The lack of an endpoint. (66) Additivity Property The result of the sum of a number of portions of x is x . *Pablo viajó en tres semanas Pablo traveled in three weeks c.M. (64) Subinterval Property Subinterval verb phrases have the property that if they are the main phrase of a sentence that is true at some interval of time I. or. John was pushing a cart.

and Bach (1986). The legs of a table are not a table. which yields bad results with count nouns (cf. However. However. In (67) the adverbial mucho refers to the intensity with which Pablo loves Mary. meaning that ‘Pablo drawing a circle takes place very often’. with activities. There are other state predicates that also allow for such a frequency reading. Quine (1960). building a house is not the result from summing portions. The correlations regarding quantification in the verbal and nominal realm are nicely captured in the set of examples (67)–(69) and (70)–(71): states and activities behave like the mass noun water. a table is not the result of minor portions of tables. Carlson (1981). Observing such properties.g. (71)). we always obtain walking or being sick as a result. Pablo is able to walk 45 miles without stopping). noticed the close parallel between the mass-count distinction in the nominal domain and the aspectual classification of predicates. However. For example. or subintervals. see section 5. 24 I thank Tim Stowell for bringing these cases to my attention. Mourelatos (1978). As is known.. among others. For more detailed discussion of frequency readings.Event Classes and Individual-Level Predicates 71 For example. (72) María está mucho en Madrid 24 Maria was in Madrid very much 23 The frequency reading is also present in the case of accomplishments (69). (67) (68) (69) (70) (71) Pablo quiere mucho a María “Pablo loves Maria very much” Pablo camina mucho “Pablo walks very much” *Pablo traza un círculo mucho “Pablo draws a circle very much” Mucha agua “Much water” *Mucha mesa “Much table” It is worth noticing that the interpretation of the quantifier mucho ‘much’ with activities and states is not completely alike.23 In this respect. it can mean that the amount of walking is big (e.2. Likewise. . of building a house. consider (72). if we sum subintervals of walking or being sick. the subdivision and the sum of the parts of a count noun and the result are not of the same nature. each of which is water. or it can also mean that Pablo goes walking very often. much is a quantifier proper for mass nouns. “water” can be divided into parts. and the sum of portions of water is always water. it is ambiguous: in (68).

I will be adding some discussion about the scenarios and the tests themselves as I treat them.2 Differences between Activities and States In this section I concentrate on those syntactic scenarios that reveal dissimilar properties between states and activities.M. These two types of events react differently to such . If John took a brief break of five minutes between. it is clear that John was not swimming at each subinterval from 2 to 3 P.M. Nevertheless. two years). asserting that if John swam from 2 P.e. 3. the difference between homogeneous and heterogeneous eventualities does not give us the differences regarding dynamicity. their different behavior in the progressive form (73). maybe not as strictly speaking as states do. On my view. and. Dynamicity is related to some important semantic issues such as the interpretation possible for subjects (whether they can be agents or not). to 3 P. a property that activities possess but states lack. a mereological perspective is in itself too strong when applied to activities. In this particular respect.M.M. the question to answer should not rely on what chemistry has to say but on what natural languages really care about. Let us begin by looking at some of the differences between states and activities that have been already pointed out here. and 2:30 P. nobody would judge that John is lying if he says I swam from 2 to 3 P.. there is no interval throughout those two years that he did not own the car. too.M.72 Individuals in Time Although the evidence for a homoemerous perspective looks quite convincing. for example..4. John was swimming at every subinterval between 2 and 3 P. Recall.. Semantic entailments such as (65) and contrasts like the one between (67) and (68) with (69) seem strong evidence to me. 2:25 P. say. If John owned a car for two years. In other words. therefore.M.. it is also true that mereological properties do not exhaust all the scenarios and the entire criteria that can be taken into account. there are some problems with it that have been acknowledged in the literature. the similar behavior as complements of stop of achievements and states. and then I went back to school. activities fulfill the subinterval property in a broader sense. As expanded in the next section and in chapter 4. and in agentive scenarios (75). A state such as owning a car can be truthfully predicated of John at every subinterval of the interval in question (i. too. I think that natural languages can be said to acknowledge the subdivisibility property. As has been pointed out many times. However. Would we say that any indefinite subpart of water is water? It seems that there are parts of water too small to be considered water.. after the verb stop (74). for example. inadequate. The real validity of the subdivision property has been debated in the metaphysical and semantic literature with respect to the nominal realm. seems excessive.M. Other event types share properties with states. Actually.. Recall.

?Juan está dándose cuenta de que su madre tiene razón Juan is realizing that his mother is right (74) As a complement of parar de ‘stop’ a. I assumed that the reason for such similar patterning is that achievements involve a (resultant) state in their structure. van Voorst 1988. (73) Occurrence in the progressive form a. Pablo paró de pasear Pablo stopped walking d. due to the subprocess that can be adduced to be involved in them. *Juan está siendo alto Juan is being tall b. Verkuyl 1993. Lo que hizo fue pasear What he did was walk d. Juan está paseando Juan is walking c. etc.Event Classes and Individual-Level Predicates 73 contexts. this can be legitimately taken as evidence for classifying them separately. Pablo paró de construir la casa Pablo stopped building the house c. As a number of authors have pointed out (Smith 1991. which differentiates them from purely dynamic events: activities and accomplishments. Juan está trazando un círculo Juan is drawing a circle d. *Pablo paró de darse cuenta de que su madre tenía razón Pablo stopped realizing that his mother was right (75) Pseudocleft with do a. achievements (76) can get . *Lo que hizo fue estar enfermo What he did was be sick b. Lo que hizo fue construir la casa What he did was build the house c. in my opinion. *Pablo paró de amar a María Pablo stopped loving María b. However.). achievements are not completely ungrammatical in contexts such as these. these contexts separate states together with one type of events—namely. and. Following Pustejovsky (1988). *Lo que hizo fue darse cuenta de que Pedro tenía razón What he did was realize that Pedro was right As noticed before. achievements.

(ix) (x) (xi) He ?dejado de/*parado de saber matemáticas cada vez más I have stopped knowing more and more mathematics La película me dejó de/*paró de gustar y me fui del cine I stopped liking the movie and I left the theater He ?dejado de/*parado de conocerla . however. but I found that scene so disgusting that I left the theater. and of the predicate (cf. Know someone. with the interpretation of ‘getting to know’ or ‘know in depth’ can be conceived of as something that can take time (i. Note that the licensing of the progressive depends on the presence of more and more (cf. pero no la compré al final I was liking the table. as a process). the good combination with the progressive is due to other factors than the predicate itself. but I did not buy it at the end A movie can be conceived of as something that develops progressively. La estoy conociendo ahora her I-am knowing now ‘I am getting to know her now’ 25 (iii) The use of the progressive in all these examples is licensed by slightly different factors. The scale interpretation also depends on the nature of the object. that is. pero encontré esa escena tan asquerosa que me fui del cine I was liking the movie. where the interval right before the resultant state is focused (77). Other predicates. (iv) (v) (vi) (vii) *She is knowing mathematics *She is knowing the answer to the question (more and more) *That apartment is belonging to me (more and more) *I am owing that apartment (more and more) Since know mathematics can be cumulative knowledge. (i) (ii) ??Está sabiendo más y más matemáticas cada día She is knowing more and more mathematics every day Me estaba gustando la película.74 Individuals in Time an inchoative interpretation. Since this is a proof distinguishing states from events. (iv)). according to Piñón (2000). (iii) is an example of inchoative meaning similar to the one pointed out for achievements. a table cannot. either. Consider the following examples. too. as its contrast with sentences like (viii) suggests. such as own an apartment do not admit degrees of participation. After parar de they give ungrammatical results. However. I consider that.e. triggered by the adjunct more and more. that all these predicates are grammatical only as complements of dejar de. (vi) and (vii)). the predicate keeps behaving as a state.25 Some predicates that behave like canonical states in some contexts can also appear in the progressive form in others.. The acceptability of the progressive in (ii) seems to be connected to the nature of the object. the answer to the question is not. it can go through a scale measuring its degree. Note. The progressive corresponds to a scale interpretation. in effect. and that process can be shown as ongoing by the progressive. as (v) shows. Finally. However. The progressive gives the interpretation that the process is in its beginning. (viii) ??/*Me estaba gustando la mesa. In (i) the state gets a gradual interpretation.

The same can be said for accomplishments (80) and achievements (81). (78) Marta debe estar enferma Martha must be sick (79) Marta debe pasear todos los días por el parque Martha must walk every day around the park (80) Marta debe trazar un círculo ahora mismo Martha must draw a circle right now (81) Marta debe llegar al final de la carrera Marta must arrive at the end of the race Note that with the stative verb be sick. about examples from English such as those pointed out by van Voorst (1988). rather than stativity versus dynamicity. (xiv) *This table is missing a leg more and more (xv) *This table is getting to miss a leg 26 On my view. the modal has two meanings— namely. Consider the interpretation of the following example: (i) The leaf could fall → ‘the speaker believes it was possible the leaf fell’ → #‘the leaf had the capability of falling’ . the modal has just an epistemic reading. the oddness of (xiv)) or an inchoative one (xv). On the former. However. the speaker expresses a guess about a particular activity she thinks the subject.Event Classes and Individual-Level Predicates 75 (76) Pablo se estaba dando cuenta de que Pedro tenía razón Pablo was realizing that Pedro was right (77) Pablo was about to realize that Pedro was right Consider now two more arguments in favor of distinguishing states and activities. with an activity such as walk around the park. epistemic and deontic. (xii) This table is missing a leg (xiii) *This table misses a leg This pair cannot be explained either by alluding to a scalar interpretation (cf. Piñón (1995) notes that modal verbs have different interpretations with dynamic and stative predicates. the simple present form being ungrammatical. On the latter the speaker expresses a command that Martha has to effectuate. Martha. Roughly described. the speaker makes a guess about Martha’s state. however.26 I have stopped knowing her/getting to know her I have nothing to say. The examples below show where stative verbs have to appear in the progressive form in English. however. the different interpretation of predicates with modal verbs diagnoses agency versus lack of agency. is usually involved in.

that events and states get a different temporal interpretation in (past) complement clauses. they get the so-called past-shifted reading) (83). at least according to the Spanish data. the stative predicate gets located in the past with respect to the main predicate. (82) a.. in Spanish. ------------walked-------said------------Utterance Time----However. Juan said that Peter walked around the park b. mainly for English. which gives the same results as stative predicates regarding its interpretation with modal verbs. It has been argued. 27 Imperfect forms can also yield a past-shifted interpretation (see chapter 5 for more discussion). I want to discuss two arguments that have been appealed to for distinguishing between states and activities. Whereas stative predicates in the subordinate clause are interpreted as overlapping the event of the main one (82). Since I have not introduced outer-aspect notions yet. Juan dijo que Pedro estuvo enfermo Juan said that Pedro be-PRET-PERF-3SG sick In (84) both sentences have a stative predicate in the subordinate clause. outer-aspect differences make a difference regarding the overlapping/past-shifted temporal contrast. -------------------///////said///////---------Utterance Time-------? ? be sick (83) a. John said that Peter was sick b.76 Individuals in Time 3. Juan dijo que Pedro estaba enfermo Juan said that Pedro be-PRET-PERF-3SG sick b.4. This leads us to conclude that the different interpretation in combination with modals diagnoses agency versus lack thereof. this can be considered as an effect due to the properties of (outer) aspect. . but they differ in their outer-aspect properties (imperfect/perfective). eventive predicates (and activities as a representative type) are not interpreted as overlapping the event of the main clause but at an interval previous to it (i. rather than due to inner-aspect properties. I will not focus on the technical treatment but limit myself to show that. My point here is just that the temporal interpretation as simultaneous or past-shifted in completive clauses is not a solid argument for a difference between states and activities. (84) a. In the first case.27 whereas in the second one (perfective). whereas the perfective cannot yield a simultaneous reading. In (i) there is a dynamic verb with a [–animate] subject (therefore incompatible with the existence of agency). the temporal interpretation is simultaneous.e.3 Some Confusing Arguments in Defense of a State/Activity Distinction Finally.

Juan dijo que Pedro caminaba por el parque Juan said that Pedro walk-PRET-IMPF-3SG around the park b. among others. For instance. ------------be sick-------said------------Utterance Time----When the predicate is an activity (i. it can be conceived that the habit has started in the past with respect to the main verb and continues.Event Classes and Individual-Level Predicates 77 (85) a. ------------walk-------said------------Utterance Time----As shown. Let me show how the reasoning goes with an example containing two telic predicates (an achievement and an accomplishment). In turn. temporal interpretation cannot be used as an unequivocal proof to distinguish between states and activities. the temporal interpretation available is a past-shifted reading (88b). nonstative). -------------------///////said///////---------Utterance Time----? ? be sick b. a stative as well as an eventive predicate can yield a simultaneous and a past-shifted reading in completive clauses. whereas eventive forms do. getting a habitual reading (87). took a book is understood as taking place at a particular interval. which happens at another interval. as discussed in Kamp and Rohrer 1983. Due to such a temporal ordering. As a last remark. -------------------///////said///////---------Utterance Time-------? ? walk b. Smith 1999. the reader feels that time has progressed between these two intervals. (86) a. in (89).. The reason is that eventives are interpreted as following each other in time. and Parsons 2000. located after walked into my office. . overlapping the interval at which the main predicate takes place (88a). Juan dijo que Pedro caminó por el parque Juan said that Pedro walk-PRET-PERF-3SG around the park por el parque todos los días (87) Juan dijo que Pedro caminaba Juan said that Pedro walk-PRET-IMPF-3SG around the park every-day “Juan said that Pedro used to walk around the park every day” (88) a. which is their (alleged) dissimilar behavior in narrative contexts. the same situations arise.e. I would like to discuss another argument used to defend a distinction between states and activities. when the activity comes in the perfective form. The gist of these works is that statives do not advance the narration. Thus. When the outer aspect form is imperfect.

played the piano. However. Estaba enfadado Juan walked into my office. Likewise. there are examples. Smith (1999) works on the interpretation of activity verbs with perfect aspect in narration.c. . be angry is interpreted at an interval overlapping the span of time of walk into my office. (90) Juan entró en mi oficina. strolled in the park. which suggests that pragmatic factors may be what precludes simultaneity in (91)–(93). and strolled in the park.78 Individuals in Time (89) Juan entró en mi oficina. All of the predicates in these examples are interpreted as ordered between one another. This fact causes them to be understood as discrete entities that are able to be ordered between each other. it presents the eventuality as delimited or bounded. in these examples activities are interpreted as bounded—that is. Cogió un libro de biología John walked into my office. as the picture from states gave us. and went to the movies. time does not progress in the text but keeps equal. between each activity and the next. in (92). Finally. ate breakfast. and. (91) He got up. He was angry. as explained before. The interval at which be angry is understood overlaps the interval at which walk into my office takes place. rather than “overlapping” with each other. it being impossible to have a forwarded or pastshifted temporal ordering between them. (Since I have not introduced technical notions about aspect yet. time moves.) Smith claims that activities do advance the progression of narration. such as (i). In (91). (93) They rehearsed. with Smith. If they were not bounded. when overlapping takes place. time does not move forward. that when a sentence comes in perfect aspect. Activity predicates are underlined. where two activities can be interpreted as simultaneous. with an (arbitrary) endpoint. they would overlap.28 28 As Tim Stowell (p. in (93) rehearse is taken as happening before stroll in the park. strolled in the park and listened to music. as a consequence. in (90). or between an activity and the accomplishments or achievements in the examples. and. play the piano is understood at an interval clearly ordered before the other activity stroll in the park. According to Smith. stroll in the park is interpreted as ordered before listen to music. (92) She ate breakfast. The following examples are from her work (Smith 1999:491).) points out. He took a book about biology However. let me just assume for the moment.

Event Classes and Individual-Level Predicates

79

One question that naturally arises from Smith’s (1999) perspective (which she does not treat) is what happens if a state comes in perfective form. Consider the minimal pair in (94) and (95), where the stative predicate be dark only differs in the aspect form (imperfect vs. perfective). (94) Me levanté, fui hacia la puerta. Estaba oscuro, me tropecé I stood-up went to the door be-PRET-IMPF-3SG dark I tripped (95) Me levanté, fui hacia la puerta. Estuvo oscuro (unos segundos), me tropecé I stood up, went to the door. Be-PRET-PERF-3SG dark (for a few seconds), I tripped According to Parsons (1990) and Smith (1999, 2004), a stative predicate such as be dark should be interpreted as simultaneous to the preceding and following cited events, which would not advance time narration. However, although that is the case with the imperfect form in (94), with perfective aspect, the reading is available according to which the predicate be dark can be understood before trip and after go to the door: (96) ----------stand up-----go to the door------be dark----trip The possibility of examples like (95) and their corresponding temporal interpretation as (96) shows that inner aspect is not, by itself, a determining factor in narrative advancement. Other factors such as outer aspect (imperfect vs. perfective) play a role, too. By the same token, these cases show that narrative ordering of eventualities does not provide the most solid evidence to argue for a distinction between states and activities.29

(i) She strolled in the park and sang to herself Although it is usually taken it that nonstative predicates advance narration, there are cases that suggest that something else must be at stake in narrative advancement. Consider (i). (i) En verano, salgo con los amigos, me levanto tarde, duermo la siesta, nado… “In summer, I hang out with friends, I get up late, I take a long nap, I swim…” In (i), there are a number of nonstative events, one following the other. Crucially here, unlike the other cases seen above, the hearer does not establish a temporal order among them. That is, take a nap is not understood as following hang out. As a consequence, no narrative progression is derived from the events’ citation. This is another proof showing that inner-aspect properties are not the sole factor in discourse advancement. It may be the case that other pragmatic properties (e.g., mode of discourse) precede the way hearers face temporal ordering and interpretation. Thus, if they are before a type of discourse where the order of the events does not give any relevant hint, the kind of inner-aspect features of the verb does not yield any type of
29

80

Individuals in Time

3.4.4 Summary of Section 3.4 In this section I dealt with the issue whether the distinction between states and activities is legitimate. We have seen two opposite points of view defended in the literature. It is important to note that the motivation for each perspective stems from the features they assume as a criterion to categorize events. The group of authors arguing for no distinction between states and activities (Bennet & Partee 1972, Hinrichs 1986, Bach 1986, Dowty 1986, Herweg 1991, Reinhart 2000) take mereological entailments between eventualities as the most powerful criterion for deciding the categories for predicates. Such a view is known in the literature as the “strong mereological perspective.” The crucial point for them is that both states and activities are homogeneous (or homoemerous). On the other hand, for the authors defending that a distinction between states and activities is legitimate, the presence or absence of dynamics is what establishes the cutting line among eventualities. Situations with dynamism, or “energeia,” take place in time and, in Comrie’s (1976) words, are subject to a new input of energy. When the input of energy ceases, so does the event. Therefore, dynamism entails the assumption of an initial point and the possibility of a final point. I will line up with the distinguishing group and assume that dynamics is behind relevant grammatical properties. The kind of interpretation that DP subjects can have, given a particular predicate, is derivable from properties like dynamism or lack thereof, not from mereological properties. To the extent that the differentiation between dynamic versus stative is relevant to account for these aspects of eventualities, states and activities can legitimately be considered distinct groups whose differences have to be accounted for. The following chapter is dedicated to accounting for such a distinction in the realm of copular adjectival predicates. In the course of the discussion, I have also made some considerations regarding the tests used to classify event types. We have been able to note that the tests proposed in the literature (notably in Vendler 1967 and Dowty 1979), aiming at articulating the criteria for categorizing eventualities, serve only as a rough guide, since the results are not as clear cut as desired in all cases. We have seen that activities share features with states, but we have also noted that this is not peculiar just of these two event types. The two types of events unquestionably considered as a natural and separated class by the strong mereologicalist group (i.e., achievements and accomplishments) share properties with both states and activities, respectively. Regarding the differentiation of states and activities, I took as some of the most reliable proofs those based on the possible complements for parar de or dejar de. I will take up the discussion concerning the stative/dynamic oppositemporal interpretation. For a review of the factors intervening in the temporal interpretation of events in discourse, see Smith 2004; Arche, to appear.

Event Classes and Individual-Level Predicates

81

tion in chapter 4, centered in the realm of copulative clauses. I also discuss there the issue as to where in the grammar (and how) the event type differences are encoded. 3.5 Summary of the Chapter This chapter presented some basic concepts I need to work on IL predicates. In the first section I introduced the notion of inner aspect and the elements that play an important role in its determination. We have seen that aktionsart is not a lexical matter of the verb itself, as originally conceived by some authors (Aristotle, Vendler 1967), but something that concerns, at least, the whole VP (Verkuyl 1972, 1993). In particular, the role played by the internal argument is of high relevance (Tenny 1987, 1989, 1994; van Voorst 1998). I have also reviewed the different behavior of aspect types through a set of tests proposed by Dowty (1979) and others. By their systematic application we have distinguished distinct event types as well as distinct thematic properties of the predicates. One of them, agency, proves to be highly relevant since its availability depends on the type of event. Some remarks regarding the notion of “agency” were made in section 3.2. Next, in section 3.3, we observed the reactions of IL predicates under the set of aspectual tests, and noted that not all of them behave alike. They show an opposition between states and activities. The nature of the contrast between states and activities is discussed in section 3.4, where I considered some semantic and discursive arguments. I concluded that the difference between states and activities is grammatically relevant and needs an account. I take care of such a contrast in the IL (adjectival) realm in chapter 4 in detail.

Chapter 4
Aspectual Alternations in Individual-Level Predicates
In the previous chapter I showed that the combination of the copular verb ser with a certain group of APs did not behave as states but as activities. This chapter is devoted to examine the properties of such cases in detail. One of the main points will be to show that it is not the case that all the constructions that these APs enter in behave as activities. Rather, such a patterning seems to correlate with the presence of another constituent, which I will treat as a case of aspectual alternation. The fact that it is due to a specific syntactic configuration of the APs heading the small clauses taken by the copula that the copular construction gains active properties will lead me to refuse those accounts according to which it is the verb be which displays different properties (section 4.1). To the contrary, my explanation of these constructions stems from the particular properties of the syntactic configuration where APs themselves are. In section 4.2 I argue that the adjectival predicates usually involved in dynamic copular clauses describe properties that can be understood in relation to another individual, which is expressed by a “relational PP” (e.g., “cruel to someone”). I will pay special attention to their ability to combine with a PP complement and I explore (in section 4.3) whether the relational constituent is obligatory or optional, its correlation with other characteristics (the need for an animate subject, for example) and its contribution to the type of eventuality. The conclusion I draw is that the relational constituent correlates with the dynamic aspectual nature of the construction. As a consequence, I argue that the aspectual properties, as well as the interpretive properties of the DPs appearing in the construction, correlate with a concrete syntactic configuration, where a PP is present. Then, in section 4.5, I justify the approach I take to account for the aspectual alternation (stative/dynamic) in copular clauses. I will argue that lexical approaches cannot make sense of the correlations found between the presence of a determined constituent, the thematic properties of the DP subject and the aspectual nature of the construction, whereas a syntactic approach such as that proposed by Borer (2005) provides a natural frame to explain such correlations. In section 4.6, where I develop the core of my proposal, I discuss the syntactic representation of homogeneous predicates (states and activities). Differing from Borer (2005), who argues that activities are the event type by default, while statives (aside from accomplishments) emerge in the presence of dedicated functional structure, I propose that the stative version of copular cases is simpler in structure than the dynamic version, which results out of the

I will defend the preposition heading the PP complement as an actual aspect head. First. This strengthens the hypothesis that the properties exhibited in the dynamic copular clauses are due to the properties of the structure where the adjectives taken by the copula are inserted. In section 4. 2004). properties like agency cannot be attributed to the lexical entry of the verb but to the whole configuration. therefore. Recall pairs like (4) and (5). where the verb is the same but it is observed that agency is a property just of the animate DP subject. and empirically. 4. among others. this is an uneconomical move (and. semantically null. and Demirdache and Uribe-Etxebarría (2000. as has been maintained thus far in the literature.1 The Two-Copulas Hypothesis Partee (1977) and Dowty (1979). with a meaning close to act (2). Theoretically. rather than to the properties of the copular verb itself. Specifically.1. based on Hale (1984). as I discussed in chapter 3. of cases such as (3).84 Individuals in Time presence of a constituent acting as a dynamicity inductor.1 Previous Explanations of Nonstative Copular Clauses In this section I summarize previous proposals concerning the active behavior observed in certain copular clauses. (4) (5) John broke the window on purpose The ball broke the window (*on purpose) . different from the “regular copular be”. 4. The last section summarizes the chapter. (1) (2) (3) John was cruel on purpose John acted cruelly John is tall (*on purpose) Among the limitations of this analysis. Second. as shown in chapter 3. undesirable in itself). the two-copulas hypothesis is unappealing because it multiplies lexical entries with no independent evidence to support it. They proposed a theta-assigning active copula for cases such as (1).7 I discuss the behavior of the dynamic AP predicates when they are taken by verbs other than the copula and show that the properties attributed to their structure make sense of the possible/impossible combinations. Stowell (1993). attributed the agentive properties displayed by adjectives such as cruel to a different lexical entry of the copula. it finds the obstacle that there is no evidence proving the actual existence of such two homophonous copulas. there are the following three.

locatable in time and space. since subjects of activities and accomplishments are typically understood as agents. 2 Rothstein (1999) also notes this is not the case for all activities. To cover this question. it is understood as a property predicated of the whole person. First. In a nutshell. The proposal by Rothstein leaves the following points unanswered. depending on the context. cannot have an agentive reading. if it is the copula that shifts a state into an activity. given the role of be to shift an eventuality from a state into an activity. where Rothstein explicitly acknowledges that activity and agentive properties are not obvious. Rothstein claims that when the adjective comes naked (6). Rothstein (1999) argues (quite unspecifically) that “be + AP” constructions move quite freely among event types. the analysis of the two copulas does not capture the fact that such “agentive properties” are activated with a specific set of adjectives. is more salient. if the small clause (SC) of (7) be polite is eventive by virtue of the presence of be. Rothstein argues that. (6) (7) Mary made Jane polite Mary made Jane be polite Rothstein assumes that all adjectives denote states and argues that the copula be in (7) maps state-denoting predicates (like polite) onto eventualitydenoting predicates (be polite).1.2 an agentive interpretation is the expected one given the event type of the outcome (an activity). the subject of sleep. the copula shifts a state into an eventuality with activity-like properties.2 Be as a Copula Shifting a State into an Activity Rothstein (1999) proposes an alternative account to that based on two copular entries. 1 . with be present. Rothstein argues that. other adjectives (blond. it is not clear how the account can be extended to simpler inflected cases like (8).1 Regarding the interpretation of the subject. Rothstein (1999) resorts to the conversational principles (Grice 1975) of Quantity (“do not say more than required”) and Manner (“avoid obscurity”). (8) Dafna is polite Also. Eskimo) do not appear in activity-agentive contexts. leaving the eventuality in (7). as a state. That is. an activity. She offers a semantic investigation about the behavior of the predicates showing an agentive behavior as a complement of causative verbs like make. for example. As shown in chapter 3 (section 3.Aspectual Alternations in IL Predicates 85 And finally.3). would violate these maxims of Actually. a temporary reading. associated to a particular eventuality. 4. it is left unexplained why the stative reading is also available in such cases. whereas when the copula is overt (7).

age. (9) John was cruel to punish the dog (10) It was cruel of John to punish the dog Stowell argues that adjectival predicates involving the attribution of Mental Properties (MPs) (cruel. in cases such as (11) and (12). Finally. when the adjectives are predicated of two arguments at the same time (an individual and an event). kind. (15) *John was premeditated (16) *John was premeditated to invest in the bonus market . (11) John was cruel (12) To punish the dog was cruel So. as in (11). If the adjective cannot be predicated of an individual (15). at best. mean. I will mention four points. the adjective must be able to be predicated of individuals. shape. Thus. this proposal leaves unexplained why it is a concrete type of adjectives which activates the set of dynamic properties.1. whose behavior I described in chapter 3 as proper to activities in English constructions such as those in (9) and (10).3 Adjectives Predicated of an Implicit Event Argument. it gets excluded from the dyadic usage (16). no adjective referring to physical properties (dimension. as in (9) and (10). With respect to the dyadic usage. to be able to participate in the so-called dyadic constructions. 4.86 Individuals in Time Quantity and Manner. or dyadic. of an individual and an event simultaneously. as the account of the two copular entries. The speaker would be adding something unnecessary. Stowell (1991) suggests that the active behavior observed in certain copular clauses is due to an implicit event argument. when the adjectives are predicated of just one argument. for the preference of one reading over the other. intelligent) can be predicated of an individual. color. and. optionally. and therefore creating confusion to the hearer. MP adjectives can be monadic. First. of an event (12). Stowell studies the set of adjectival predicates.) can appear in dyadic sentences. (13) *To invest in the bonus market was wide/green/squared/old (14) John was wide/old to invest in the bonus market Second. the adjective must be able to be predicated of events. Rothstein’s Gricean explanation does not account for the fact that (7) is ambiguous but just. etc.

from a finer grained typology of eventualities. (21) Accomplishments It was very kind of you to bring me that book/to read my paper/to explain me the problem/to walk me home/to cook dinner for me… (22) Activities Having such heart decease.” Compare (17) and (18).Aspectual Alternations in IL Predicates 87 Third.” (19) States *It was very kind of John to know mathematics/to own a house/to be an African/to want that coat (20) Achievements *It was very cunning of John to reach the top/to recognize the thief/to find the needle. As the following contrasts show. only activities and accomplishments fit. and the tree in (24) depicts the dyadic one. the infinitival arguments can be neither states nor achievements. the event. refers to “actions” rather than “facts. The structures Stowell (1991) proposes for these potentially dyadic APs are the following. The tree in (23) represents the syntax of the adjective in its monadic usage. represented by the infinitive clause. the adjectives that can participate in these dyadic constructions have to fulfill the twofold property of being able to be predicated of an action and of an individual. it is very imprudent of you to run every day/to swim in the ocean In sum. (23) John/To tell the truth is intelligent AP 5 Spec A′ | | | A John intelligent To tell the truth . This is expected. given the fact that only activities and accomplishments refer to “actions. (17) To invest in the bonus market is a cunning/mean action (18) *To invest in the bonus market is a cunning/mean fact Finally.

by the same token. Thus. those that can be agentive. in the dyadic cases the adjective is predicated of the action denoting argument. which is the DP (John).3 I will not undertake an analysis of these dyadic constructions in this work. The controller of this PRO is the DP subject (John). whereas in English the event denoting argument cannot co-occur with the complement some MP adjectives can have. in Spanish there is no such a restriction. (iii) Pedro fue muy amable conmigo al ayudarme a terminar el trabajo Pedro was very kind to-me PREP to-help-me to finish the paper (iv) Pedro was very kind (*to me) to help me finish the paper . Stowell proposes (24). the dyadic interpretation is obtained this way: the adjective is predicated of the action denoting argument and. (i) (ii) Pedro fue muy inteligente *(al) decir la verdad Pedro was very intelligent (PREP) to-tell the truth Pedro was very intelligent to tell the truth 3 Also. whereas the event denoting argument in English is conceived as an infinitive. precisely. states and achievements are excluded). whose PRO subject is understood as an agent (recall that the only predicates possible here are. The adjective takes two arguments via a double-shell structure.88 Individuals in Time (24) John was intelligent to tell the truth AP wp Event A′ | 5 A AP | | | 5 | Spec A′ | | | | | | | A 6 [e] John intelligent to tell the truth In the two monadic examples in (23). the subject of the predication is in the specifier of the adjective. which allows predicates to project additional maximal projections to accommodate all of their arguments. the performer of such an action. Compare (i) and (ii). formed by the preposition (a) (‘to’) plus the definite article. since. in Spanish it has to be introduced by the contraction al. in Spanish they are not entirely parallel to those in English. As Stowell notes. In the first place. gets also qualified as intelligent for having performed the action. based on Larson’s (1988) proposal of phrase structure. For the cases where the adjective is understood as simultaneously predicated of the individual and the action.

which is the lexical marking for SL-hood. 1995). I will make two observations. and the contrast between the examples with ser and those with estar (27) can be captured by paraphrases such as the one in (28).2). necessarily. In Spanish. this proposal would entail that in all instances of APs followed by a relational PP. that is not case. and between the dyadic one and stage-level (SL) predication. to admit there is an eventive argument that is implicit amounts to admitting that (25) is a SL predication. I take up the discussion related to the interpretation of the “infinitival argument” in Spanish and its status (as argument or adjunct). pero normalmente Juan estar-PRES-3SG very cruel to Pedro this evening but usually no es cruel con él not ser-PRES-3SG cruel to him “Juan is being very cruel to Pedro this evening. However. I repeat one of the examples here in (29).1) regarding the APs that can appear with either copula (cf.2. . Ser is completely grammatical. as (26) illustrates. (25) John was cruel to Peter Regarding Stowell’s analysis. Note that the contrast between ser and estar in this case falls in line with the contrasts studied in chapter 2 (section 2. in chapter 6 (section 6. the event argument represented by the infinitive is implicit in simpler copular cases such as the ones under investigation here. the copular verb should be. Following Kratzer (1988.3). (28)). in his account. estar. 1995) account. According to Stowell. he argues that the action denoting argument is an overt version of the (Davidsonian) eventive argument. Specifically. (26) Juan es muy cruel con Pedro Juan ser-PRES-3SG to Pedro (27) Juan está muy cruel con Pedro Juan estar-PRES-3SG to Pedro (28) Juan está muy cruel con Pedro esta tarde. see chapter 2 (section 2.4 he takes it that the presence of such an argument is proper of SL predicates. but he is not cruel to him usually” Later.1. First.2. 4 For details of Kratzer’s (1988.Aspectual Alternations in IL Predicates 89 Stowell (1991) establishes a correspondence between the monadic version of these adjectives and individual-level (IL) predication.

I will first elaborate an explanation as to what specific kind of adjectives are the ones that show an activity-agentive patterning and why. Second. the properties of copular cases such as (25) (the active-agentive characteristics) cannot be attributed to or derived from those of the implicit event argument. As shown in the previous chapter. More specific- . 4. which is proved by their acceptation of during-adverbials and the rejection of in-adverbials. since it is not the copula that varies in the minimal pairs that can be construed (be Eskimo/be cruel) and. These two facts lead me to discard this hypothesis based on the existence of implicit event arguments. Since the activity-like behavior depends on the adjectival predicate combining with the copula. we observed that there were certain copular combinations that behave like states and others that behave like activities. as a consequence. My purpose in the remainder of the chapter is twofold. the event argument Stowell proposes corresponds to an accomplishment (21) or activity (22). pero está muy (29) Pablo no es Pablo not ser-PRES-3SG handsome but estar-PRES-3SG very guapo con ese traje handsome in that suit “Pablo is not handsome. I have argued in the previous section that the activity-like behavior should not be attributed to properties of the copular verb. I will consider these sentences as instances of IL predication.90 Individuals in Time guapo. to account for nonstative IL copular clauses. but he looks very handsome in that suit” Given the grammaticality of ser.2 Adjectival Predicates Showing an Activity-like Behavior In chapter 3 (section 3. (8) above). it cannot be said it is the copula that induces the active features. as shown above. The idea I will defend here is that nonstative properties of copular clauses are due to the characteristics of the APs heading the SC taken by the copula. these nonstative copular clauses pattern with activities and not with accomplishments.3). my proposal will not be based on properties of the copular verb. Thus. I will then offer an explanation regarding the stative reading also available with such predicates (cf. since they are not the same. cruel con Pedro (*en una hora/durante (30) Juan fue Juan ser-PRET-PERF-3SG cruel to Pedro in an hour/during toda la entrevista) all the interview “Juan was cruel to Pedro (*in an hour/during the whole interview)” However.

slow Apt.Aspectual Alternations in IL Predicates 91 ally. dense White. shrewd. farsighted. blue. and as complements of force or regret). Dixon distinguishes different classes of adjectives according to the type of concept expressed. e. 4. b. recent Beautiful. That is. where a set of adjectives is conceptually more natural than others. kind. I will attribute such an active behavior to a particular syntactic structure. capable. *La mesa estaba siendo ancha/pequeña The table was being wide/small (33) *El mueble estaba siendo ligero/pesado The piece of furniture was being light/heavy (34) *La mesa estaba siendo blanca/azul/marrón/redonda/cuadrada The table was being white/blue/brown/round/squared . old. d. small Light. cunning. wide. *Juan estaba siendo alto/bajo Juan was being tall/short b. intelligent. I will argue that the APs usually presenting an active behavior are those that enter more easily (for pragmatic reasons) into the syntactic structure that gives rise to such dynamic properties. (31) a. heavy. Progressive Form (32) a. I begin by describing the type of qualifying adjectives giving rise more often to the nonstative patterning. c. cruel. stupid. given their lexical meaning.1 Lexical Semantic Classes of Adjectives I will take the basics of the classification of adjectives proposed by Dixon (1977) as a reference. brown. nice The APs of the table react differently in contexts testing out stativity versus nonstativity (such as the progressive form) as well as in scenarios diagnosing for agency (the suitability of adverbials such as on purpose.2. g. short. f. new. Dimension Physical property Color and Shape Age Evaluative Velocity Human aptitudes and dispositions Tall. horrible Quick. I will not argue that activity-like properties are due to the intrinsic lexical properties of the adjectives. squared Young. mean. Concretely. he distinguishes the following classes of semantic concepts. round.

*Juan era alto/bajo a propósito Juan was tall/short on purpose b. *Juan era joven/viejo a propósito Juan was young/old on purpose b.92 Individuals in Time (35) a. *Juan estaba siendo joven/viejo Juan was being young/old b. *La mesa era ancha/pequeña a propósito The table was wide/small on purpose (40) *El mueble era ligero/pesado a propósito The piece of furniture was light/heavy on purpose (41) *La mesa era blanca/azul/marrón/redonda/cuadrada a propósito The table was white/blue/brown/round/squared on purpose (42) a. *La noticia era nueva/reciente a propósito The piece of news was new/recent on purpose (43) *La mesa era preciosa/horrible a propósito The table was beautiful/horrible on purpose (44) Juan era rápido/lento a propósito Juan was quick/slow on purpose (45) Juanera*apto/*capaz/*inteligente/*ingenioso/cruel/amable a propósito Juan was apt/capable/intelligent/cunning/cruel/kind on purpose . *La noticia estaba siendo nueva/reciente The piece of news was being new/recent (36) *La mesa estaba siendo preciosa/horrible The table was being beautiful/horrible (37) Juan estaba siendo rápido/lento Juan was being quick/slow (38) Juan estaba siendo *apto/*capaz/inteligente/ingenioso/cruel/ amable Juan was being apt/capable/intelligent/cunning/cruel/kind Combination with on purpose (39) a.

.Aspectual Alternations in IL Predicates 93 As complements of force (46) *Juan forzó a Pedro a ser alto/bajo Juan forced Peter to be tall/short (47) *Juan forzó a Pedro a ser ligero/pesado Juan forced Peter to be light/heavy (48) *Juan forzó a Pedro a ser blanco Juan forced Peter to be white (49) *Juan forzó a Pedro a ser joven/viejo Juan forced Peter to be young/old (50) *Juan forzó a María a ser preciosa/horrible Juan forced Maria to be beautiful/horrible (51) Juan forzó a Pedro a ser rápido/lento Juan forced Peter to be quick/slow (52) Juan forzó a Pedro a ser *apto/*capaz/*inteligente/*ingenioso/ cruel/amable Juan forced Peter to be apt/capable/intelligent/cunning/cruel/kind The results of the progressive test and the suitability of the combination with on purpose and force suggest the following classification: (53) Semantic concept Dimension Physical property Color and shape Age Evaluative Velocity Aptitudes and human dispositions a. inside the group patterning with activities only a subset of them involves agentive properties. A subset of them pattern with states and the two other subsets pattern with activities. cunning c. capable b. those APs referring to velocity and those referring to aptitudes and human dispositions or mental properties (MPs). Apt. Intelligent.6 These are adjectives referring to properties that 5 Observe that although APs such as intelligent can refer to agentive events in constructions of the type discussed by Stowell (1991) (John was intelligent to close the window). this does not make them agentive. Furthermore.5. kind Inner aspect patterning States States States States States Agentive activities States Activities Agentive activities Most of them show a uniform patterning with states. More specifically. there are two groups that pattern with activities: namely. Cruel. the group of MPs manifests a variety of behavior. However.

instigators’. performers. Chichesaw chokma-LI SA-chokma (iii) (iv) 7 “I act good” “I am good” Recall that the interpretation of eventive verbs in present tense is ‘habitual’ (John drives to work. As I will amplify later in the chapter. That is. Following Stowell (1991). The point I want to make with the classification of (53) is that whereas some adjectives do not seem to easily have the possibility of behaving like activities. their interpretation as processes is easily obtained (55). and ma as a marker for ‘patients’. Central Pomo to· goes with semantic patients. it takes a different marker than when it is stative (iv). Mithun describes Lakhota prefix wa as a marker for ‘agents. however. whereas the interpretation of stative verbs can be habitual but need not. “take place”. The data are from Munro and Gordon (1982). somehow. others do because they describe properties that can have certain types of complements. The data are from Mithun (1991). 6 In Lakhota (a Sioux language) and Central Pomo (spoken in the Clear Lake in Northern California). to· mká·t’ “I am careful” “I am mean” “I am cold” “I’m surprised” In Chichesaw (a Western Muskogean language.94 Individuals in Time can be understood in relation to another individual: be cruel/kind to someone. I do not mean that each type of the adjectives of the table lexically belongs to states or to activities. and ?a· reflects semantic agency. describe the fact as no systematic. See chapter 3 for discussion. however. In fact.7 Note. as their interpretation in present tense (not necessarily ‘habitual’) proves. I do not mean that the inner aspect patterning is lexically given. the core of the following sections is dedicated to show that such complements are the source of the dynamic behavior. I consider that intelligent or cunning in examples such as (54) are states. ‘John usually drives to work’). ?a· ?eč·baya c. . that when a complement (involving nouns that can be said to. malákhota “I am prudent” “I am Sioux” (ii) Central Pomo a. who. to· kasíla d. such as business or jokes) is added. adjectives of this type morphologically show agentive markers. I will call them “relational MPs”. spoken in the American Southwest) there is also evidence that when an adjective has an agentive sense (iii). waksápa b. ?a· yá · qač’in b. Very similarly. (i) Lakhota a.

Consider the following contrasts: (56) a.e. giving pieces of the picture bit by bit. 4. those referring to velocity and a subset of the APs referring to MPs. I will start in the next section by treating the property that makes this group of adjectives a natural class. *Sacar las cosas del armario fue rápido/lento por parte de Juan “Taking the stuff out of the closet was fast/slow of Juan” (57) a. a complement of the type of those in (55) is understood. the fact that they are predicates that can be interpreted in relation to another individual. Juan fue cruel con Pablo al insultarlo en público Juan was cruel to Pablo in insulting-him in public “Juan was cruel (to Pablo) to insult him in public” b. relational ones). as is the case of MPs. I will deal with different aspects in turns. Since their analysis is complex.2 Summary of Section 4. which is an (agentive) activity. In the remainder of the work I will leave aside these adjectives and I will concentrate on the properties of MPs. Insultar a Pablo en público fue cruel por parte de Juan “Insulting Pablo in public was cruel of Juan” Juan fue rápido/lento al sacar las cosas del armario “Juan was fast/slow in taking the stuff out of the closet” b. although I will make some considerations about all the sets behaving as activities. As we can see.. Juan es inteligente en los negocios/ingenioso en sus bromas Usually. Juan is intelligent at his business/cunning at his jokes When these adjectives appear in the progressive form. we cannot say (57b) from (57a). I will be chiefly devoted to analyze the properties of the agentive MPs (i.2. namely.Aspectual Alternations in IL Predicates 95 (54) Juan es inteligente/ingenioso Juan is intelligent/cunning (55) Normalmente. they do not qualify the individual herself. I will propose that such a property is in strict . Although slow or fast also refer to the way an individual does something. Regarding the agentive-activity patterning of APs referring to velocity. They are predicates referring to ways of accomplishing a trajectory (physically or figuratively). I do not have a specific proposal to offer at this moment. Thus.2 In this section I have described the semantic class that adjectives showing an activity patterning belong to: namely. whereas out of (56a) we can assert (56b). since velocity APs work differently in many other respects.

it is understood that Juan was the agent of an action that had Pedro as its affected argument. 1988. even if we finally concluded that the PP is an “affected argument”. set on fire and bother. abuse. 4.” This captures his intuition that such a DP refers to the individual that gets affected by the (underlying) action undertaken by the subject DP. although it seems intuitively clear that the PP specifies the goal addressed by the subject’s action.3 Relational Mental Properties: The Relational PP Complement As just described. whether or not the PP expresses an “affected argument” depends on the concrete type of action undertaken by the subject. harass. we would have still to discuss the aspectual role of such a PP regarding the delimitation of the event (bearing in mind the correlation between affectedness and event delimitation pointed out by Tenny [1987. In this section. (58) Juan fue muy cruel con Pedro Juan was very cruel to Pedro The subject is understood as the agent of an action that has the DP inside the PP as the affected argument. offend or regale. Thus. 1994] and . or by acting in a certain way.3. and such an action is qualified as cruel. note that the DP inside the PP has to have a [+animate] noun: (59) Juan fue cruel con el gato/Pedro/*el armario Juan ser-PRET-3SG cruel to the cat/Pedro/the closet Although the concrete action itself is left unspecified (we do not know exactly what Juan did to Pedro). to name just a few. for example. the set of adjectives that behave as agentive activities in their combination with the copular verb are those MPs that can be understood in relation to another individual (“AP to someone”). and discuss all the consequences derivable from this. focusing on two aspects: its interpretation and its optionality or obligatoriness. Note that referring to some action without specifying what particular action is involved is quite common among verbs too: think of verbs such as humiliate. Likewise. Incidentally. can be considered as an “affected goal. 4. One can. I investigate the nature of the PP complement.1 On the Interpretation of the Relational PP Stowell (1991) suggests that the DP inside the relational PP. offend someone else either by saying something unpleasant.96 Individuals in Time correlation with their aspectual characteristics. in italics in (58). I will show that its syntactic presence correlates with the aspectual patterning of the construction.

” Tenny (1987. as the suitability of in x time adverbials show. Getting back to the relational PP of our adjectival cases. in what sense can we say that the relational PP is an “affected” goal? Although. 1988. there are internal arguments that can be considered as “affected. nor do they delimit the event. like the city in (60). where.” since they undergo a change. there is no delimitation of the event in (63). after this brief discussion. however. (62) Bill pushed the cart to New York/into the house/over the bridge (in/*for an hour) (63) John chewed/kneaded/jiggled/spun the loaf of bread (for/*in an hour) In (62). (64) *Juan fue cruel con el entrevistador en una hora Juan ser-PRET-3SG cruel to the interviewer in an hour The situation of MP relational cases resembles the situation of other verbs such as humiliate. but a PP. over the bridge) delimit the event. distinct PPs (into the house. Jackendoff (1996) challenges Tenny’s correlations and points out that it is not always an internal affected argument which delimits the event and. Both examples below are from Jackendoff (1996).Aspectual Alternations in IL Predicates 97 Pustejovsky [1988] mentioned in the previous chapter). in effect. Humiliate has direct internal arguments and. a delimited event can be tested by the suitability of in x time complements: (61) The soldiers destroyed the city in two months The correlation established by Tenny among internal argument. also. nevertheless. The first point I will address is the meaning of the notion of “affected argument. it has become clear that the correlation among internal argumentaffectedness-delimitation as conceived by Tenny is too strict. our cases are not like the ones in (62) or (63) since it is not a direct internal argument (a naked DP) the (arguably) affected one. I will deal with all this in turns. event delimitation and affectedness has been discussed by other authors. and delimits the event. (60) The soldiers destroyed the city That the event in (60) is. as the ungrammaticality of (64) shows. 1994) describes “affected argument” as the direct internal argument which undergoes some change. In turn. it is not the case that all affected internal arguments actually delimit the event. it .

2 On the Optionality of the (Affected) Goal PP 8 Now we have established the interpretive status of the relational PP. by hitting him in such an aggressive way that Pedro undergoes a change. we would not say that Pedro has undergone a change. Again. harass. humiliate (like offend. if one (Juan) humiliates someone else (Pedro) by. As Stowell (1991) also notes. it seems possible for one person to be kind or cruel to another without this actually affecting the other person in any way. that is the case in . maybe. the DP inside the PP in the relational MP cases could be argued to be “affected” or not. as an optional one. By ‘optional complement’. However. I am going to treat the PP as a “goal. can humiliate someone else (Pedro). As Bosque (1999) points out.98 Individuals in Time would be debatable whether they can be considered as affected or not. Similarly. by uttering Juan has been cruel to Pedro we are just describing a scene where Juan has ridiculed Pedro. Or. criticizing him in public. another important property of obligatory complements is that they are headed by a specific preposition. that is. the internal argument of humiliate can be considered as a licit affected argument.” which can be considered as “affected” just potentially. In this case. whether they undergo any change. I have argued that its status as an “affected argument” depends on the nature of the action actually undertaken by the subject. on the contrary.9 8 9 I want to thank Tim Stowell for all the conversations around this section. As will be specified. I mean a complement that is not necessarily present. whereas in the first case the DP inside the PP can be considered an affected argument. in the second case it is not so obvious. either phonetically overt or not. it is considered implicit. say (Juan). very similarly to the destruction of a city examples (60). depending on the action itself. (65) *Juan humilló a Pedro en una hora Juan humiliated Pedro in an hour As mentioned earlier. As seen in (65) humiliate someone is not a delimited event either. by ‘obligatory complement’ I refer to a complement that has to be syntactically present. 4.3. I will explore whether it should be taken as an obligatory complement or. One. for example. regale or bother) does not refer to any kind of action in concrete. All this suggests that we should keep a differentiation between the action referred to as humiliate or cruel and the constructions of these predicates themselves. In sum. As to relational PP complements. if it is not phonetically overt. We can be saying that one (Juan) has been cruel to someone else (Pedro) because he has hit Pedro in such a way that all his face is unrecognizable.

it is quite clear that someone has to be guilty and a supporter of something. Spanish. constructions with adjectives such as eager. In English. maybe even like a contradiction. they cannot be taken as ungrammatical as they appear. but.Aspectual Alternations in IL Predicates 99 I will start by observing the behavior of other APs that also have prepositional complements. since all of them are introduced by the preposition con. The complement is interpreted as something specific which is not necessary to be repeated for whatever reason: guilty of such and such crime. although with can appear as well. although their presence can be phonetically null and its content is recovered from the context. on the other. which is the status of the PP complements in the relational MP cases? (73) Juan es cruel/amable/mezquino Juan is cruel/kind/ mean In principle. the complements in these cases are obligatory too. In other words. distressed or prone become ungrammatical if the PP complement is not explicitly present. Now. we may judge a sentence like (74) as odd. Examples (69)–(72) are trickier. (66) Los reporteros parecen ávidos *(de noticias) Journalists seem eager (of news) (67) Parecía aquejado *(de una enfermedad crónica) He seemed distressed (from a chronic disease) (68) Era propenso *(a la gripe) He was prone (to flu) (69) Era culpable He was guilty (70) No era partidario He was not a supporter (71) Era inmune He was immune (72) El programa es compatible The program is compatible As the judgments of (66)–(68) indicate. immune to something and compatible and consistent with something. the PPs are massively headed by to. kind and mean to someone else. The following examples are from Bosque (1999). one can think that if a person is cruel or kind or mean he has to be cruel. . On the one hand. For example. supporter of such and such proposal and compatible with such other computer features.

Genericity (as well as habituality) refers to a significant proportion of iterated instances. where gustar (‘like’) and distraer (‘distract’) or ayudar (‘help’) always involves “gustar/distraer/ayudar a alguien” (‘to someone’) . I give a more detailed account of habituality in chapter 5. in general) The evidence brought up thus far suggesting that the PP is always syntactically present has consequences on temporal interpretation. from (79). en general) Juan is cruel/kind/ mean (to people. In sum. as stative. for example.10 (75) Esta película gusta mucho this movie like-PRES-INDIC-3SG a-lot “Everybody likes this movie a lot” (76) a. 10 . in principle. The null nominal in the (null) PP is interpreted as a nominal with a generic interpretation.100 Individuals in Time (74) #Juan es cruel. This way. the DP inside the PP can be considered in a similar vein as other null nominals. pero nunca ha sido cruel con nadie Juan ser-PRES-3SG cruel but never was cruel with anyone “Juan is cruel. be paraphrased as (77). is not ‘John is cruel to everyone’ or ‘Everyone likes this movie’. Genericity refers to a significant proportion of individuals and habituality to a significant proportion of occasions at which an eventuality takes place. like those in (75) and (76).78)) amounts to claiming that sentences such as (78) are habitual sentences. such as ‘like/distract/help to people’.12 The interpretation of a null nominal as generic is not ‘everyone’. then. b. crucially distinct. I conceive generity and habituality as phenomena based on quantifiers that share the semantic components of iteration and proportion. The claim that the PP complement is always present (implicit with a generic interpretation in the present tense (cf. the latter over event variables. simply. 11 Recall that as many authors have argued (Enç 1991b) habitual interpretation is not available with stative predicates. See chapter 5 for further discussion. whereas the former quantify over individual variables. (77) Juan es cruel/amable/mezquino (con la gente.11. which are taken. That is. as “phonetically” null when no overtly pronounced. but he has not been cruel to anyone” This case could lead us to think that the relational PP is present in all crueltype occurrences. the interpretation of (73) and (75). Los ruidos distraen mucho Noises distract-PRES-INDIC-3PL a-lot Este profesor siempre ayuda This professor always helps The interpretation of (73) could.

it is true that states and habituals behave the same way in certain contexts. These contrasts suggest that dynamic predicates keep as such even in the contexts they share with states. states are excluded from such contexts (v) and (vi). so that there is no crucial difference between (78) and (79). which invites us to take the parallelism between habituals and statives with a bit of caution.Aspectual Alternations in IL Predicates 101 (78) Juan es cruel (habitualmente) Juan is cruel (habitually) (79) Juan es sabio (*habitualmente) Juan is wise (habitually) In what follows I will show that the presence of the PP is optional and. whereas (74) above was considered a contradiction. as already shown. . To begin. In fact. as has been claimed in the literature. I will keep the idea that the habitual interpretation should not be considered stative in inner aspect terms. the interpretation as habitual in present tense is not the unique reading available. habituals of non state verbs keep their dynamic and agency properties in the habitual forms. which suggest that habituals and statives should not be considered to be exactly the same thing. the opposite assertion does not sound that strange: One could argue that. as a consequence. note that. While we can have habitual predicates (go walking every day) combined with volition adverbials (iii) and as complements of verbs like force (iv). (i) Juan ha dejado de tener sed Juan has given up being thirsty (ii) Juan ha dejado de pasear todos los días Juan has given up going walking every day 12 Even though habituals of nonstate verbs behave as statives in some contexts. (iii) Juan paseaba todos los días por el parque a propósito Juan used to go walking every day around the park on purpose (iv) Pedro obligó a Juan a pasear todos los días por el parque Pedro forced Juan to go walking every day around the park (v) *Juan tenía sed a propósito Juan was thirsty on purpose (vi) *Pedro obligó a Juan a tener sed Pedro forced Juan to be thirsty Bearing these caveats in mind. habituals are stative. Examples (i) and (ii) illustrate this. One context is after dejar de ‘≈give up’.

pero ha sido cruel con Pablo Juan no ser-PRES-3SG cruel but has ser-PART-3SG cruel with Pablo alguna vez some time “Juan is not cruel. SCs complement of consider can perfectly have SL predicates. Stowell (1991) was right in considering the simple versions of relational MPs as IL and the ones accompanied by the PP as SL. (81) Juan considera a Pedro cruel Juan considers Pedro cruel (82) ??Juan considera a Pedro cruel con Maria Juan considers Pedro cruel to Maria To elucidate the kind of predicates licensed in the SCs of consider is quite delicate. without any apparent contradiction. have pointed out. From these cases. However. as other authors. The first scenario where cruel cannot appear followed by a complement PP is small clause (SC) complements of epistemic verbs such as consider. at the same time. I will conclude that cruel-type APs are not inherently relational.102 Individuals in Time (80) Juan no es cruel. I will concentrate on some syntactic scenarios where an overt relational PP cannot appear. such as Demonte and Masullo (1999). (83) Me considero de vacaciones “I consider myself on vacation” (84) Los diputados consideraron acabado el debate The congressmen considered the debate finished The predicates on vacation and finished can be defended as SL predicates. but he has been cruel to Pablo some time” Example (80) shows that we can be claiming that one person has been cruel to someone else and. be claiming that the person is not cruel. and rejects SL predicates. To show that the PP is optional. as their obligatory combination in Spanish with the copula estar proves: de vacaciones (85) Estoy/*soy estar/ser-PRES-1SG on vacation “I am on vacation” . Some authors have claimed that consider selects for SCs containing IL predicates. in fact. This could in principle lead us to think that. Sentences (83) and (84) are based on examples from Demonte and Masullo (1999). In the remainder of the chapter I will leave aside these (somewhat confusing) interpretive contrasts.

since. I conclude. (87) *Considero a Pedro padre I consider Pedro a father (88) Considero a Pedro un buen padre I consider Pedro a good father However. In sum. one can agree as a matter of opinion about whether someone is cruel to someone else or not. Surely. The second piece of evidence suggesting that cruel-type APs do not always involve a PP complement comes from the examples where the DP subject is not animate. there should not be any difference between cruel by itself and “cruel + PP”. As observed in the examples below. if cruel-type APs always involved a PP complement. In the first place. that the reason why (82) (Juan considers Pedro cruel to Maria) is odd is because the predicate cruel to Maria has the properties of activities. (89) *I consider John to build houses (90) *I consider John to walk around the park (91) I consider John to know mathematics (92) I consider John to be on vacation I take the contrast between (89) and (90) versus (91) and (92) to mean that consider selects for stative predicates and rejects eventive ones. which leads to the conclusion that it is not always there. a PP involving an arbitrary nominal could be syntactically noticed. Consider now the following examples from English. where just some infinitival predicates are possible as predicates of consider SCs. for example. this does not seem to give us the reason of the contrast in (81) and (82). when the DP is inanimate the relational PP is excluded. then. . the predicate should refer to something which can be a matter of subjective opinion. Consider the contrasts between (87) and (88). such as activities or accomplishments. The contrast found in (81) and (82) suggests that the presence of the PP makes a difference.Aspectual Alternations in IL Predicates 103 acabado (86) El debate está/*es the debate estar/ser-PRES-3SG finished “The debate is finished” It seems that in the licensing of the predicates of consider two factors play a role. even in the case where it appears alone.

be cruel!” (99) As a complement of forzar ‘force’ *El director forzó a la imagen a ser cruel “The director forced the image to be cruel” . b. the cold is really cruel to the habitants Esas imágenes son crueles Those images are cruel *Esas imágenes son crueles con el espectador Those images are cruel to the spectator Ese trabajo es muy cruel That work is very cruel *Ese trabajo es muy cruel con los obreros That work is very cruel to the workers These examples indicate that the subject of cruel does not have to be [+animate]. ¡sé cruel! “Image. (95) a.104 Individuals in Time (93) a. siendo cruel (96) *Esa imagen está that image estar-PRES-INDIC-3SG ser-ing cruel “That image is being cruel” (97) *Esa imagen ha parado de ser cruel that image has stopped ser-ing cruel “That image has stopped being cruel” As can be expected. En Canadá el frío es muy cruel In Canada the cold is really cruel *En Canadá. b. Examples (93)– (95) demonstrate that the possibility of the PP to appear depends on the nature (animate/inanimate) of the DP subject. when the DP subject is inanimate. they show a correlation between the relational PP and the properties that the subject has to possess. such as the progressive or as a complement of parar de ‘stop’. the construction cannot appear in contexts proper of dynamic eventualities. as a complement of verbs like force or in combination with volition adverbials: (98) Occurrence in command Imperative *Imagen. such as the imperative form. el frío es muy cruel con los habitantes In Canada. b. (94) a. the construction cannot appear in agency scenarios either. When the subject is inanimate. and. more interestingly.

otherwise. If the PP complement is not overt. This is because they describe a property that can be understood in relation to another animate entity. which enables agency. while the subject of cruel is not an agent. but just a “theme. etc. as mentioned before. Consider (101) and (102).” if we use traditional vocabulary. not included in the group of relational MPs in the table above which. kind.3 The Relational PP with Other APs There are other MPs. conclude that the structure of cruel-type APs does not always involve a PP complement and aspectually behave as states in principle.). can also take a relational PP complement. since. it is not inferred from the context or interpreted as generic. its impossibility to appear with certain subjects would remain unaccounted for. which are not totally excluded. dynamicity. namely. as Violeta Demonte makes me notice. Summarizing. (101) ?Juan fue estúpido con el entrevistador (y no le contestó a ninguna pregunta) “Juan was very stupid to the interviewer (and did not answer any question)” (102) ?Juan fue muy inteligente con su jefe (y consiguió lo que quería) “Juan was very intelligent to his boss (and got what he wanted)” I will make two brief considerations in this respect. since it needs to bear the required properties for that (and animacy is the most basic one). although these adjectives referring to human (or animate) aptitudes can appear in combination with a relational PP. as it is possible with the other adjectives (cruel. also.3. In the first place. with particular characteristics of the construction. (103) Juan es inteligente Juan is intelligent . 4. I therefore. I will consider that the presence of the PP correlates with a subject involving particular properties and. (104) is not the interpretation of (103). In the next sections. they do not have the same relationship with the PP.Aspectual Alternations in IL Predicates 105 (100) Combination with volition adverbials *La imagen fue cruel intencionalmente “The image was cruel intentionally” I take these contrasts to suggest that the subject of “cruel to someone” is a real agent. I consider these cases with inanimate subjects as evidence that the PP complement cannot be an obligatory complement of cruel. I account for the relation between the two inner aspect forms.

in the be cruel to someone constructions. interestingly. Examples (105) and (106) show that. I first showed a syntactic scenario (relational MPs as . I have debated the correlations between affectedness (understood as ‘change of state’) and event delimitation (Tenny 1987.4 Summary of Section 4. In sum. although this complement does not maintain the same relationship with all adjectives. Whereas without the PP.106 Individuals in Time (104) #Juan es inteligente con la gente en general “Juan is intelligent to people in general’ As a second remark. they can be said to gain agentive properties. note that.3. Jackendoff 1996). Along the same line as Stowell (1991). either explicitly or covertly. I considered two facts. Second. I argued that whether the DP goal can be considered affected depends on the action that the subject actually undertakes and is qualified by the adjective. 1989. In this regard. Agentive properties are in direct correlation with the presence of the PP: for the PP to be grammatical. with the PP present they become acceptable. Interestingly. (105) *Juan fue inteligente a propósito Juan was intelligent on purpose (106) ?Juan fue inteligente con su jefe a propósito Juan was intelligent to his boss on purpose This fact argues in the same direction as the examples analyzed above with [–animate] DPs as subjects of the cruel-type. I do not believe we can talk of a real affected DP. volitional adverbials are excluded. Compare the following sentences.” although I discussed the status as “affected” that Stowell also attributes to them. whereas. 4. aspectual and thematic properties of the constructions are proved to depend on the presence of such a PP. I will consider that a relational PP can be a complement of those adjectives that express a property that can be interpreted in relation to another animate entity. Finally. The following sections elaborate on this point. when the PP is added to these adjectives. it can be understood or inferred from the context if phonetically null. With some of them.3 In this section I have been concerned with different properties of relational MPs. with other adjectives. I first discussed the interpretation of the PP complements. In copular constructions of the type of be cruel to someone. it must be overt. when the PP is added. I defended that the DP inside these PPs should be analyzed as a “goal. I studied whether the relational PP is always present. which strengths the relationship proposed between the PP and agency. the subject is understood as an agent. the subject must be able to involve agentive properties. In this respect.

Aspectual Alternations in IL Predicates 107 predicates in SCs complement of “consider”) where the plain form (the adjective by itself) and the one with the PP complement (adjective + PP) are not fully interchangeable. I will propose that we can have cruel. the thematic interpretation of the DP subject in each case (stative or dynamic) was different. I will argue that an “aspect-shift” (from state to activity) occurs when the PP is plugged into the structure and enters into the picture. In other words. whereas it must be overt to be understood with APs such as intelligent. I briefly discussed the combination of the relational PP with adjectives other than the ones considered “relational” by Stowell (1991). interestingly. I also pointed out that. which correlates with the syntactic presence of an argument (a PP). I showed that these nonagentive APs gain agentive properties as they combine with the relational PP. I then looked at cases where the subject DPs are [–animate]. but the activity behavior ensues with those APs admitting a PP relational complement. the matter is reducible to an aspectual alternation affecting the thematic properties of the arguments. I will defend that there are no inner aspect properties decided from the lexicon but it is the elements present in the syntactic structure that gives the aspectual nature of the construction. In the copular cases in question. behaving as a state. The remainder of this chapter argues that there exists the possibility that cruel projects by itself and has a structure distinct from that of “cruel + PP”. It can be inferred if it is not present with the cruel-type. I took this as additional proof that these adjectives are not inherently relational.4 The Issue: Aspectual Alternation in IL Copular Clauses In the previous section I showed IL copular clauses showing up in two different appearances: one of them stative. 4. I showed that. the relation between the AP and the PP is different. Finally. Otherwise. the issue is not very different from other contrasts noticed in the literature and men- . Concretely. and “cruel + PP”. the cited active properties emerge. a [–animate] DP subject is not compatible with the PP and it cannot be understood as an agent. Put in these terms. and the other dynamic. which strongly suggests that their properties are different. I argued that this aspectual alternation does not emerge with every type of copular predicate. whereas predicates describing properties that can be projected onto an animate entity can appear with a relational PP. I suggested that the APs that take a relational PP are those referring to MPs. on the one hand. all of which are odd with a relational PP. and the type of DPs allowed was correspondingly distinct. although I also showed that if we create a suitable scenario and add a relational PP to other adjectives. More precisely. Whereas a [+animate] DP subject is compatible with the presence of a PP complement and it is interpreted as an agent. activity properties are due to the presence of the PP. behaving as an activity. such cases would be unexpected. In particular.

). according to which aspect properties are fully encoded in the lexical items. . two different entries (stative and dynamic) for each adjective (cruel. This perspective cannot capture any of the other properties (interpretation of the subject. María tocó la verja (dos veces) María touched the fence (twice) The alternation here is of states versus activities. a good number of facts suggest that the interpretation of arguments does not depend on properties of the “verbs” but are properties of the “constructions.5 Justifying the Approach 4. syntactic presence of a PP) as correlations since. since there are syntactic and semantic facts suggesting its actual substantiation. the interpretation of the DP subject as an agent and the aspectual characterization of the eventuality as an activity provide further evidence against lexicalist approaches. by definition. I argued that such a contrast is grammatically relevant. mean. La pared tocaba la verja The wall touched the fence b. In (107) it is a PP which is exerting a crucial aspectual role in making the event an accomplishment. In the remainder of the chapter I will be concerned with such a contrast in the realm of adjectival copular clauses. it does not allude to the other elements present in the syntactic configuration. That is. we would be forced to assume that each aspectual behavior of adjectival copular cases (stative and nonstative) corresponds to different (but homophonous) lexical items. From a lexical point of view.108 Individuals in Time tioned thus far. both atelic eventualities.” Recall in this regard from chapter 3 (section 3.5. the thematic interpretation of the subject and particular aspectual properties. (107) El delantero avanzó hasta la portería The foremost advanced until the goal (108) a. kind. etc. Along the lines of Borer (2005) (among others). The pair in (108) from Borer (2005) shows an alternation stative/dynamic which correlates with the [±animate] properties of the DP subject. 4. whose contrast has been debated in the literature (see section 3. I begin by discussing which approach to event representation can make sense of the observed correlations among the syntactic presence of a constituent.4).2) the contrast between animate and inanimate DPs in pairs like (109) and (110).1 Lexicalist and Logico-Semantic Approaches The observed correlations among the presence of the PP relational complement. However.

Observe the following: (111) Martha drew a circle ((e) (drawing (e) & agent (e.13 Logical-semantic approaches. and there is a time (t). The aspectual class of verbs gets registered in the logical representation. The event holds now (at a time simultaneous with now). the theme of the event being a circle. the agent of the event is Martha. The “active” meaning of be as a verb assigning an agentive role to its subject was also mentioned in chapter 3 (section 3.Aspectual Alternations in IL Predicates 109 (109) (110) *El portazo rompió deliberadamente el cristal The bang deliberately broke the glass Juan rompió deliberadamente el cristal Juan deliberately broke the glass In essence. which applies to the event taking place at time t. which has an agent. and a theme (mathematics). (112) says: there is an event. achievements and accomplishments. Predicates are aspectually different because they have different aspectual terms. Such terms are two: Cul (for ‘culminate’) and Hold. The former corresponds to those eventualities that culminate—that is. framed in the tradition of Davidson (1967). argue for the explicit presence of the eventuality (events and states) as a variable in the logical representation of the sentence. 13 . according to which the dynamic properties found in these copular constructions were attributed to two lexical entries of be. Martha) & (theme (e. Parsons (1990) further adds extra terms. and culmination. which is an event of loving. but hold: activities and states.t)) (112) Martha loves mathematics ((e) (loving (e) & exper (e.2) in relation to the passive cases where a volition adverb can ambiguously refer either to the surface subject or to the agentive by-phrase (McConnellGinet 1982). which has an experiencer (Martha). mathematics) & Hold (e. to assume different entries (stative and dynamic) for each adjective would not be different from the proposals cited above (Partee 1977. and it has a theme. Dowty 1979). Martha) & (theme (e. distinct from the event variable itself. The latter to those that do not culminate. which is located before now. corresponding to the event type of the predicate in the logical representation. which is an event of drawing. one stative and another one active. now) The logical formula in (111) reads: there is an event. circle) & ((t) (t<now & Cul (e.

and the presence or absence of other elements in the sentence (the PP complement in our cases) that. its syntactic position and its role in the determination of the aspectual properties of the sentence. First.” Second. arguably. 1989. among others.110 Individuals in Time This approach leaves several questions unanswered. each entity corresponds to an aspectual notion. 4. In particular.5. as captured in the Event Correspondence Rule (116). which opens the possibility of establishing a direct relationship between the thematic properties of the arguments present in a sentence. As discussed earlier (see section 3. This perspective does not make any direct prediction about the correlation between the aspectual properties of a VP and the thematic properties of the DPs. Tenny (1987. in principle. authors such as Verkuyl (1972).1.2 Syntactic Approaches 4.1). Tenny (1987. both would contain the term “hold. (113) •---------------------------------------object of origin/actualization (114) •----------------------------------------• object of origin object of termination (115) ----------------------------------------• object of termination Activities Accomplishments Achievements (116) Event Correspondence Rule (Van Voorst 1988) Subject NP Direct object NP •----------------------------------------• event-object of event-object of termination origin/actualization . these authors emphasized the role of the internal object as an event delimiter. there is no obvious way to distinguish between the two constructions in point (stative and dynamic copular clauses). since.5. and van Voorst (1988). this approach does not establish any relationship between the aspectual term (Hold).2. 1989. Dowty (1979). the dynamic or stative properties.1 Event Roles.” given the fact that both are atelic and do not “culminate. As van Voorst puts it. 1994) and van Voorst (1988) elaborate both on the observation that arguments grammaticize the points that give the temporal contour of an event. 1994). correlate with the aspectual outfits of the construction. the notions of origin and termination give the characterization of events (113)–(115) and the temporal points identifying the beginning and the end correspond to physical entities (the arguments): that is. showed the relevance of arguments in the inner aspect properties of a sentence.

this means that thematic roles do not participate in determining syntactic structure. since it is an accomplishment. external and oblique (internal indirect) arguments constrain the kinds of event participants that can occupy these positions. taken from van Voorst 1988. according to the role that an argument plays in the structure of the event. their syntactic positions can be predicted. (119) Uniformity of Theta Assignment Hypothesis Identical thematic relationships between items are represented by identical structural relationships between those items at the level of D-structure. everything is mediated by inner aspect. The aspectual properties associated with internal (direct). This motivated Tenny (1987) to propose the Aspectual Interface Hypothesis (118) as an alignment principle. The entity denoted by the direct object identifies the termination of the event. Only the aspectual part of the cognitive or thematic structure is visible to the syntax. two points. corresponding to the two objects participating in the event. (120) The army officer bought a car from the old lady goal theme source (121) The old lady sold a car to the army officer source theme goal . The syntactically relevant roles are those aspectually relevant. From a broader theoretical perspective.Aspectual Alternations in IL Predicates 111 (117) is an example. (118) Aspectual Interface Hypothesis The mapping between cognitive (thematic) structure and syntactic argument structure is governed by aspectual properties. consider (120) and (121). (117) John drew object of origin the circle object of termination The entity denoted by the DP subject brings about the event. These facts led Tenny (1987) and van Voorst (1988) to propose that arguments should be described not as bearers of theta-roles but as bearers of eventroles. Since particular event-roles are played in particular syntactic positions. enunciated in (119). can be distinguished (origin and termination). To illustrate the difference between these two alignment principles. in contrast to the Uniformity of Theta Assignment Hypothesis (UTAH) of Baker (1988).

In other words. 2003). Compare (120) and (121) with (122) and (123). Travis (1994. regarding the stative/dynamic contrast in IL copular clauses. the work of authors such as Borer (1994. . but not so their syntactic positions. 1998. which I have argued are in strict correspondence. appears in the subject position. this assumption makes them similar to lexical approaches. The old lady instigates the selling in (123). and. Ritter and Rosen (1996. since. As Rosen (1999) points out. Although these event-role approaches provide accurate insights in important respects (the fact that arguments can affect the event type. van Voorst’s and Tenny’s proposals do not provide a direct way to capture the correlation between the presence of a PP and the aspectual behavior as an activity. For our present concerns. they cannot capture aspectual alternations such as the one we are interested here in a natural way. Incorporating the insights from the lexical argument-oriented approaches just reviewed. as such. (122) The army officer bought originator/actualizer (123) The old lady sold originator/actualizer a car delimiter a car delimiter from the old lady source to the army officer goal The NP the army officer instigates the event of buying a car in (122). and. the systematic correlation between object and delimitation and subject and origin). different lexical items have different ways to project. but the position of event-roles seems to be so. However. correspondingly. The car is the theme in both cases. 2000). In a sense. 2005). and Sanz (2000) elaborates on the idea that argument structure is licensed by functional syntactic structure. The position of thematic roles is not restricted. Benua and Borer (1996). although both Tenny and van Voorst explicitly acknowledge the compositionality of events. the parameters controlling the event type are assumed to be encoded in the lexicon itself: the event type is determined by the way an event maps from the lexicon. Note further that the NPs play different event roles in each one of them. Furthermore. a natural account of such compositionality cannot be derived from their lexicalbased approaches. 1998.5.2 Syntactic Structure as Event Structure. she does nothing that can be called buying in (122). However. 4.2.112 Individuals in Time In these two sentences the theta-roles are the same. In sum. these approaches would have to assume that each aspectual behavior corresponds to different adjectival projections. therefore. 2000. for them. he does nothing that can be called selling in (123). the event-role perspective predicts the mapping of different arguments to particular syntactic positions more accurately than the UTAH. occupies the subject position.

14 .Aspectual Alternations in IL Predicates 113 and functional structure is interpreted as event structure. whereas a DP headed by a numeral.17 Borer proposes that the core feature of aspectual distinctions is [±quantity]. 1989. A [+quantity] object. with telicity. 1998. 2001a.15 In what follows. That is. As mentioned in chapter 3. 2000. In other words. Borer (1994. 16 The work by Borer has appeared progressively in a number of articles. Mostly along the lines of Verkuyl (1972. 1999. I will refer to her book (Borer 2005). Thus. More accurately. which I take as a frame of reference for the remaining discussion. all the relevant grammatical processes get a unified account. It is the syntactic structure. where functional projections are in charge of Case assignment and agreement processes. see Kratzer 2004 and Kiparsky 1998. 1998. Likewise. 15 For further discussion about the relationship between case and event structrure. in turn. 2000). a verb projects an unordered number of arguments whose interpretation depends just on the functional projection they move to. 2005) defends the idea that Case-checking processes are satisfied through aspectual projections. that is. Compare: (124) John drank beer (125) John drank two beers A bare mass noun gives an atelic interpretation. I introduce some points of the work by Borer (2005). that gives the type of aspectual interpretation.16 For Borer. an object that can be marked as [–quantity] correlates with atelicity. In turn. Event-roles are not assigned by a particular head but are simply derived as an entailment from the aktionsart of the whole event. Ritter and Rosen (1996. argument structure is computed on the basis of the syntactic structure. events are not lexically specified for any type of aktionsart.4). whose core is constituted by aspectual projections. she argues that aspectual properties of a determined event All the authors cited make proposals in a similar spirit in this regard. 17 This was introduced in chapter 3 (section 3. 2001b). These works are framed in the Minimalist Program (Chomsky 1995. 2001. the quantification properties of the DP direct object make a direct difference regarding telicity. Borer (2005) elaborates on the intimate relationship between quantification and telicity or lack thereof. since it presents the latest versions of her ideas and proposals. Borer proposes that the core functional projection of the clause is Aspect. Assuming the parallel distinctions between mass/count nouns and atelic/telic verb interpretation. aspectual properties can be recast into quantity properties. 2000) argue that functional projections responsible for Case processes (Agreement projections) assign the event-roles of delimitation and initiation. This head has the property of “quantity” (AspQMAX or Quantity). a telic one. When considering that functional projections are in charge of event-roles too.14. events are not states or accomplishments from the lexicon.

and the one in (127) the structure of an activity (a nonquantity event). This amounts to saying that.114 Individuals in Time structure derive from the presence or absence of the functional node Aspect. Borer argues that. as telic). when AspQMAX is not projected. As the contrast between (126) and (127) suggests. whereas atelicity emerges in the absence of telicity. The idea. More concretely. that is. (126) Event Phrase 2 <e>E … 2 AspQ 2 two books VP g write (127) Event Phrase 2 <e>E … 2 VP g walk Borer argues that it is in the specifier of AspQ where DPs marked for quantity (quantified NPs. two books) check their quantity features. atelicity does not emerge from dedicated structure. (I ignore details of the representation that are not of interest at this moment). telicity is structurally represented.18 The tree in (126) depicts the structure of an accomplishment (a quantity event). the result is an atelic interpretation. in principle. . all verbs are inherently atelic and it is just a quantized theme that makes the telicity job. triggering the corresponding interpretive consequences. is that a verb stem has no inherent quantity. According to these authors. such as Krifka (1992) or Schein (2002). since it is what emerges in the absence of an optional and additional quantity head. Correspondingly.19 The bottom line is that the DP assigns quantity to the event. between the two 18 Other authors. the result by default is atelicity. 19 They also get their Case checked. AspQMAX can participate in a certain derivation or not. involving (positive) quantitative features: AspQMAX. as developed by Borer (2005). have argued in similar terms. It is when it meets the quantity phrase (AspQMAXor Quantity) that it gets interpreted as a quantity event (that is.

state.Aspectual Alternations in IL Predicates 115 types of homogeneous eventualities (activities and states). which denotes the presence of an originator (130). Pedro molestó a María intencionadamente Pedro bothered Maria intentionally b. this would capture the fact that adjectives are predicates of stative events. That is. 1996. since they are compatible with both of them. and activity). or others such as quickly. such as intentionally. do not make any distinction between quantity and nonquantity (dynamic) predicates. Borer argues that notions such as ‘activity. a view commonly upheld in the literature (Kratzer 1994.’ ‘state. among many others) and explicitly by Borer herself. On the other. She further argues that there are no modifiers restricted to activities. *El ruido molestó a María intencionadamente The noise bothered Maria intentionally (131) Pedro pintó la pared intencionadamente Pedro painted the room intentionally (132) La hoja se cayó rápidamente The leaf fell quickly (133) Juan hizo el examen rápidamente Juan wrote the exam quickly In sight of the absence of modifiers referring specifically to activities. who argues that adjectives occur in statives but not in eventive atelic predicates. Such dedicated stative structure preempts the lexical stem entered into the derivation from verbalization. According to Borer. Borer (2005) proposes that states emerge in the presence of specific functional structure (a sort of “stativizer”). adverbials blocking a stative interpretation. 2000. activities are the event type by default. examples such as (128)–(131) show that adverbials denoting lack of telicity are compatible not only with activities but also with states. (128) El geranio tuvo flores durante/*en una semana The plant had flowers for/in a week (129) María paseó durante/*en toda la tarde María walked for/in the whole afternoon (130) a. Note that this entails the claim that adjectives can only be stative. activities are the option by default and states are the result of some specific structure. Bennis 2004.’ ‘originator. . which denotes dynamicity but does not necessarily requires an originator (132)– (133). out of the three event types possible (quantity. Borer concludes that they are a derived notion and it is states that emerge from dedicated structure. On the one hand.’ or any other are grammatically real insofar as modifiers can make reference to them.

I gave reasons to discard a lexical and a logical-semantic approach by arguing that they cannot make any prediction regarding the relationship among the syntactic presence of arguments.116 Individuals in Time Regarding this specialized stative structure. “default. Regarding the main aspectual issue I am concerned with (the state/activity contrast observed in IL copular clauses). The next section develops the idea that dynamic properties of copular constructions (of adjectives with a relational PP complement) come from the relational PP complement itself. but it is the syntactic structure which gives the type of aspectual interpretation. More in the spirit of the seminal work by Hale and Keyser (1993). I have suggested.” and dynamic properties come induced from separate projections. as argued in section 4.3 Summary of Section 4. I assume that eventualities are not specified from the lexicon as states. which has been proved to be essential in explaining aspect properties. achievements or activities. and the aspectual properties of the construction. is that the stative status is. accomplishments. in some sense. that states are the type by default. Likewise.5 In this section I introduced the approach I will take for the analysis of the aspectual properties of the IL copular constructions. Since. .5. by creating the structure based on functional nodes. Focusing on adjectival copular cases. My discussion of inner aspect properties of IL copular sentences in Spanish takes Borer’s (2005) work as a frame of reference. differing from Borer’s idea. I also assume that heterogeneous and homogeneous eventualities are distinguished by a concrete syntactic projection (Quantity). The dynamic properties exhibited by the adjectival constructions studied here (those of the cruel-type) directly challenge this hypothesis. in the spirit of Borer (2005) and Ritter and Rosen (1996. differing from Borer. provides the skeleton where lexical items merge and their interpretation is obtained as a by product. I will discuss the claim that adjectival predicates can only be stative. Such a perspective is in itself an account of compositionality. I will propose that activity properties and the related thematic characteristics of the subject (recall it is understood as an agent) become a part of the construction by the syntactic intervention of one particular projection.3. this complement is not obligatory. I assume that it is syntax itself which. Specifically. their interpretation. and its absence correlates with the stativity of the construction. the conclusion I draw. and. 2000). 4. in particular. by the intervention of the PP complement that some adjectives can have: John was cruel to Pedro. I assume that the interpretation of arguments emerges as an entailment from the aspectual structure.

the following two properties: (a) When there is a relational PP. That is.6 An Account Based on the Relational PP Complement My proposal about the aspectual alternation observed in the copular realm can be sketched out as follows. but it correlates with the existence of the PP relational complement. the construction has characteristics proper of processes. As I intimated before. it has become clear that the account for the cruel-type constructions has to capture. As I advanced. I will discuss the syntactic treatment of the opposition state/activity. based on Hale (1984). .6. As I develop the proposal. the PP makes them a dynamic predicate. which gains dynamic properties from the preposition heading the PP denoting the DP goal. the dynamic aspectual behavior) actually resides in the preposition itself. the aspectual alternation (state/activity) is not unpredictable at all. and I will also be concerned with the aspectual treatment of the arguments expressed via a PP. as soon as they combine with a (relational) PP. Let us begin by discussing point (b) concerning the aspectual contribution of the preposition. I will address two theoretical points. I will argue that the stative version of the cruel-type APs corresponds to their most “basic” structure. 2000.3. that prepositions can be conceived of as heads encoding aspect properties.6. behave as activities. among others.) in the active cruel-type constructions is a projection (the PP) that combines with the “basic” structure of the AP.Aspectual Alternations in IL Predicates 117 4. I propose that the dynamic properties become a part of the construction through the preposition heading the PP. all of them very likely emerging from animacy itself: the DP has to be able to behave as a controller agent. 4. the codependency between the presence of the PP and the peculiar properties of the DP subject—I propose that both are arguments of the same head. Point (a) is discussed in section 4. (b) When an animate DP subject and the PP are present. 2004). Stowell (1993). and Demirdache and UribeEtxebarría (1997. To capture (b). In support of this proposal. at least. To capture (a)—that is. I argue.1 Prepositions as Aspect Encoders I have shown that cruel-type APs. I propose that the source of dynamicity and its related properties (agency etc. I will take such a correlation seriously and I will propose that the source of the aspectual properties (specifically. since their intervention in the event structure is not obvious (given that they are oblique arguments). From the description in the previous sections. the DP subject has to involve a particular set of properties.

In the previous section. In other words. Finally. he explains. the presence of a complement and particular aspectual properties are both due to the preposition. Aspectual clitics. temporal interpretation is derived from the meaning of locatives. expressing a close meaning. He notices that. In the absence of specific temporal marking. Consider (136). and also. the predicate unambiguously behaves as a dynamic predicate. in Warlpiri. the directional enclitic of central coincidence -yi. as the oddity of (136) shows.” (134) Juan fue muy amable con Pedro Juan was very kind to Pedro It is clear that such objects are not naked DPs. it introduces an important aspectual change: whenever the preposition is present. In what follows. in preparation. but they have to be inside a PP. Demirdache 1997. as an activity. As will be shown. language spoken in the southwest of British Columbia. This makes adjectives differ from verbs. prepositions are seen as carriers of aspect. based on the distinction of visibility/proximity relative to the speaker versus 20 . among others).118 Individuals in Time I will first consider the role of the preposition to make the (affected) goalcomplement syntactically available. “affected-goals. spatial oppositions appear replicated in the set of enclitic elements constituting the aspectual system. Davis. has the meaning of ‘durative’. One realm where this can be noted is its determiner system.” which can be. (Matthewson 1996. whose complements can be added directly. Hale (1984) notices a strong correlation between the spatial and the temporal system in Warlpiri (spoken in the central western part of Northern Australia). the assimilation between spatial meaning (cross-linguistically encoded through the preposition system) and temporal meaning can be observed with clarity and purity in a good number of (unrelated) languages. The mutual resemblances between prepositions and temporal content have been noted by many authors. morphologically belong to the spatial clitic system. I will then study the relation between prepositions and aspect.20 Spatial marking is used to express time relations also in Lillooet Salish. For example. I will motivate the preposition as a predicate conveying aspect information. That is. potentially. (135) *Juan was very kind Pedro (136) Juan regaled Pedro The preposition is needed to make the (affected) goal DP available. I offer a formal account for these predicates. I argued that DPs like the one in italics in (134) should be analyzed as “goal/destination-objects.

the central coincidence versus noncentral coincidence opposition is a fundamental semantic distinction that is universally present in grammatical relations. 21 . In their invisibility. out of. such as the ones in the domain of time. over. at.Aspectual Alternations in IL Predicates 119 Hale argues that the oppositions underlying the spatial system are reducible to the more basic and abstract opposition of central coincidence versus noncentral coincidence of a figure with respect to a place or ground (Talmy 1978). form the basis of the aspectual system of many languages. in. or through. onto or into. from. kel7ßqsten-s-a ti-United States-a] sÚcsec [ni chief-3SG. along. Examples of noncentral coincidence are from. not visible) president IS a fool kel7ßqsten-s-a ti-United States-a (iii) sÚcsec [ti PRES-DET chief-3SG. at.POSS-DET DET-USA-DET fool “The chief of the USA is a fool” (i) Hale (1984) proposes that the central/noncentral coincidence distinction is also present in the semantics of complementizers. out of. such as those expressing temporal relations (when and while clauses) and those expressing circumstance or condition (conditionals with if or comparisons with whereas). The drawings below illustrate these two relations: the white circle stands for the figure and the black square for the ground. (137) Central coincidence PP 2 figure P 2 in ground (138) Noncentral coincidence PP 2 figure P 2 to ground ◙ ○■ According to Hale. *The (past. whereas proximity or presence in place (iii) correlates with the present tense. (ii)).POSS-DET DET-USA -DET fool absent-DET “The chief of the USA was a fool” (ii) a. Perkins. Examples of prepositions of central coincidence are on. to. not visible) president WAS a fool b.21 Along the same lines. over. to. The (past. and Pagliuca (1994) argue that prepositions such as those just mentioned (on. along. Bybee. into). in. Invisibility or absence in the utterance place (i) correlates with a past interpretation (cf.

they show that prepositions of central coincidence are a very common source for progressive aspect. 2000) point out that the progressive form in Basque is construed by combining the verb ari ‘engage’ with the nominalized form of the verb suffixed with the locative postposition -n 22 Although I am focusing on the aspectual value of prepositions. “sleep”/“spend the night” (Bemba [Bantu]). (141) A: ¿Has hecho la cama? Have you made your bed? B: No. The purity of the temporal meaning emerges when the locative sense is lost. and Pagliuca (1994). where the auxiliary is the locative copula estar. As can be appreciated. “sit” (Siluyana [Bantu]. these authors also note the important role of verbs as aspectual sources. They suggest that locative prepositions were first used to indicate that the subject was involved in an activity at a certain location.22 For example. In Spanish. the English locative preposition on or at (shortened to a-. According to Vlach (1981) and Bybee.120 Individuals in Time historic and cross-linguistic study. Perkins. this preposition in combination with a gerund noun phrase is the origin of such an aspectual form. as in asleep) is considered the historical antecedent for the English progressive. Juba Arabic). and spatial and temporal locations conflated. “be there” (Krio). Sentence (i) illustrates a pure locative context of estar. estoy en ello No. the auxiliary in progressives is a verb with a locative component. Such an idea is nicely illustrated by the progressive forms in Spanish. . whereas those of noncentral coincidence are on the basis of prospective aspect. “be with” (Swahili). “live/reside” (Hindi). “lie down/stand”. I am at it Demirdache and Uribe-Etxebarría (1997. (ii) is an example of progressive usage. The following examples are from these two works respectively: (139) John is on/at hunting/building a house (140) He was a. María está en casa Maria is at home (ii) María está estudiando Maria is studying (i) This situation is found across languages.coming home Other pieces of (synchronic) evidence argue in the same direction. the use of locative (central coincidence) prepositions to express ongoing activities can be nicely illustrated with examples such as (141B). Observe the verbs used as progressive-durative markers. The data are from Givón (1982): “stay” (Hawaii-Creole).

similar cases are found in Romance languages. In turn. in Spanish and English. such as from. appear in the form used to express close future (going to).23 Prepositions like the Italian per (‘for’. For example. where “come from” can be used with the aspectual meaning of the perfect. the combination of a motion verb with an elative component and the preposition from yields a perfect interpretation: (146) Max COME FROM home ‘Max has come from home’ As Demirdache and Uribe-Etxebarría (1999) notice. the prepositions a (142). in Nigerian Margi and Palaung (from the AustroAsiatic family of languages). in Dutch. in contrast to stative locative verbs participating in progressive forms. Also. prepositions denoting ‘movement from’ (centrifugal movement). 23 Note also the presence of a verb denoting movement. are present in perfect aspectual forms in some languages. the locative preposition at directly combines with the verb in infinitival form to construe the progressive. For example. ‘to’) (144) or the English about to (145) are constituent parts of prospective forms. to (143). 2000. they notice that. (142) (143) (144) (145) Voy a hacer la cama I am going to make the bed Sto per uscire I am about to leave According to Givón (1982). Demirdache and UribeEtxebarría (1997. 2004) conclude that Aspect can be analyzed as a head establishing relations between two terms. as prepositions do. (147) French Je viens d’être malade I come from be sick “I have been sick” (148) Spanish Vengo de matricularme en la Universidad come-I from register-me in the university “I have just registered at the university” Since prepositions express aspect relations. indicating movement toward (centripetal movement).Aspectual Alternations in IL Predicates 121 ‘(with)in/on/at’. . prepositions of noncentral coincidence are present in the auxiliaries expressing perfect and prospective aspect across languages. Specifically.

2004) argue that Aspect can be conceived as a head establishing a temporal ordering relationship between two times. he characterizes tense distinctions as relations of priority. Demirdache and Uribe-Etxebarría argue that Aspect can be conceived in parallel terms. simultaneity and posteriority describable in terms of central/noncentral coincidence between the speech time and the Reference Time and aspect distinctions in terms of central/noncentral coincidence between the Reference Time and the Event Time. Following Reichenbach (1947). 2000) proposal inspired in Hale 198424 and Stowell 1993. In (149b) the TT is located “after” the span of time the event of cradling takes place. Prospective AspP 2 t2 Asp′ t1 (within) t2 Asp′ t1 (after) t2 Asp′ t1 (before) 2 2 2 Asp Asp Asp 24 Hale (1984) proposes to analyze aspect and tense forms as central/noncentral coincidence relations. Perfect AspP 2 c. . when he entered the room. Maria had cradled the baby [----------------]///////// c.” The discontinuous line represents the event and the slashes the TT. When I entered the room. Progressive AspP 2 b.122 Individuals in Time Demirdache and Uribe-Etxebarría take as a point of departure Stowell’s (1993. the kind of the relations being either of central coincidence (progressive aspect) or noncentral coincidence (perfect and prospective aspect). which establish a temporal ordering between time-denoting arguments. 1996) proposal that the semantics of tenses can be described in analogy to prepositions like before or after. When I entered the room. When I entered the room. and in (149c). In (149). Demirdache and Uribe-Etxebarría (1997. (150) a. This time is the Topic Time (TT). Maria was cradling the baby [----------////////------------b. Some clarifying examples appear below. It is conceived as “contained” in the span of time the event of cradling takes place in (149a). Based on Klein’s (1994. the time the speaker is talking about is a concrete time. the TT is captured “before. 1995) idea that Aspect orders the event time with respect to the time the speaker refers to in the sentence (the “Topic Time” or “Assertion Time”). namely. (149) a. Maria was going to cradle the baby //////////[------------The trees below schematize Demirdache and Uribe-Etxebarría’s (1997. 2000.

6. denoting the goal. I propose that this is the case. and. as seen in the examples above. con-PPs can also appear with motion verbs. I am going to argue that the aspectual shift from state to activity found in the cruel-type constructions is due to the presence of the PP relational complement.Aspectual Alternations in IL Predicates 123 Assuming that aspectual meaning is reducible to prepositions (and. conversely. and assuming with Demirdache and Uribe-Etxebarría that all temporal relations can be reduced to the same set of primitives. in these cases the preposition con is in complementary distribution with the directional preposition a ‘to’ when the DP complement is [– animate] and is interpreted as a location: . I am going to argue that the preposition introducing the (affected) goal carries inner aspectual properties affecting the aspectual nature of the whole construction. Preposition con establishes a central coincidence relation between the figure and the ground (in the terms introduced before). con ‘with’. This preposition is associated with stative locational meanings.2. the preposition con. ¡corre con tu papa! Pablito. as its adequate combination with location verbs shows: (151) Juan permaneció/se quedó con María Juan remained/stayed with María con Pablo (152) Pedro está Pedro estar-PRES-3SG with Pablo “Pedro is with Pablo” However. 4. that prepositions have aspectual meaning). run with your daddy! (154) Me volví con mis padres I went-back with my parents “I went back to my parent’s place” Importantly.6. The preposition introducing the PP in Spanish is.1 Con: A Locative and a Directional Preposition. 4. paying special attention to the APs I am taking care of in this chapter. like in examples such as the following: (153) Pablito. it is conceivable that prepositions can also convey inner aspect properties.2 The Preposition Introducing the “(Affected) Goal”: An Activity Inductor In this section I am going to describe the preposition heading the PP that refers to the (affected) goal in cruel-type construals in Spanish.

in previous periods of Spanish. Most were locative predicates that can also have a directional meaning. the prepositions introducing the relational complement in cruel-type cases were of same characteristics as con. S. the complement of adjectives of the type of cruel appears introduced by prepositions such as a (‘to’). contra (‘against’). XIII] (160) Paulo orosio (…) que era muy cruel sobre los ciudadanos Paulo orosio (…) who was very cruel over the citizens [Alfonso X: General Estoria IV. (157) Con + [+animate] DPs A + [–animate] DPs Therefore. . he was very cruel against whom did not deserve it [Alfonso X: General Estoria V. con (‘with’) and a (‘to’) can be said to share properties and can be both analyzed as prepositions denoting a trajectory with a destination. XIII] (159) (…)& guardandosse segunt su aluedrio fue muy cruel contra los que lo non merecían & keeping himself according to his will. or hurt them… [Alfonso X : Siete partidas. the only difference between con ‘with’ and a ‘to’ is the nature of the DP acting as their complements. as a path with a goal (a Goal Path. as (157) summarizes.25 As reported in the corpus. XIII] 25 I would like to thank Isabel Pérez for bringing this source into my knowledge. All this suggests that con is not (only) associated with stative locational meanings but can be analyzed as a predicate associated with directed motion— in particular. It is interesting to note that. Consider first the following examples taken from the corpus of Spanish from the Illinois State University (Davies 1999). in previous centuries.124 Individuals in Time (155) ¡Corre a la farmacia antes de que cierren! Run to the pharmacy before they close! (156) Volví a la tienda a devolver los zapatos I went back to the store to return the shoes Thus. Svenonius 2004). with a goal. S. S. and sobre ‘over’: (158) (…) Otrosi dizimos que si algund onbre fuese tan cruel a sus sieruos que los matase de fanbre: o les firiese mal (…) And we also say that if some man were so cruel to his servants that he killed them hungry.

Aspectual Alternations in IL Predicates

125

Crucially for our purposes here, prepositions a (‘to’), contra (‘against’), and sobre (‘over’) have both a situational import (they create Place PPs) and a directional one (they head GoalPath PPs, in our terms). Consider the following pairs, where (a) examples show the locative meaning and (b) examples the directional one. (161) a. Se quedó al otro lado del país He remained at the other side of the country b. Emigró al otro lado del país He emigrated to the other side of the country (162) a. Puso la escalera contra la pared He put the ladder against the wall b. Dirigieron el misil contra un campamento de refugiados They directed the missile against a refugee’s camp (163) a. Dejó el libro sobre la mesa She left the book over the table b. El alud se precipitó sobre los excursionistas The avalanche cast down over the excursionists To account for the directional meaning of these prepositions, Bosque (1996)26 argues that these cases involve a null directional preposition heading the PP.27 This null preposition, with the semantic features proper of directional heads like a (‘to’) takes the PP headed by the situational preposition. Due to a process of syntactic incorporation at Logical Form, the directional interpretation is obtained. (164) GoalPathP 3 GoalPath PlaceP (∅) 3 Place DP (over/against)

I would like to thank Luis Sáez for bringing this paper to my attention. Bosque (1996) notices that, although the concepts of situation, origin and destination look analog semantic categories, they do not share all the properties among themselves. Surprisingly, as the examples show, destination predicates pattern with those denoting situation instead of with those denoting origin.
27

26

126

Individuals in Time

Bosque (1996) shows that the patterning regarding the combination of prepositions is that where the situation P is taken by the P denoting origin (Cogí un libro de sobre la mesa ‘I took a book from over the table’) or direction (examples just cited). Thus far, I have shown that, across centuries, the prepositions introducing the relational complement with MPs of the cruel-type in Spanish are those that show an ambiguity between a locative and a directional reading (a, sobre, contra, con). Given the interpretation of these relational MPs followed by a PP, according to which the DP complement of the preposition is understood as the ‘goal of somebody’s actions’, it is reasonable to think that the meaning of con ‘with’ in these cases is directional (i.e., it is a GoalPath P), rather than situational (a Place P). The application of the analysis that Bosque (1996) suggests for the PPs headed by these prepositions to the cruel-type cases we are interested in, would give a structure like (164) above, with a null directional preposition into which con would incorporate in syntax. 4.6.2.2 Con in Relational MPs. In Spanish, the goal in relational MPs was also introduced by another directional preposition, para (‘for’), in previous centuries. In modern Spanish, para is not used in this context (166). (165) (…) porque despedaça sus hijos y es cruel para ellos (…) because she tears to pieces her children and is cruel for them [Vocabulario eclesial; S. XV] (166) *Juan es cruel para María28 Juan is cruel for María Para ‘for’ is a directional preposition denoting movement toward. It denotes a trajectory with orientation of the type of hacia ‘toward’ (Jackendoff 1976, Morimoto 2001).29 (167) Salgo para Madrid mañana de madrugada I leave for Madrid tomorrow early in the morning (168) Lo pusieron mirando para la pared They put him staring for the wall

28 This example is possible under the reading ‘According to María, Juan is cruel’, which is not of interest here. 29 For and toward in English and para and hacia in Spanish have minimal differences between each other. Jackendoff 1976 conceives for as an intentional directional preposition expressing ‘mental direction’. See Jackendoff 1976 and Morera 1988 for details.

Aspectual Alternations in IL Predicates

127

Regarding para in combination with relational MPs, it is important to note, at least, two things. First, para appears in combination with con not only in old Spanish (169), but is also a possibility in modern Spanish, specially with nouns derived from this kind of adjectives (170): (169) (…) ¿por qué serás tan cruel para contigo y tan enemigo de ti mismo “why will you be so cruel for with-you and so enemy for yourself?” [Luis de Granada (1567), Guía de pecadores] (170) la crueldad de Juan para con María the cruelty of Juan for with María “John’s cruelty to María” Second, para alternates with con (in current Spanish) depending on the [±animate] nature of the DP subject and the DP taken by the preposition30: (171) Juan es bueno {con/*para} María Juan is good {with/for} María “Juan is good to María” (172) Las zanahorias son buenas {*con/para} la vista the carrots are good {with/for} the sight “Carrots are good for the sight”31 Again, preposition con forms minimal pairs with unambiguously directional prepositions. Also, the fact that when the subject is a [–animate] DP the preposition has to be para confirms the idea of con as an agentive predicate. The fact that para and con can co-occur could suggest that each one of these prepositions specifies a particular subevent. Actually, this structure could be considered parallel to others where PathPPs take PlacePPs. The co-occurrence of the two prepositions could be described as the overt instantiation of a complex structure, in the spirit of the ones proposed by other authors such as Higginbotham (2000), Svenonius (2004) or Folli and Ramchand (2005)32. These authors suggest that PPs headed by prepositions such as to in examples like (173) below are complex functional structures containing both a direction and a final location.

I owe this insight to Olga Fernández Soriano. Note that the Spanish opposition between con and para is parallel to the one existing between to and for in English. 32 See also Koopman (2000), Tungseth (2005).
31

30

128

Individuals in Time

(173) John ran to the store GoalPathP 3 GoalPath PlaceP (to) 3 Place DP (∅)

Thus, to, despite being a morphologically simple preposition is subeventually complex. English prepositions like into or onto show their complexity also morphologically. In these cases, in/on incorporate into the preposition to in morphology: (174) GoalPathP 3 GoalPath PlaceP (to) 3 Place DP (in/on)

As proposed by these authors, path prepositions and place prepositions enter two different event projections. While path PPs would correspond with the subevent head denoting the ‘process’, place PPs, the final location, would constitute the ‘resultative’ part of the event, which gives a telic interpretation as a result (cf. the correctness of John ran to the store in five minutes). However, some clarification regarding the parallelism with our examples is in order. In the first place, as I have largely shown in previous chapters, relational MP constructions lack a “telos,” which would derive from the presence of a PlaceP understood as a resultative phrase. In the second place, the meaning of the examined examples does not change at all either if para appears in combination with con, or if it does not. Para is completely optional (175). (175) su crueldad (para) con María his cruelty for with María This makes these cases different from those where the presence of two prepositions (one heading a GoalPathP and another heading a PlaceP) can be at-

Aspectual Alternations in IL Predicates

129

tested, as in the following English pair, where the meaning differs, hinging on the overt instantiation of the stative preposition: (176) Bill walked to the room (177) Bill walked into the room To recapitulate, it seems that the aspectual properties exhibited by cruel + PP constructions are those associated with the GoalPathP. In this respect, recall examples from other languages such as English, where the preposition introducing the PP complement in cruel-type cases is the directional to. The preposition combination para + con behaves the same way as other Spanish preposition complexes such as por + entre ‘along + among’. This preposition complex, formed by a directional preposition denoting extension followed by a situational one, yields atelic constructions, where a culmination is lacking: (178) Juan paseó por entre los árboles (*en/durante veinte minutos) Juan walked along among the trees (in/for twenty minutes) Zwarts (2005, 2006) argues that the unboundedness of examples containing prepositions like along (Spanish por) directly derives from the properties of the preposition introducing the directional expression. This author argues that prepositions of the type of along describe a path denotation which is unbounded (we could say homogeneous, in the terms used in this work), as the sum of such paths show. If two paths are along the river, then, their concatenation is also along the river, because the property of having all points beside the river is preserved under concatenation (sum). In contrast, other directional prepositions, such as the English over in over the line, do not have this property, since a concatenation of two paths that go over the line yields a path that goes over the line twice (that is, there are two points of the path that are on the line). The same can be said of para (‘for’), the directional preposition with the meaning of ‘toward’ involved in the PP complements of the cruel-type constructions. If two paths are toward a place, their concatenation is also toward such a place. 4.6.2.3 Summary of Section 4.6.2. In these sections, I have described the preposition that in current Spanish introduces the goal of MPs such as cruel or nice, the preposition con. I have shown that this preposition is a central coincidence predicate which also shows dynamic (directional) meaning. In this sense, I have shown that the prepositions heading the relational PP complement were of this same type in previous periods of the history of Spanish.

Also. para + con. When it is phonetically null. we can account for the event properties from the properties of the path itself. The structure I thus propose for cruel-cases goes along the lines of (180). it can be assumed that preposition con incorporates into its position. a semantic function such as SHAPE. a directional preposition with the meaning of ‘toward’ (cf. I argue that it is due to the properties of the path described by this preposition that the event is unbounded. it semantically entails that the destination has not been reached. assuming with Zwarts (2005. (179) GoalPathP 3 GoalPath PlaceP para/(∅) 3 Place DP (con) I propose that it is due to the dynamic properties of this preposition with the meaning of ‘toward’ that cruel-type constructions in the presence of the relational PP complement pattern with activities (instead of with states) in all the tests previously shown. As a noncentral preposition. As suggested above. para-paths are homogeneous.130 Individuals in Time Based on the existence in old and modern Spanish of a combination of prepositions to introduce the MP goal. which explains the directional meaning it acquires. In this sense. which in the aspectual realm translates into an unbounded eventuality. since the concatenation or sum of two para-paths is also a para-path. 2006). Morimoto 2001)—be it overt or not. I propose that this dynamic preposition is present in all relational MP cases. where the GoalPathPP involves a null directional preposition of the type of para. I have argued that the PP has the properties of the preposition para (‘for’)—that is. . which maps paths onto events. Jackendoff 1976.

(182) summarizes the description of the preposition. I would like to argue that in the inner aspect realm directional (noncentral) prepositions can be associated with dynamic eventualities where a culminating point has not been . …● ⌂ In section 4. associated with ‘future’ in the realm of Tense and with ‘prospective’ in the domain of outer aspect. that is.1. it locates the figure in its trajectory toward the ground. The two terms are put under a specific relationship according to the meaning of the preposition. it designates the movement of a figure toward a place or ground. (The bullet stands for the figure and the little house for the ground. the discontinuous dotted line for the trajectory. and (183) represents it graphically.6. I consider that the relation between the DP subject Juan and the goal Pedro is not very different from the one existing between Juan and Pedro in (181). therefore.Aspectual Alternations in IL Predicates 131 (180) Juan fue cruel con Pedro ser 3 ser cruel 3 cruel GoalPathP 3 Juan GoalPathP 3 GoalPath PlaceP (para-Ø) 3 Place DP (con) → dynamicity Bearing in mind this discussion. 2004) proposed that centripetal noncentral prepositions translate as “before” in the temporal domain. As a directional preposition. 2000. the DP subject can be considered as the figure and the DP inside the PP as the ground.) (182) (183) TOWARD: preposition that establishes a centripetal noncentral coincidence relation between the figure and the ground. (181) Juan went to Pedro In the terms of figure and ground presented before. The two of them are related by a preposition meaning ‘toward’. we saw that Demirdache and Uribe-Etxebarría (1997. They are.

consider the temporal contribution that motion verbs make in Lillooet Salish.) shows motion verbs in the mentioned language. an incomplete process.e. The following table (Davis. then.). both refer to an ongoing process. they gain dynamic and agentive properties. Since the figure is captured “before” the ground. but not in past. by their lexical meaning. Due to this reason. illustrate this point. (i) Motion toward speaker Motion away from speaker Destination not reached ts7as Nas Destination reached t’iq Tsicw 33 Verbs ts7as and nas share the property of referring to “movement” whose destination has not been reached. which translates as ‘nonaccomplishment’. In principle. in inner aspect terms.3. there are other adjectives (stupid. which argues in favor of the hypothesis that the preposition heading the complement is an actual dynamicity inductor. Because of the same reasoning. in other words. in prep.132 Individuals in Time reached (“the destination has not been reached”). indicating a reached destination (i. are interpreted in present tense (or future). admit a relational PP complement. sentences containing this verb and lacking an explicit temporal marker. to some extent. although. (ii) Ts7áswit ets7á Sát’a lhláku7 Lh7úsa “They are coming here to Sat’ from over there at Lh7us” (iii) T’iqwit ets7á Sát’a lhláku7 Lh7úsa “They came to Sat’ from over there at Lh7us” (iv) Náswit áku7 Lh7úsa lhelts7á Sát’a “They are going over there to Lh7us from here at Sat” (v) Tsícwwit áku7 Lh7úsa lhelts7á Sát’a “They went over there to Lh7us from here at Sat’ . shrewd) that.3. as mentioned in section 4.. we can say that such prepositions convey also the meaning of ‘before’ in the inner aspect domain. That is. when these adjectival predicates combine with the PP. cunning. (184) Juan fue estúpido/ingenioso (*a propósito) Juan was stupid/ cunning (on purpose) (185) ?Juan fue estúpido/ingenioso con su editor (a propósito) Juan was stupid/ cunning to his editor (on purpose) This contrast shows that it is by virtue of the conjunction of the adjective and the PP in the syntactic configuration that the inner aspect nature of the In a similar vein. The examples (ii)–(v) from Davis (in prep. Interestingly. no process has been fulfilled33. sentences with t’iq or tsicw. Finally. but merely as a ‘process’ or ‘activity’. the PP is not contextually inferred if it is not overtly present. a concluded process) are interpreted as past tense sentences. it cannot be said it has reached its destination.

2. it suggests that the properties of the noun in the DP subject and the PP stand in a codependent relationship. as a complement of verbs like force or in combination with volition adverbials: (190) Occurrence in command imperative Juan/*imagen.6. when the DP subject is inanimate. As shown in section 4. ¡sé cruel! “Juan/image. (188) Juan/*esa imagen está siendo cruel Juan/that image estar-PRES-INDIC-3SG ser-ing cruel (189) Juan/*esa imagen ha parado de ser cruel Juan/that image has stopped ser-ing cruel Likewise.3 The DP Subject of Relational MPs As I mentioned in the beginning of the section. it suggests that the PP is not a plain complement of the adjective. be cruel!” (191) As a complement of force El director forzó a Juan/*la imagen a ser cruel “The director forced Juan/the image to be cruel” . This fact suggests two things. and we could not explain why it cannot appear in sentences like (186). And second. the construction cannot appear in agency scenarios either. I repeat the contrasts below. If that were the case.3. eventuality properties are obtained from the very syntactic structure. such as the progressive or as a complement of parar de ‘stop’. First. the construction cannot appear in contexts proper of dynamic eventualities. In sum. when the subject is inanimate. Consider the following contrast as a reminder (186) Esas imágenes son muy crueles (*con los espectadores) Those images are very cruel (to the spectators) (187) Juan es muy cruel (con Pedro) Juan is very cruel (to Pedro) As we already know. the PP complement would be possible independently from the properties of the DP subject.Aspectual Alternations in IL Predicates 133 construction is decided. such as the imperative form. when adjectives such as cruel are predicated of an inanimate subject. 4. the relational PP complement cannot appear. one important property we want the structure proposed for the cruel-type captures is the correlation between the presence of the PP complement and the properties of the DP subject.

which. makes it be interpreted as an agent. In this vein. while the subject of cruel is not an agent. but just a “theme. which I have justified above as an aspect head. but the stronger configurational perspective (Hale and . This way. the DP is the subject of a movement predicate. In other words. I argue that the DP subject gets interpreted as an ‘initiator’ simply by virtue of the dynamic properties of the head on whose specifier it is generated. a process event or a subevent which is qualified as cruel: (193) cruel 2 cruel Juan PP 2 P 2 to Pedro I will not take the view that the dynamic preposition assigns a certain interpretation to the DP. the syntactic realization of an agent and that of an experiencer or goal comes together.” to use the traditional vocabulary. the structure I propose is the one already introduced above. located in the preposition. One way to capture the codependency between the properties of the DP subject and the presence of the PP is to make them co-arguments of the same head.134 Individuals in Time (192) Combination with volition adverbials Juan/*la imagen fue cruel intencionalmente “Juan/the image was cruel intentionally” Above. Just by virtue of the fact of being located in the specifier of a movement predicate. Therefore. where the PP stands for a process predicate. since it needs to bear the required properties (and animacy is the most basic one. I would like to argue that the interpretive properties of the DP subject are naturally derived from the aspectual characteristics of the construction. as an ‘agent’ just by virtue of the fact of being an animate figure of a dynamic projection semantically conveying ‘movement’. together with its animacy properties. The DP subject is interpreted as an ‘initiator’ and an “undergoer” of the process. Since I have shown that relational MP cases are aspectually dynamic (due to the presence of a head inducing such dynamicity). as I mentioned). I consider that the element that triggers the aspectual peculiarities of these copular constructions is the preposition heading the PP itself. The properties of the DP subject come as a simple entailment from their interpretation as an animate figure in a “trajectory”. As I have argued. I took these contrasts to suggest that the subject of “cruel to someone” is a real agent.

6 In this section I analyzed the state/activity alternation in IL copular clauses.4 Summary of Section 4. 4. Borer 2005) that “thematic roles” are not “assigned”. in very simple terms. the DP in its specifier is understood as the figure moving toward a goal. the DP moves to the specifier of Tense. the idea that the DP is. a ‘goal’. the interpretation that can be legitimately attributed to the DP inside the PP is. the property (cruel) and the trajectory described by the PP. it moves to the specifier of cruel. but it correlates with the possibility of the adjective of being able to have a relational PP complement. I have argued that this aspectual alternation is not unpredictable at all. This plus the fact that the DPs subject of these constructions must be animate (cf. but emerge as an entailment from the aktionsart of the event. This proposal captures. Finally.Aspectual Alternations in IL Predicates 135 Keyser 1993. (195) Juani T TP 2 T 2 ser 2 ser cruel (= SC) 2 ti cruel 2 cruel ti PP 2 P 2 P Pedro The DP subject is generated in the specifier of the PP. the subject of two predicates.6. in a sense. (194) Juan/el león/*el frío fue cruel con Pedro Juan/the lion/the cold was cruel to Pedro A more detailed structure appears below. simply. provided that the precise status (either affected or not) of the DP inside the relational PP is left in the dark. In particular. I have proposed that the source of the aspectual . because the preposition conveys dynamicity content and the PP can be considered as a predicational structure with the properties of activities. Thus. From there. In a nutshell. 194) makes the DP in the specifier of the PP be interpreted as an agent.

which in the inner aspectual realm translates into an un-bounded eventuality. Consider the following examples from Spanish. I have argued that inner aspect properties can also be reduced to the same set of primitives. Regarding the interpretive properties of the DP subject (which is understood as an agent). they semantically entail that the destination has not been reached. the dynamic aspectual behavior) actually resides in the preposition heading the complement itself. which relates a figure (the DP subject) and a goal (the DP inside the PP). Based on Zwarts (2006). Specifically. Stowell (1993) and Demirdache and Uribe-Etxebarría (1997. I have argued they ensue as an entailment from its configurational position (the specifier of the dynamic preposition) plus its properties of animacy. such as to in English and a (‘to’). I have argued that the inner aspect properties of relational MP constructions can be derived from the properties of the path described by the preposition introducing the relational complement. the “goal. I have proposed that. sobre ‘over’ and contra (‘against’) in old Spanish.136 Individuals in Time properties (specifically. prepositions of the type of “toward” have been shown to be “homogeneous” (toward a place + toward a place = toward a place).” of somebody’s actions. as directional prepositions. which gains the activity-like properties from the preposition heading the PP denoting the DP goal when the PP complement is “plugged” into the structure. ‘to’). All this reasoning has led me to conclude that the stative version of adjectives accepting relational complements corresponds to their most basic structure. this analysis suggests that states do not need a dedicated The aspectual alternation observed in the cruel-type constructions is very much the same as the one that can be argued there to be in the formation of deadjectival verbs in languages such as Spanish. The PP complements of relational MPs denote the destination. Specifically. I propose that the process properties they show also come from the preposition participating in its constitution. 2000. 34 . describable all in terms of central/noncentral coincidence between a figure and a ground. The prepositions involved denote the orientation of a trajectory without alluding to the end points. I have shown that the relational complement is introduced by prepositions with a clearly directional meaning. 2004) that prepositions can be conceived as heads encoding tense and aspect properties. Taking as a point of departure the proposals by Hale (1984). para (‘for’. I have described the preposition heading the complement as a preposition of centripetal noncentral coincidence with the meaning of ‘toward’.34 That is. I have argued that the directional meaning of the preposition currently used (con) is accounted for by a syntactic process whereby it incorporates into a directional preposition which appears overtly in old Spanish as well as with nouns derived from the adjectives under discussion in modern Spanish (para).

the paraphrases of both being ‘to’ (The Latin preposition in could be followed by a nominal either in ablative case. which renders parallel deadjectival verbs and cruel-type cases. The structure I have proposed for deadjectival verbs exploit the participation of a preposition to account for their dynamic properties. I suggest that such dynamic properties come from the semantics conveyed by the prepositions that participate in their formation. If followed by a nominal in ablative case. the preposition had a locative (stative) meaning (English ‘in’). Etymologically. red) and refer to dynamic processes. That is. Both cases are similar regarding the elements that participate in them: a preposition. it had a directional meaning (‘to’). 4. contra Borer’s (2005) idea that statives are due to the presence of specialized functional structure (a sort of a “stativizer”). since it is by virtue of a preposition that an adjectival construction becomes dynamic. or in accusative case. whereas if followed by an accusative. black. As surveyed in the beginning of the (i) El sastre alargó la falda The tailor lengthened the skirt (ii) La tela ennegreció the cloth blackened “The cloth turned black” (iii) La cara de María enrojeció the face of Maria reddened “Maria’s face turned red” These verbs are formed on the basis of an adjective (long.7 The Cruel-type as a Small Clause Integrant The account I have developed thus far differentiates from previous explanations according to which the peculiar properties (which I have described as an activity-aspectual behavior) displayed by predicates of the type of cruel were due to the properties of the copular verb. I propose that it is by virtue of the combination of the adjective with the temporal content contributed by the preposition that the structure gains dynamic features: (iv) PP 2 P la tela 2 P A en negro This would explain the aspectual properties of deadjectival verbs. they come from the Latin “in” and “ad”. (v) en–negre–c–(er) ? ? ? ? P A v (infinitive-ending) ? ? ≈? con cruel ser . The prepositions participating in the cases above are en (en-negreció) and a (a-largó). a `centripetal noncentral coincidence preposition. an adjective and a verbal piece.Aspectual Alternations in IL Predicates 137 structure to emerge.

1 “Cruel-type” Small Clauses Taken by Verbs Other Than the Copula I will start by considering the existing contrast between cruel and “cruel + PP” as a complement of parecer ‘seem’. However. traditionally considered very close to be. volverse ‘become’. a stative SC (with the bare AP) is accepted. but when the SC contains the adjective plus the PP (197) it gets degraded. the constructions get degraded in some cases. can take an SC where the AP appears by itself (196). we observe. In the set of cases above. when the PP is present. The aim of this section is to analyze whether this is borne out. the verb seem. Partee (1977) and Dowty (1979) argue for a homophonous agentive copula with a meaning close to act.138 Individuals in Time chapter. and Rothstein (1999) defends that the copular verb maps states onto eventuality-denoting predicates. at least. As the examples show. the activity-like properties are derived from the properties of the (adjectival) predicate heading the small clause (SC) taken by the copular verb. First. a verb traditionally considered “pseudocopulative. 4. Second. in all the examples. three things. This hypothesis makes the prediction that such active properties should be retained when the taking verb is other than the copula.7. as a complement of hacer ‘make’ and as a complement of volver(se) ‘become’. it is not the case that “cruel + PP” fits in all cases. (196) María parece cruel Maria seems cruel (197) ?María parece cruel con Pablo Maria seems cruel to Pablo (198) María se volvió cruel Maria CL-REFL become cruel (199) María se volvió cruel con Pablo Maria CL-REFL become cruel to Pablo (200) María volvió cruel a Juan Maria turned Juan cruel (201) ??María volvió cruel a Juan con Pablo Maria turned Juan cruel to Pablo (202) María hizo cruel a Juan Maria made Juan cruel (203) ??María hizo cruel a Juan con Pablo Maria made Juan cruel to Pablo Note that. can take . however.” with an intrinsic inchoative meaning. In the following paragraphs I attempt to explain the contrast between allowing and disallowing an “AP + PP” SC. In my proposal.

only the bare AP is good. I will begin with parecer ‘seem’. a state such as be tired looks good. are states. the preposition introducing the (affected) goal is actually an aspect head conveying activity-like properties to the construction. For the same . Parecer takes SCs containing predicates which. I argue that “AP + PP” is only compatible with predicates selecting for SCs with certain aspectual properties (process-denoting) or with predicates allowing for SCs with any kind of aspectual properties. As a consequence. parallel to the examples with hacer ‘make’. I repeat here the structure I give for the SC the copular verb takes. And third. (200) and (201). Consider the following contrast: (205) *María parece [obtener el premio] Maria seems to-get the prize (206) *María parece [nadar] Maria seems to-swim (207) María parece [estar cansada] Maria seems to-be tired Whereas both an accomplishment and an activity such as to get the prize and to swim are excluded. aspectually. the SC can be considered to bear activity-like properties. I will account for such contrasts by arguing that the compatibility or incompatibility is due to the aspectual matching between the verb and the SC.Aspectual Alternations in IL Predicates 139 both (198) and (199). I will argue that only those verbs selecting (or allowing) for SCs containing dynamic predicates are compatible with an “AP + PP” SC. under a causative form. (204) ser Juani ser 2 cruel 2 cruel 2 PP (con) = activity inductor 2 (∅toward) con 2 con DP cruel As I largely argued.

When a QP rises from the lower subject position across. for example. quantifier lowering is blocked across “seem + PP”. whereas the simpler one is more defective (looking very close to a modal). “seem to me” in (iiib). whereas the AP with the PP (an activity SC) gets degraded. Nobody seems to have left (scopally ambiguous: seem > negation and negation > seem) Nobody seems to me to have left (only main clause negation scope: negation > seem) It could be argued a “richer” structure for the form with the clitic. in English. Therefore.c. and does not sound natural in other tenses other than the present. it cannot “lower” at Logical Form. therefore behaving as a state. I argue. is accepted. b. (The judgments are from Spanish). Although I do not have an explanation to it now. This hypothesis could explain. and the scopal ambiguity disappears. for example. I will bring up two remarks suggesting that seem and “seem + PP” are actually different. precisely. the contrast between (i) and (ii) is a piece of evidence suggesting that parecer has different properties than “clitic-parecer”.140 Individuals in Time reason. the aspectual property they involve is. (iv) ??María pareció cruel Maria seemed cruel (v) ??Maria ha parecido cruel Maria has seemed cruel . Tim Stowell (p. (208) √ cruel 2 volverse cruel 2 cruel con 2 Juan con 35 con 2 Pedro It is worthwhile to notice that parecer ‘seem’ in its combination with a clitic in Spanish (with a PP in English) gives different results. María me parece cruel con Juan Maria me CL-seems cruel with Juan “Maria seems cruel to me with Juan” (ii) ?María parece cruel con Pablo Maria seems cruel to Pablo (i) In this respect. (190) with the plain adjective. even in the simple “parecer + AP” cases.35 Regarding inchoatives like become. the same one I propose for the PP in these cases (both involve process properties). why the clitic forms can be found in all tenses. they are aspectually compatible.) observes that. (iii) a.

Causative constructions can be considered telic constructions.36 36 Before proceeding further. as a property. would make telic something which is not. to make someone cruel to someone else is not. it can be taken as the deglutinated version of a de-adjectival inchoative verb. which does not exist but could have existed. The tree wants to suggest that hacer. a telic constellation against an atelic one: (209) AspQ 2 hacer cruel (–Quantity) 2 cruel con 2 Juan con con 2 Pedro We can assume that Quantity is associated to the lexical entry hacer ‘make’. rather than as an event. at least. when volver is a synonym of hacer ‘make’. for instance).Aspectual Alternations in IL Predicates 141 However. is grammatical (actually. the sentences improve. [+quantity]). According to the examples above. an “aspectual clash” with the consequent oddity. or cannot be so. which seems the . something like encruelizar ‘cruelize’. I argue that is due to an aspectual mismatch too. it is expected it cannot be a part of a construction which is telic. I would like to remark on the nature of the DP inside the relational PP. whereas to make someone cruel. therefore. whose markers (the explicit causative verbs) can be analyzed as aspect markers (in concrete. as a quantity head. (i) ??María parece cruel con Pablo Maria seems cruel to Pablo (ii) ?María hizo cruel a Juan con Pablo Maria made Juan cruel to Pablo (iii) ?María parece cruel con los animales Maria seems cruel to animals (iv) María hizo cruel a Juan con los animales Maria made Juan cruel to animals These examples suggest that when the DP inside the PP is a generic plural. In aspectual terms. Compare (i) and (ii) with (iii) and (iv). with a stative SC. The SC is understood. triggering. as causatives are. If we take it that cruel to someone involves and gives dynamic atelic properties. sounds worse. It would be like having a causative with inchoative properties. since it encodes the causative meaning. the active SC seems excluded. or.

for example) the preferred selection of these predicates. Inchoative volverse and causative counterparts get odd in the imperative when the PP is added. and AP+PP). That is. its oddity in the imperative form is not a surprise. inchoative volver(se) ‘become’. I have to conclude that ser seems to be “looser” in its selections and can get either a stative SC or a dynamic unbounded eventuality SC. Cruel to animals seems. on the one hand. the combination of ser with AP or AP+PP gives good results in the imperative. then. it appears that the more semantically vacuous and syntactically simpler (note causative make has quantity in its structure. . Finally.142 Individuals in Time With this discussion I have tried to show that the properties deriving from the preposition matter when they combine with the verb selecting for the SC. to have been reanalyzed as a property at some level. Given the similarities existing between habituals and generics and statives (see chapter 3). In the same vein. does not take the active SC but just the stative. which can be considered semantically more vacuous than the causative hacer. The different behavior of ser. However. additionally. is not as odd with the PP as the causatives. Obviously. However. The command cannot refer to the event involved in the SC. and the command over the particular eventuality referred to by “cruel + PP” is the command of the whole sentence. the next question is what happens with ser. confirms ser as a very light verb. Since parecer is a state and. Regarding the inchoative (212) and causative cases. which so nicely accepts both forms (AP. this is not surprising. when the active SC is present. when the command appeals to the main verb (become. the command seems to sound better when the PP is not present. and volverse and hacer. make) the imperative seems grammatical. A perspective where the (mis)-matching is conceived in aspectual terms seems to make sense of the body of contrasts in (196)–(203). it is degraded. I would like to note contrasts such as the following: (210) ¡Sé educado (con Pablo)! Be polite (with Pablo)! (211) *¡Parece educado (con Pablo)! Seem polite (with Pablo)! (212) ¡Vuélvete educado (?con Pablo)! Become polite (with Pablo)! (213) ¡Vuelve educado a Pablo (??con Juan)! Turn Pablo polite (with Juan)! (214) ¡Haz educado a Pablo (??con Juan)! Make Pablo polite (with Juan)! As shown. In sum. parecer gives an odd sentence under such a form either way. on the other. Bearing in mind the discussion thus far. (213) and (214).

This fact led me to reject in section 4. no null PP would be proposed. Although null. Rothstein 1999). 4. mean. I explained as aspectual (in)-compatibilities the appropriateness of the relational MP SCs with each verb. Correspondingly.2 Summary of Section 4. accordingly. in the cases where an explicit PP is odd. according to which such dynamic properties are in correlation with the status of the predicate as SL. I have shown that this is not necessarily the case.7 I have analyzed the behavior of relational MP SCs as a complement of verbs other than be.7. I also discussed Stowell’s (1991) work. the better it accepts the active SC. and make. as the full grammaticality of Spanish copular clauses with ser suggests. unlike previous accounts (Partee 1977. the P brings the dynamic properties enabling forms such as imperatives.8 Summary of the Chapter In this chapter I have accounted for those copular constructions that behave as activities from an inner aspect perspective. kind to someone). and Rothstein (1999). 37 In imperative forms with ser. which proves its empirical superiority. Such a hypothesis has to admit that in a number of cases it is not possible to say at first sight whether the sentence is an activity (with a PP null) or a state. rather.1 previous accounts by Partee (1977). since they attribute the dynamic behavior found in copular clauses to the copula itself and do not make any prediction as to why the dynamic properties are manifested in the presence of a certain group of adjectives. I have argued that APs that can take a relational complement are stative in the absence of such a PP complement (although the PP complement can be phonetically null too). I assume a null PP when it is not overt. the better it bears a command imperative. Dowty (1979).37 4. I have shown that the dynamic behavior is not inherent to this type of APs. but it is only present if the PP complement is also present. I argued that to attribute the dynamic properties to pieces inside the SC.Aspectual Alternations in IL Predicates 143 verb is. I have shown that this aspectual behavior is not idiosyncratic or unpredictable but. I showed that the properties attributed to the SC headed by the adjective are “active” independently of the taking verb. makes more predictions. which can make sense of the preferences of different predicates such as seem. more suitable with those adjectives describing properties that can be understood in relation to another animate entity (cruel. . become. Taking into consideration the aspectual properties of the SC. corresponds to the presence of a (relational-directional) PP. Dowty 1979. I have shown cases where the PP cannot appear overtly. where they were attributed to the copula. In order to support the hypothesis that cruel-type APs can also appear without the PP complement.

Aspect-Quantity being the unique relevant head and it being absent in both cases. all temporal relationships: Tense. Once in such a structure. both homogeneous predicates. Stowell (1993). (outer) Aspect. triggered by the preposition. I have treated the alternation state/activity as an instance of aspect shift. it becomes an activity because the adjective and the PP are met in the structure. bald) would acquire a relational meaning if inserted in a relational structure and. but the syntactic structure corresponding to a relational meaning they are inserted in that yields such particular aspect properties. 2004). taking seriously the correlation between the PP complement and the active behavior. the account exposed here highlights the relationship between prepositions and event structure. 2000. Likewise. That is.). Any adjective (e. Rather. The interpretation of the subject (as an agent) has been explained as an entailment from the aktionsart properties. Therefore. and others that prepositions can bear aspect information. Theoretically. as a consequence. Based on the adjectival constructions I have examined. While it is due to the lexical properties of the adjectives that they can easily take a relational complement or not. those adjectives that more naturally allow for such a relational complement are the ones that manifest active aspect properties.144 Individuals in Time Examples of this are sentences with inanimate DP subjects and clauses where the adjectival SC is a complement of the verb consider. I proposed that the dynamic properties characterizing our activity-like APs come from the aspectual information conveyed by the preposition introducing the (affected) goal argument (to Pedro). the hearer would have to come up with a suitable interpretation for the construction. have been reduced to the same set of temporal primitives. which can be brought onto the construction through lexical indirect or oblique arguments (PPs). Based on evidence brought up by Hale (1984). Davis (in prep. Thus. and (inner) Aspect.g. this amounts to saying that aspectual properties are not lexically encoded but are decided through the syntactic construction of the clause. it is not adjectives themselves. As a consequence. Demirdache and Uribe-Etxebarría (1997. I have concluded that they are not to be structurally distinguished through a purely functional projection. I have considered that the (most) basic structure corresponds . it would involve dynamic properties.. I have argued that the difference between these two types is dynamicity. following a strong configurational perspective such as that defended by Borer (2005) or Ritter and Rosen (2000). Regarding the broader theoretical issue about the real grammatical difference between states and activities. it is not because of the lexical properties of the adjective that the construction is an activity.

Aspectual Alternations in IL Predicates 145 to the stative counterpart. which allows for any type of SC. a VP (process event) and a final state (the result). In this concern. 38 In a similar vein. that ser is looser in its selection taste and I have argued it accepts stative as well as active SCs. which just accepts stative SCs. I have concurred with Schmitt’s idea. of a v projection. This makes a difference with respect to previous proposals of copular verbs in Spanish. According to Ramchand. In my proposal. simply. instead of a full procesual VP. syntax at the lexical level) of event structure can be decomposed into a vP projection (initiational state). though. authors such as Luján (1981). I do not consider necessary special functional projections dedicated to stative predicates. Ramchand (2003) argues for a simpler structure in the case of states. Schmitt (1992). ser takes a SC containing a PP which has been justified as an aspect head. the first phase syntax (roughly said. and Fernández Leborans (1999) consider that only estar-predicates involve aspectual content. As mentioned in chapter 2. .38 Therefore. She argues that states are eventualities that consist. however. I have shown contrasts between ser. and other pseudo-copulative verbs such as parecer ‘seem’. Throughout this chapter I have shown. that SC complements of ser actually involve aspectual content.

.

Chapter 5 Outer Aspect and Individual-Level Predicates In the two previous chapters I dealt with inner-aspect properties. I will first describe the meaning of different aspectual viewpoints and the way outer aspect works (section 5. Outer aspect refers to the perspective from which a certain eventuality is considered: in its progress in time or after its completion or before its starting. Following Klein (1994. outer-aspect properties do not fulfill quantity (i. Centered on individual-level predicates. I also introduced some discussion about the difference between the two types of homogeneous predicates—states and activities.2 that aspect viewpoints also involve quantification over occasions. and its absence amounts to leaving the predicate homogeneous. and the progressive. telicity) to the construction and the imperfect is compatible with quantity predicates (telic predicates). so that the complete articulation of Tense and Aspect and their parallel behavior can be easily contemplated. I will also present how Tense works here. namely.e. for example. the perfective. Since. In the previous chapter I advanced the essentials of the semantics of aspect with the aim of showing that inner-aspect properties can be reduced to the same type of primitives that Aspect and Tense involve.1 Besides the ordering component just mentioned. Throughout the discussion I will focus on three particular viewpoints: the imperfect. 1995) and Demirdache and Uribe-Etxebarría (2000). I am going to deal now with another aspectual realm. 1 The particulars of temporal interpretation are the topic of the next chapter. in Spanish. specifically. Its presence makes the predicate heterogeneous.. .3. I propose in section 5. which I will capture through a quantifier inside the Aspect composition. through the functional projection of Quantity. inner aspect) properties. I discuss the relationship between inner and outer aspect. I assumed that different event types are decided through functional structure. Following Borer (2005). I show that the perfective form does not contribute quantity (i. the external argument of Aspect is the internal argument of Tense.1). outer aspect. I conclude that. I argue that Aspect is a predicate (located structurally higher than Quantity) that establishes an ordering relationship between two interval arguments. In section 5. The main points I focus on are: (a) the description of a number of outer-aspect forms (“viewpoints”) and (b) the relations and restrictions that can be argued to exist or not exist between inner and outer aspect..e. Focusing on the imperfect and perfective. according to Demirdache and Uribe-Etxebarría’s proposal.

1 Tense and Aspect as Ordering Predicates In this section I deal in greater detail with the conception of Tense and Aspect as ordering predicates. which I mentioned briefly in chapter 4 (section 4.1). posteriority. atelic).148 Individuals in Time In section 5.. like to. I show that states are compatible even with the perfective. in Spanish at least. for example) and. I argued that dynamicity properties (that activities involve) can be described through prepositions. in some languages (e. ter Meulen. I turn my attention to adjectival IL predicates and show the different interpretations available with each type of viewpoint mentioned. inner-aspect characterization may work as an aspect indicator (an accomplishment can be taken as a perfect action. and simultaneity). but in dynamicity versus lack of dynamicity. the interested reader is referred to Partee. I argued that inner aspectual properties could also be conceived in ordering terms.e. Consistent with the conclusion that viewpoints do not interfere with quantity properties. as something that took place in the past). For an introduction of such notions. Therefore. The idea is depicted in (1). 5. Reichenbach gave a formalization of tenses in natural languages based on three intervals (Reference Time. as Tense and Aspect are considered ordering predicates. and Wall 1993. and the possible relationships that can be established among them (anteriority.1 Tense. Lillooet Salish). 1996) have captured this relational property of tenses by arguing that Tense is a dyadic predicate taking two temporal arguments: the time of the event and a time of reference. this being why the eventuality is computed as “dynamic and unbounded” (i. inner aspect can be as well.3 Authors such as Zagona (1990) and Stowell (1993. 2 . and Eventuality Time). In particular. 3 Anteriority. specifically.6. the progressive seems to impose restrictions on the type of eventuality. I show that. therefore. aspect properties can serve as indicators of temporal properties (a perfect action is computed as finished and. Since the line of this work is not formal semantics I do not discuss them here. At that point. by the same token.. Speech Time. I tried to show that all aspectual properties could be reduced to the same set of primitives. These predicates translate into the temporal domain as ‘before’. However.2 One of the facts I alluded to was that. It turns out that this cannot be described in terms of presence or absence of quantity.g.1. and inclusion are semantically defined as partial order relations fulfilling a number of properties. Stowell (1993.5. 5. that establish centripetal noncentral coincidence relations between a figure and a ground. based on Zagona 1990. 1996) The idea that temporal interpretation can be conceived as a relationship between the time of the event the sentence is about and a given time traces back at least to Reichenbach 1947. posteriority.

. which means the interference of syntactic constructions on temporal interpretation can be directly accounted for. (3) a. Stowell (1993. whose structure and working is analogous to the structure of DPs. as (3) schematizes. (2) TP 3 ZP (PRO) T′ 2 T ZP 2 Zi VP 2 ei VP 2 Subj V′ 2 V … In Stowell’s proposal. ZP 2 Zi VP 2 ei VP 4 From the German Zeit ‘time’. The novelty of this approach is that temporal arguments are represented in the syntax. DP 2 Di NP 2 NP ei b. called Zeit-Phrases (ZP). Elaborating on this idea.Outer Aspect and Individual-Level Predicates 149 (1) TP 2 T-ext arg T′ 2 T T-int arg Tense takes an external argument corresponding to the Reference Time (RT) and an internal argument that corresponds to the time of the event. 1996) argues for a more complex structure (2). Tense takes two time-denoting arguments.4 The external argument corresponds to the RT and the internal argument to the Eventuality Time (ET). as Enç (1987) had already noted. The ZP is conceived as a temporal quantifier phrase.

depending on the value of the predicate (Tense) and the value of the RT.150 Individuals in Time Just as a determiner or a quantifier binds the variable acting as the external argument of the NP (Higginbotham 1985.e. tenses resemble prepositions like ‘after’ or ‘before’. 1996) conceives the external ZP as an argument whose concrete features depend on syntactic conditions independently established. The external ZP is understood as an element comparable to PRO. consists of ordering the internal ZP with respect to the external ZP. the eventive variable is present only in SL predicates.. 1996) notices. since they do the same job—that is. and present can be said to mean ‘in’ or ‘within’. future means ‘before’. The job of Tense. but the UT is one of the possible denotations the RT can have. Past tense locates the RT. It is worthwhile to note that in this framework the Utterance Time (UT) and the RT are not different primitive notions (as they are in Reichenbach 1947 or Hornstein 1990). the UT. in this sense.////////---------↑ UT As Stowell (1993. Stowell 1989). except that the content of T would change from ‘after’ to ‘before’ and ‘within’. In contrast to Zagona (1990). The external ZP is subject to control. Z binds the external eventive argument (e) of the VP (Kratzer 1988). after the ET (4). in sum. they locate an interval with respect to another. 5 Recall that. Past means ‘after’. The tree in (7) represents (4). from the nominal realm. it gets the Speech Time as a default value. past.5 Time-denoting phrases are viewed as referential or quantificational categories. the temporal interpretation (i. respectively. and the present locates the RT coinciding with the ET (6). Stowell (1993. What are the values of Tense? These are what we usually call “tenses”: present. In the absence of a controller (which is the case for matrix clauses). for Kratzer (1988). future. the location in time of the eventuality) is derived. who conceives the external argument as specifically denoting the Speech Time. Abney 1987. This way. and its controller is the closest c-commanding ET ZP. the future does it before (5). (4) Mary took the book ---------------///////////-------UT (5) Mary will take the book --------------UT-------////////// (6) Mary knows mathematics -------------. . which Stowell considers present in every type of predicate. The representations of (5) and (6) would be basically identical.

ZP(ET) 2 ET PP (at 5 P.M. in embedded sentences the situation is more complex since the specification of the external ZP becomes affected by the internal ZP of the main clause. However. for the moment. prior to 5).e. temporal relations are accounted for by arguing that Tense is a dyadic predicate that takes two temporal arguments (ZPs) plus general syntactic rules. the value of the external ZP is the UT. the grammatical existence of such ZPs is demonstrated. These authors argue that time adverbials can be considered as modifiers of temporal arguments. For example.) (iii) To the extent that time adverbials can be considered modifiers of time-denoting arguments. where I explore the temporal interpretation of IL predicates in simple and compound sentences. In turn. yielding the reading according to which Maddi’s leaving takes place at 5. Demirdache and Uribe-Etxebarría (2004) provide empirical evidence supporting their actual representation in the syntax. I examine this situation in chapter 6.M.M. with no additional stipulation. In sum. Modifiers become the evidence for modified arguments. The different readings thus emerge depending on which temporal argument they modify. nothing else hinges on this. In (ii).6 I will adopt this conception of Tense 6 As for ZPs. . the interpretation is the so-called reference-time reading: the time of Maddi’s leaving is presented as culminated prior to the TT (i.. if the adverbial modifies the TT (iii). the two interpretations of (i) are derived from the different ZP modified in each case.) ZP(TT) 2 TT PP (at 5 P.Outer Aspect and Individual-Level Predicates 151 (7) TP 2 ZP T′ (RT=UT) 2 T ZP (after) 2 Zi VP 2 ei VP 2 Mary V′ 2 take the book Because these sentences are matrix sentences. the adverb restricts the reference of the ET. (i) (ii) Maddi had left school at 5 P.

Now. However. he intends (10) rather than (9).2 The Internal Argument of Tense: The External Argument of Aspect. The former would correspond with readings of the sort in (10). whereas the latter with those along the lines of (9).1. in principle. The analogy between DPs and ZPs enables Stowell to distinguish between “definite” and “indefinite” (with just pure existential force) ZPs. Smith (1991) makes an insightful comparison between aspect and the lens of a camera. the total interval during which María has been washing the car is left in the dark. at that time. but the interval the speaker is referring to. for example. makes them visible for the receiver to be able to catch them in a photograph. the obvious question is: what tool picks out the interval the sentence makes an assertion about? Following Smith (1991) and Klein (1994). as Klein (1994) pointed out: Tense does not order the ET with respect to RT. 1984b) and was also noted in Stowell’s (1993. That time is in the past. be paraphrased as (9) or (10). 5.152 Individuals in Time as an ordering predicate. Depending on the focus . by focusing objects. when someone utters (8). 1996) work. The observation that Tense can refer to a concrete interval traces back to Partee (1973. (10) seems the most accurate paraphrase of (8). It is clear that it seems desirable to establish that the interval located in time is not the whole span of time the event is taking place. (10) There is a specific time at which Maria was involved in the task of washing the car. (8) María estaba lavando el coche Maria was washing the car (9) Maria was involved in the task of washing the car at some point in the past. I assume that the element that does this job is Aspect. According to (9). It orders the time the speaker talks about with respect to a time of reference. I call the interval the sentence is about (or “the interval the speaker talks about”) “Topic Time” (TT). María was involved in the task of washing the car. Aspect as an Ordering Predicate The temporal interpretation of a sentence like (8) can. That is. Following Klein (1994). This leads to a big conclusion. Between the two. In (8). In the next subsection I will amplify the discussion about the internal argument of Tense and the conception of Aspect as an ordering predicate too. the speaker is asserting that María has been washing the car some time in her life. The lens of a camera. (10) wants to capture a different assertion: the speaker has a concrete time in his mind and he is asserting that.

” as opposed to focus. the progressive makes visible just some portion inside the event.) points out. as they appear in García 1999. …and he is still working on it. Before proceeding further. as a whole. some things or others will appear in the picture. In sum. nothing can be deduced about its ending bound. the portion of the situation that Aspect “privileges” is the portion of the situation asserted by the speaker in the utterance. Similarly. 7 . we see neither the initial nor the final bound. …but he did not finish it. in (11).7 Only what is made visible is asserted in the sentence. some portions or others of the situations will get “registered” and become relevant for the interpretation of the sentence. The perfective lens allows us to see the entire event: (11) Pablo made a cake [/////////////////////////] Compare (11) with (12). However. Distinct “aspectual viewpoints” (in Smith’s words) correspond with distinct lenses. Smith argues. the event is presented as completed—that is. b. the appropriate analogy here seems to be “wide angle” versus “telephoto. Whereas the existence of the initial bound (onset) gets presupposed because the asserted part belongs to the inside part of the process. which I consider a significant work of reference for Spanish.c. For example. (13) a.8 As Tim Stowell (p. 8 García 1999 is part of the Gramática descriptiva de la lengua española (edited by Bosque and Demonte 1999).Outer Aspect and Individual-Level Predicates 153 setting. there is an asymmetry between these two. (12) Pablo was making a cake [------//////////-------In (12). depending on what Aspect focuses. let me introduce a few notes of clarification about the aspect nomenclature I assume and its correspondences with the verbal forms in Spanish. As a consequence. for instance. That is why (12) can be continued either of the ways in (13). Focus would make things blurry as opposed to excluded.

as Demirdache and Uribe-Etxebarría (2000) put it. I follow Klein’s (1994) proposal that there is an ordering relationship between the TT and the ET.3SG Trabaja-ba he-work-IMPF-PRET. Returning to the internal working of Aspect. how does the interval corresponding to the singled out portion of the situation (the TT) relate to the rest of the ET? As mentioned earlier. 9 As many authors have pointed out. I mention them here only as clarifications. More strictly speaking. whereas with the perfect. which is the element in charge of establishing a relationship between these two intervals. I restrict my attention to the forms called “imperfect” and “perfective” (and concretely to the perfective simple form) as well as with the progressive.3SG Había salido a las tres he-had left at 39 Había salido ya a las tres he-had left already (at 3) Trabajará he-will-work Trabajaría he-would-work Table 5. Aspect. I am not going to analyze all these aspect viewpoints. the next question is how the singled-out portion of the situation (TT) relates to the rest of the event. In the glosses for my Spanish examples. That is. Tense and Aspect consist in the same mechanism—namely. The verb estar acts as an auxiliary and can appear in any tense and aspect form. Aspect nomenclature (García 1999) As indicated earlier. (i) Perfective → the leaving occurs at 3 (ii) Perfect → the leaving occurs prior to 3 . The progressive is construed by the verb estar (locative be) plus the ing-form of the verb (Spanish -ndo). the difference between the composed forms of “have” corresponding to the perfective and the perfect is that. is an ordering predicate.1. the adverbial refers to the time at which the event takes place. with the perfective. the event has already taken place at the time denoted by the adverbial. I will thus use “perfective preterit” and “imperfect preterit” to refer to the two simple past forms. Thus.3SG Trabajó he-work-PERF-PRET. that of ordering temporal arguments.154 Individuals in Time Aspect name Imperfect Verbal forms Present and imperfect preterit Perfective or aorist Simple perfective preterit (and composed forms with “have”) Perfect Neutral Composed forms with “have” Simple future and simple conditional Examples Trabaja he-work-PRES.

As shown in chapter 4 (section 4. Likewise. The structure in (14) captures all this. for the sake of completeness I repeat it here. without entering into the differences it displays with the perfect. (i) and (ii). The slashes indicate the part of the situation asserted (TT). With the progressive (15).10 and when the prospective is involved. depending on the inner-aspect properties of the predicates. As Bertinetto (1986) points out. Although the working of aspect was also described in chapter 4.Outer Aspect and Individual-Level Predicates 155 As these authors also point out. As Klein (1994:109) points out. aspectual relationships can be described through the same type of prepositions as Tense’s relationships (‘after’. ‘before’.1). Aspect orders the TT with respect to the situation time or ET. Tense and Aspect only differ with respect to which arguments these predicates take to order. or to the end of the event (iii). ‘[with]in’) work in the tense realm. the assertion time is after the situation. ‘after’.6. However. ‘within’). with the perfective. I will assume that the ordering predicate for the perfective is ‘after’. while with the perfect the TT is conceived ‘after’ the ET. the perfective can refer either to the time at which the event started. and the dotted line the entire situation. (14) RT T TP 2 T′ 2 AspP 2 TT Asp′ 2 Asp VP 2 ET VP 2 e VP In the previous section I showed the precise ways the different values of T (‘before’. 10 Actually. the asserted part is within the situation. it appears before. Tense orders the TT with respect to an RT. the TT is partly included in the interval at which the event takes place. since nothing of what I will comment on hinges on these distinctions. la canción a las 3 (i) Juan cantó Juan sing-PERF-PRET-3SG the song at 3 “Juan sang the song at 3” Accomplishment . with the perfective. I will not make distinctions among the different types of perfective. the content of both predicates can be reduced to the same set of primitives. this is an oversimplification.

‘toward’ the ground). (16) Mary took the book ……………. In (19) and (20) the relation between figure and ground is of noncentral coincidence.156 Individuals in Time (15) Mary is taking the book …………. the figure is ‘within’ the square. which represents a relation of central coincidence. In (21) these three relationships are depicted structurally. this means that the TT is captured ‘within’ the ET./////////// (17) Mary is going to take the book //////////……………… Progressive Perfective Prospective As mentioned earlier.////////……. The latter is. following Demirdache and Uribe-Etxebarría (2000) and corresponding to the forms in (18)–(20). described as a centrifugal relation (i. temporally.e. (18) (19) (20) ◙ ○■ ■○ In (18). The former is described as a centripetal relation (i... each of these relationships can be described in terms of figure and ground (see Talmy 1978. respectively. (ii) Juan paseó a las 3 Juan go-PERF-PRET-3SG for a walk at 3 “Juan went for a walk at 3” a las 3 (iii) Juan llegó Juan arrive-PERF-PRET-3SG at 3 “Juan arrived at 3” Activity Achievement . in turn. ‘from’ the ground). means that the TT of the sentence refers to a moment ‘previous’ to the interval covering the event itself. The white circle stands for the figure and the black square for the ground.e.. but in each case the figure and the ground are in an opposite ordering relation. where the TT alluded is ‘after’ the ET. Hale 1984) as follows. which. In temporal terms..

perfective aspect has been associated with “completion” and “delimitation. the ordering one just presented. the event can be conceived as open. same contents. just different arguments to order. there are “two floors”—namely. If the TT is located ‘after’ the ET. I defend the position that Aspect contains a quantificational component referring to the number of occasions a particular eventuality is said to have taken place. Whereas Tense orders the TT with respect to an RT whose content depends on the syntactic environment. 5. Traditionally (Comrie 1976.” These characteristics can be derived from the fact that the perfective locates the TT ‘after’ the ET. Aspect orders the TT with respect to the whole span of time the eventuality extends over. it does not say anything about its beginning or ending. Finally. it is hard to demonstrate that factors such as “open” or “closed” have any real grammatical import where aspect is concerned. prospective aspect is described as an indicator of futurity. Progressive AspP 2 TT Asp′ 2 Asp ET (within) b.2 Aspect as a Quantifier over Occasions In the articulation of Aspect I propose here. Prospective AspP 2 TT Asp′ 2 Asp ET (before) The properties commonly attributed to each aspectual viewpoint are derivable from the ordering properties just described. the event can be seen as closed. In sum. which is explained given the fact that the TT is located just ‘before’ the event. where perfective events correspond to closed intervals and imperfect ones to open intervals. If the TT is located ‘within’ the event. . If the TT is located ‘within’ the event.” because if completed. among many others) are those of closed versus open interval. Tense and Aspect are explained via the same set of primitives: same mechanism. just about its developing. “delimitation. delimited. 2000). Perfective AspP 2 TT Asp′ 2 Asp ET (after) c.11 Similarly. and a quantificational one. Following Verkuyl (1972. Although these metaphors may be illustrative and clarifying. among others). this is distinct from the projection 11 Two other popular notions to account for Aspect (Comrie (1976) and Bertinetto (1986. 1999). as a consequence. As I will argue. Closedness and openness are descriptions that can be derived from the ordering properties of aspect. This posteriority ordering explains by itself “completion” and. 1993. the widespread description of the progressive as an “ongoing viewpoint” is derived from the ordering property.Outer Aspect and Individual-Level Predicates 157 (21) Aspect structures a.

5. here in section 5. can be either |1| or |>1|.2 and next in section 5. also contain information about the number of occasions an eventuality occurs. (22) Pablo estaba paseando por el parque Pablo was walk-ing around the park (23) Pablo paseó por el parque Pablo walk-PRET-PERF-3SG around the park (24) Pablo paseaba por el parque Pablo walk-PRET-IMPF-3SG around the park These sentences all refer to a (homogeneous) eventuality that took place in the past. how many occasions each viewpoint alludes to. However. as the reader may have conjectured already. I address these points in turn. As I will repeat later in section 5. who observed that the difference between the progressive and the habitual rests on the number of occasions they refer to. However. The continuous imperfect does not count the number of occasions a determined predicate takes place. The values of such a quantifier. In this latter regard. which I analyze as complex phrases containing a quantifier over occasions and ordering information.3.2. what the relationship is between the number of occasions alluded to by each viewpoint and the quantity nature of predicates (that is. and the progressive. I will consider two points—first. as shown in (25). aspectual viewpoints. I propose that aspectual viewpoints involve. a quantificational component. besides ordering. besides the ordering component. It appears . the habitual imperfect. Let me start by considering sentences (22)–(24).158 Individuals in Time of Quantity. the three forms differ in the number of occasions they allude to.1 Quantifying over Occasions As just mentioned. and second. inner aspect. the functional projection which makes the eventuality heterogeneous and whose absence leaves it homogeneous. The number of occasions the progressive and the perfective form allude to is just |1|. whose head is a quantifier over occasions (Q<occ>). as in (26). (25) ----------------------x---------------------(26) ------x------x--------x-------x---------x-In the spirit of Verkuyl (1999). In this section I am chiefly concerned with the perfective. giving rise to another interpretation. Q<occ> can also have an existential value.1. the imperfect in (24) refers to a plural number of occasions that can be described as greater than one (|>1|). the nonhabitual or “continuous” imperfect. 5.

Progressive AspP 2 TT Asp′ 2 Asp Q<occ> (within) 2 1 walk b. therefore. (i) Voy a pasear I am going to go for a walk (ii) Sospecho que va a ser rubio cuando sea mayor I think he is going to be blond when he gets older (iii) a. Prospective ∃ AspP 2 TT Asp′ 12 2 Asp Q<occ> (before) 2 1 walk 13 2 Asp (before) (Q<occ>) 2 ∃ be blond Although continuous imperfect typically appears with states. Prospective |1| AspP 2 TT Asp′ b. be conceived independently from a particular number of instances. con insistencia (i) Durante la reunión me miraba during the meeting me he-look-IMPF-PRET-3SG with insistence .12 Note that it is not only the homogeneous property that counts (activity predicates like walk and SL statives like be sick are also homogenous) but the property of counting (or not) the number of occurrences the predicate is true of a subject. Whereas for sentences like (i). where we can distinguish the habitual and continuous forms.Outer Aspect and Individual-Level Predicates 159 with predicates that “hold. García (1999) mentions (i) as an example of continuous imperfect. I would like to point out that its articulation regarding the quantification over occasions is similar to the imperfect. the quantifier can be argued to be ∃ (iiib). it can be defended a cardinality of |1|.” rather than “take place.” and can. it can be also present with eventive verbs. Perfective AspP 2 TT Asp′ 2 Asp Q<occ> (after) 2 1 walk Although I will not work on the prospective aspect. for other sentences (ii) with a different predicate. as in (iiia).13 In (27) the differing ordering and quantifying components of each aspect form are represented: (27) a.

the imperfect form corresponds to (27d). When the predicate is stative. Imperfect forms of eventive verbs like paseaba (walk-PRETIMPF-3SG) in (28) have two possible interpretations: progressive. María was walking around the park por el parque a las tres (30) Habitualmente. habitual and progressive. This progressive/habitual syncretism obviously concerns those predicates that can be conceived in progress as well as habitual. whereas the habitual. the continuous. I will ignore the progressive reading of the imperfect form. and when I deal with the properties of the progressive I will use the explicit progressive form estar + V-ing. and the imperfect continuous of (27d). María paseaba por el parque At three. Imperfect continuous AspP 2 TT Asp′ 2 Asp Q<occ> (within) 2 ∃ be blond According to this. a few clarification notes with respect to Spanish imperfect forms are in order. I argue. Before proceeding further. María walk-PRET-IMPF-3SG around the park (29) A las tres. and the progressive differ in the occasion-quantification part. eventive predicates. The progressive paraphrase of the imperfect is far more natural in compound sentences (i) than in simple sentences such as (28). Pedro leía el periódico (i) Mientras María paseaba. the perfective differs from the progressive in the ordering part. the imperfect habitual of (27c). that is. Spanish imperfect forms can correspond to the progressive of (27a).14 and habitual. (28) A las tres.160 Individuals in Time c. where the imperfect suffix has ended up expressing two meanings. Imperfect habitual AspP 2 TT Asp′ 2 Asp Q<occ> (within) 2 >1 walk d. María paseaba habitually María walk-PRET-IMPF-3SG around the park at three “(Habitually) Maria used to walk around the park at three” In what follows. like have a house (31). where. María estaba paseando por el parque Yesterday at three. I will simply assume that this is a case of syncretism. and it cannot be understood as an eventuality in progress (32) or as habitual (33). whose paraphrase appears in (30). the imperfect continuous. with a paraphrase like ‘was walking’ (29). there is a plain existential quantifier over occasions. while Maria walk-PRET-IMPF-3SG Pedro read-PRET-IMPF-3SG the paper 14 .

Likewise. the progressive and the habitual share the same ordering component. the TT is contained in the period of time when Pablo was walking. (36) -----------xxxxx[/]xxxxxx---------------↑ when I saw him (37) ------x------[x--------x-------x]---------x-ie his teens In (36). Pablo used to walk around the park” The relationship established between the TT and the ET is the same in both cases—namely. Pablo was walking around the park” por el parque (35) En su adolescencia. Although the habitual viewpoint will be more extensively discussed shortly. to go walking in the park. and maybe still use in the present. consider the following contrast to illustrate this shared characteristic: (34) Cuando lo vi. . the TT interval in his teens of (37) is included in the period of time during which Pablo used to walk in the park. Pablo estaba paseando por el parque when him saw Pablo was walk-PROG around the park “When I saw him.” The TT in (34) is identified by the temporal clause when I saw him. Pablo paseaba in his teens Pablo walk-PRET-IMPF-3SG around the park “In his teens.Outer Aspect and Individual-Level Predicates 161 una casa María tenía Maria have-PRET-IMPF-3SG a house (32) *María estaba teniendo una casa Maria was having a house (33) *Habitualmente María tenía una casa habitually Maria have-PRET-IMPF-3SG a house “Maria used to have a house” (31) As indicated in (27). for example). That is. that of “containing. He may have kept on walking afterwards (when he was 34) or he could have walked in the park when he was not a teenager yet (when he was 5. the period of his teens is included in the period of time during which Pablo used. the TT in (35) is specified in the phrase in his teens.

in and of itself. their compatibility with for-adverbials gets explained. there is no need to interpret that the paper has been read completely or the sonata has been played entirely. I would like to note that in the iterative interpretation with the perfective and for-adverbials. and. nor an account for these cases. Consider (38). Although I will discuss neither the process of coercion itself. that cake is not physically the same cake year after year. (i) Pablo llamaba a la puerta cuando apareció el monstruo Pablo knock-IMPF-PRET-3SG at the door when the monster appeared (ii) Pablo was knocking at the door when the monster appeared 16 15 In this sense. For example. In (i). However. as she conceives habituals to be. several factors seem to play a role. where this interpretation is much more difficult to get. it is interesting to consider accomplishment predicates in perfective followed by a “for + time” adverbial. where the predicate is understood as iterated: (i) Hice ese pastel durante años (como postre de navidad) I-made that cake for years (as a Christmas dessert) De Swart (1998) accounts for these cases by arguing for the existence of a coercer operator. (iii) Leyó el periódico durante dos horas he-read-PERF-PRET-3SG the paper for two hours . as Tim Stowell (p. Once accomplishments are states. as a consequence. One can be reading just sections of the paper and playing just parts of the sonata. Compare (i) to (ii). although similar cases are more difficult to find with the Note that if we force an imperfect form to be confined to one occasion (i). simply. refers to a plural number of occasions. the iterative interpretation seems possible when the DP object can be understood as different instances of the same object description.15 the perfective and the progressive can never denote a plural number of times. it is interpreted as a progressive (ii). which turns the event (make the cake) into a state. this does not mean that the number of occasions that an eventuality in the perfective takes place cannot be modified by an adverbial.) points out.16 Likewise.162 Individuals in Time Regarding the number of occasions that the different viewpoints can allude to.c. the point I want to make with (27) is that. whereas the imperfect habitual. where. excluded: (ii) *Construí la mesa del salón durante semanas I-make-PERF-PRET-3SG the living room table for weeks Other typical cases of the combination of accomplishments followed by for-adverbials are sentences such as (iii) and (iv). a Venezuela tres veces (38) El año pasado Pablo fue last year Pablo go-PRET-PERF-3SG to Venezuela three times “Last year Pablo went to Venezuela three times” The adverbial “three times” modifies the (otherwise singular) number of occasions. the sentence is.

Now suppose we interpret (40) and (41) against the following scenarios in (42) and (43). this form can also be modified with respect to the number of times the eventuality is said to take place: (39) En las dos ocasiones en que lo oí. 5.1 Iteration. and Systematicity Verkuyl (1999) and Stowell (2000). I assume that the meaning of habituality is lexically expressed in quantifier adverbials such as habitually or usually. The speaker’s assumptions about the actual number of instances of a habitually quantified event depends on heterogeneous (pragmatic) factors. In this subsection I deal with the properties of one such quantifier that gives rise to habitual interpretation. Proportion. Consider the contrast between (40) and (41) as a starting point.2 The Habitual Interpretation: Iteration. as described in (27c).2. Pablo was knocking at the door three times. whereas the progressive and the perfective (in the absence of adverbial modification) refer to a singular occasion.2. Pablo estaba llamando a la puerta tres veces In the two occasions I heard him. the exact number of event instances is not specified. (42) Juan smokes four times a year. (40) Juan fuma Juan smokes (41) Juan va a Nueva York Juan goes to New York Suppose we take a year as a reference unit.17 among others. However. Following the quantificational-based perspective defended by Verkuyl. the habitual viewpoint refers to a plural number of occasions by itself. define habituality as an iteration of instances of a given eventuality distributed in time. (43) Juan lives in Madrid and goes to New York four times a year (iv) Tocó la sonata durante una hora he-play-PERF-PRET-3SG the sonata for an hour 17 Girona International Summer School of Linguistics class notes. As just mentioned. I consider that habituality corresponds to a quantifier over occasions denoting a plurality of occasions of a particular eventuality. Specifically.2. . and propose that it is expressed through functional quantificational structure.Outer Aspect and Individual-Level Predicates 163 progressive. 5. I will consider that habituality is encoded in the “quantificational floor” of Aspect.

and Wall. Suppose now that the number of right-handed people in the class is also five. (44) Many students got an A (45) Many students are right-handed Suppose that the set under consideration is a class of 20 students. given the situations of (42) and (43). The number of times considered “average” is established taking into account extralinguistic information.” which is established by external information. In particular. I propose that habitual Q<occ> behaves similarly to canonical quantifiers such as many. Specifically. it seems that. (40) does not reach the category of “habitual. as such. (40) and (41) are judged differently. given the fact that the amount of times (per year) people in our culture visit a city abroad (such as New York for a person living in Madrid) is much lower than four. (45) does not properly describe the situation. two intervening factors: (a) iteration. How does such pragmatic information get integrated into the aspectual encoding? I propose that aspect Q<occ> behaves like regular quantifiers and.2 Proportion. note. . ter Meulen. since the eventualities take place more than once. Thus. that in the determination of “habituality” there are. The following pair is from Partee. Whereas (40) is very likely to be judged as inappropriate. In other words. It seems. If five is more than the average ratio of students that get an A.2. I will review the behavior of this quantifier in the nominal realm briefly. (44) should be judged as a sentence that properly describes the situation. Since the number of times people usually smoke tobacco in our culture highly exceeds four times a year. among others.164 Individuals in Time Although the implicated number of times is the same (four). ter Meulen. when judging whether a habitual form is appropriate. and Wall (1993). what counts as “many” depends on different contextual parameters. (46) expresses this formally. some notion of “average” regarding the number of occasions that an action is performed is taken into account. we conclude that the crucial notion in describing many resides in the “average ratio. a habitual sentence is considered appropriate for depicting such a scenario. whose definition heavily depends on contextual basis. Suppose further that the number of students who got an A is five. at least. (41) is very likely to be judged as appropriate. therefore. As Westerståhl (1984) and Partee. 5. and (b) a certain proportion with respect to the number of times the action at stake is usually performed.2.” In turn. If the average ratio of right-handed people is (statistically) higher than that. is affected by contextual parameters (Westerståhl 1984). (41) is judged as appropriate for the situation in (43). Sentence (44) would be judged inappropriate for such a scenario.

for example. The habitual quantifier refers to a number of occasions considered “average. note that ‘John is a writer’ is a possible paraphrase for John writes but not for John usually writes. This would explain the fact. Consider the paraphrases for John writes as ‘John is a writer’. who argue that not all events are quantified under the same parameters. often observed in the literature. in general. Just as the interpretation of quantifiers like many is affected by contextual information. such as Kearns (1991). are not understood as proportional statements but as simple assertions of the existence of the event. Both authors argue that if. where the contextual parameter captures the average ratio relevant in this case. but people call them “writers. One can say of John that he is a writer even though he does not usually write. A given eventuality can be described as habitual if the number of occasions in which the eventuality gets substantiated is the approximate average ratio. consider that sentences like (47). 19 Other authors.Outer Aspect and Individual-Level Predicates (46) Many AB = many A |(A ∩ B)| > c · |A| where A {students} and B {get an A} 165 The number of students who got an A can be described as “many” if the cardinality of students who got an A (that is. when speakers use a habitual form.” where the margins to consider a certain amount of times significantly close to the ratio may be flexible and subject to other pragmatic considerations. the frequency with which he has to develop such an activity must be much higher than the frequency with which he must kill people for one to claim John is a murderer. etc. we say John is a teacher. so is the interpretation of the habitual quantifier. that an action in imperfect (I would not say. but it has to be characterized by the semantics of the right quantifier. traveling to a foreign city (41).18 That is. statistically established. we understand that ‘he can walk’ rather than ‘He usually walks around the house’ or something along these lines. where we understand that Pablo (surely a toddler) has learned to walk. In the terms I am proposing here. In fact. unlike many. This makes these authors conclude that it is virtually impossible to talk about quantification over events in an effective way.19 Obviously. habitual) can be understood as a property characterizing an individual. In parallel with many in (46). the habitual quantifier does not denote an amount of instances “greater” than the ratio but an amount that represents “the ratio itself” or is “close to the ratio. (Think of those who have written only one novel in their life. A ∩ B) is greater than the cardinality of students by a contextual (“norm”) parameter (c). they are not basing it on a strict statistical background but on their perception of what the average can be.). then. if the mere assertion of the event means to remain indeterminate about the number of occasions it takes 18 . of such an eventuality. I represent the interpretation of the habitual quantifier (Hab) as in (48).” The average ratio relevant in each case is necessarily established with the help of external information (the number of times considered average for smoking (see (40)). This kind of observation is of the same spirit as that by Zemach (1975) and Carlson (1977). assuming that this perception is shared by most of the society. I am defending the opposite view: it is possible to capture habituality in quantificational terms. the aforementioned cases would correspond to the continuous imperfect (covering present tense and imperfective preterit). However.”) Other illustrative examples would be sentences like Pablo anda ‘Pablo walks’. with no overt habitual quantifiers. that is.

but. my main point in this section is to describe habituality. this is not always the case. despite the fact that. contextual information participates in establishing one of the members of the proportion the quantifier refers to. in a simple case like (47). the semantic intuition about examples like (47). . we could say that we are comparing the number of occasions people usually bring about the action of smoking and the number of occasions the DP subject Juan does. rather. A certain event can be described as habitual if the cardinality of occasions that a subject performs it (|(A ∩ B)|) is asymptotically equal to the cardinality of times the event takes place by a contextual parameter (c). As I mentioned in section 5. we can paraphrase John smokes by using ‘John is a smoker’. in summer the children play-IMPF-PRET-3SG at tennis in the garden Juan ganaba Juan win-IMPF-PRET-3SG “In the summer the children used to play tennis in the garden. The cited contextual parameter would capture.” (50) En verano los niños jugaban al tenis en el jardín. (49) En las fiestas de la empresa había marihuana. Juan fumaba.2. aspectual quantifiers give us the number of instances of a particular event. rather than discussing the semantics of the present or imperfect preterit.1. in principle. Juan used to smoke. the number of occasions that people smoke (in general). Juan used to win” In (49) and (50) we are not referring to the number of occasions that Juan smokes (marijuana) in general or wins when he plays tennis in general. In any event. As before with many. this does not fully capture. However. Consider (49) and (50) as samples of other possible cases. the contextual parameter captures the average ratio relevant in each case. it would not be appropriate to use (47) to describe a situation where John has smoked only four times in a year. which give us quantities of individuals. at the parties of the company was marijuana Juan smoke-IMPF-PRET-3SG “At the company parties there used to be marijuana around. As shown. then. In principle.166 Individuals in Time (47) Juan fumaba Juan smoked “Juan used to smoke” (48) Hab AB = Hab B |(A ∩ B)| ≃ c · |B| where A {Juan} and B {smoke} Unlike nominal quantifiers. to my understanding. to the number of occasions that Juan smoked marijuana at the parties with place.2. in this case.

Outer Aspect and Individual-Level Predicates

167

the people from the company and the times he won when he played tennis in the summer as a child. In both sentences it is understood that Juan smoked marijuana at the parties of the company and won when playing tennis in the summer a significant number of times so that we can refer to them by using a habitual quantifier. That is, unlike (47), in these cases we are not considering the number of times Juan does something and the number of times taken as average for people to do so. We are not measuring the frequency that a subject does something in comparison to the frequency that other subjects do. Instead, we are measuring the proportion between, for example in (49), the number of parties and the number of times Juan smoked marijuana in such parties. This is represented in (51). (51) Number of times Juan smoked marijuana at the company parties -------------------------------------------------------------------------------Number of company parties

Suppose the number of parties is 10. If Juan smoked marijuana at just two of them, a habitual form would not be considered appropriate to describe such a situation. Only if Juan smoked marijuana at six or more of them would a habitual quantifier be judged as appropriate. In sum, although in the interpretation of habituality there are a number of occasions of reference with respect to which the number of instances of an event is evaluated, such a reference number is not always the frequency that “subjects” usually perform the action at stake. Other factors can play a role in determining the comparison class. As Tim Stowell (p.c.) brought to my attention, focus is one of them. Consider the following examples: (52) (53) (54) (55) John smokes on the train John SMOKES on the train JOHN smokes on the train John smokes on the TRAIN

Each of these examples has the following different implications, respectively: (56) (57) (58) (59) ‘When(ever) John is on the train, he usually smokes’ ‘When(ever) John is on the train, what he usually does is smoking’ ‘When(ever) someone smokes on the train, it is usually John’ ‘When(ever) John smokes, he is usually on the train’

As before, (52) is judged as appropriate if John smokes on the train a significant (greater or close to the average) number of times with respect to the number of occasions he takes the train.

168

Individuals in Time

(60)

Number of times Juan smokes on the train ----------------------------------------------------------------------Number of times John takes the train

However, when focus is applied to some of the elements of the sentence, the components of the proportion change. (57), (58), and (59) correspond to (61), (62), and (63), respectively. (61) Number of the types of things Juan does on the train ----------------------------------------------------------------------Number of times John smokes Number of people that smoke on the train ----------------------------------------------------------------------Number of times John smokes on the train Number of times Juan smokes ---------------------------------------------------------------------Number of times John smokes on the train

(62)

(63)

That is, we interpret the habitual quantification with respect to different situations of reference, which are set up by the intervention of focus. In (61), the habituality of smoking is measured with respect to other activities that John brings about when he goes on the train. In (62), it is measured with respect to who it is that usually smokes on the train, and in (63), with respect to the place where John smokes. Focus reshapes the restriction of the habitual quantifier in each case. (I cannot go into deeper detail on this topic in this work. For the role of focus in quantifier restriction, see Herburger 2000 and references therein.) 5.2.2.3 Systematicity or Regularity. Thus far, I have accounted for the properties of iteration and (high) proportion that I propose the habitual quantifier contains. I will argue now that habituality also involves some notion of “systematicity” or “regularity.” To describe this characteristic I will introduce other frequency adverbials that also involve iteration and (high) proportion but, nonetheless, are not a paraphrase of the habitual quantifier. Consider the following examples: (64) El mes pasado fui a Madrid tres veces Last month I went to Madrid three times (65) El mes pasado fui a Madrid a menudo Last month I went to Madrid very often

Outer Aspect and Individual-Level Predicates

169

In both, the event of going to Madrid is understood as repeated or iterated. However, as many authors have pointed out (see García 1999 and references therein), whereas the adverbial three times in (64) counts the number of times such an event has taken place, very often in (65) does not count the instances of the event but establishes a proportion between the number of occasions an event takes place in relation to a particular period of time (namely, last month). In particular, (very) often means that the frequency of times the subject went to Madrid last month is high. We see, then, that adverbials such as (very) often refer to iterated occasions and establish a proportional interpretation. The point I want to make now is that iteration and proportion do not suffice to make a quantifier habitual. Consider these examples: (66) Voy al trabajo en coche habitualmente I usually drive to work (67) Voy al trabajo en coche a menudo I drive to work often (68) Voy al trabajo en coche a menudo, pero no habitualmente I drive to work often, but not usually Both (66) and (67) refer to an event that takes place iterated times. However, they do not make the same assertion, as can be tested out with (68), in which we explicitly assert that an event takes place very often and deny its habituality without entering into contradiction. This suggests that the two quantifiers over occasions make different claims. The adverbial a menudo ‘often’ tells us that a certain eventuality takes place a high proportional number of times but does not give us any information regarding the systematicity with which a certain eventuality is iterated. When an event is considered “habitual,” it is entailed that, under normal circumstances, the event can be expected to be iterated. However, the adverbial a menudo ‘often’ is vague with respect to the regularity with which an eventuality takes place. Habitual adverbials allude to events that occur systematically, in the sense that the occurrences of the eventuality are distributed in time with regularity. Interestingly, claiming the reverse of (68) is not possible: (69) ??/*Voy al trabajo en coche habitualmente, pero no muy a menudo I drive to work usually, but not very often In (69) it is asserted that the event takes place habitually but not very often, and the result is a contradictory sentence. This reinforces the idea that habitual adverbials mean that an eventuality takes place a significant number of occasions.

170

Individuals in Time

Likewise, although “regularity” seems to be an ingredient of the meaning of habituality, I do not consider “regularity” to be enough for an eventuality to be considered habitual. Rather, a medium/high proportion is needed. Observe the following cases: (70) María prepara la cena los primeros viernes de cada mes María prepares the dinner on first Fridays of each month (71) María prepara la cena Maria prepares the dinner “Maria usually prepares the dinner” Consider a scenario such as (70), where María is in charge of taking care of the dinner the first Friday of each month. This is a scenario where a particular action takes place regularly: (72) --------F---------F----------F----------F-------However, if someone uttered (71) to describe such a situation, we would not consider it an appropriate description. If we are told that Maria prepares the dinner, we are strongly inclined to understand that Maria prepares the dinner every day (or a high number of days per week), and not just some days regularly distributed over the calendar, since preparing dinner is something that can take place every day. Thus, the action of Maria taking care of the dinner does not take place habitually, taking into account the number of times it takes place. As I argued above, contextual information intervenes in the component with respect to which the number of occasions the eventuality takes place is measured. 5.2.2.4 Summary of Section 5.2.2. I have argued that the expression of iteration, proportion, and systematicity or regularity is represented by a particular aspectual morpheme in languages such as Spanish. I have also mentioned that this meaning is lexically spelled out by adverbials like usually or habitually. I have considered habituality as an interpretation based on quantificational parameters. I argued that the meaning of the habitual quantifier does not simply denote a vague plural number of instances but refers to the average ratio that a certain eventuality is brought about. Thus, an eventuality can be described as habitual if the number of instances it is substantiated approximately squares with the number of instances that such an eventuality is performed on average. The quantity considered “average” is independently established, taking into account heterogeneous parameters, based on external statistical background. As a final remark, it is worth noting that habitual adverbials can also appear with aspectual forms other than the imperfect. The following sentences illustrate the fact:

Outer Aspect and Individual-Level Predicates

171

(73) Este año María ha venido al seminario habitualmente This year Maria has attended the seminar habitually (74) Estos meses Pablo ha faltado a clase habitualmente These months Pablo has missed class habitually (75) Este verano María ha paseado por la playa habitualmente This summer Maria has walked along the beach habitually In sum, the appearance of habitual adverbials is not tied to the occurrence of the imperfect habitual morpheme. Habitual adverbials seem licit with other aspect forms that, by themselves, refer to just one event-occasion. In these cases, habitual adverbials can be taken as modifiers of such Q<occ>, yielding the interpretation that the event has taken place repeated times regularly. Likewise, the imperfect habitual form can appear with adverbials modifying the meaning expressed by the morpheme. Consider the following sentences: (76) Rara vez Pablo nadaba en la piscina rarely Pablo swim-PRET-IMPF-3SG in the pool “Pablo used to swim in the pool rarely” en la piscina muy a menudo (77) Pablo nadaba Pablo swim-PRET-IMPF-3SG in the pool very often “Pablo used to swim in the pool very often” Rarely and very often modify the habitual quantifier in these cases. Rarely denotes a low average, and often a high average.20 5.2.3 On the Relation between the TT and the Habitual Q<occ> In (27) above, repeated here as (78), I argued for a quantificational component of Aspect, giving the number of occasions an eventuality occurs. This quantified set of occasions is ordered (‘within’, ‘after’, or ‘before’) with respect to the TT by the Aspect head.

20 It is worthwhile to note the different sentence position of the adverbials in all these examples. Although, unfortunately, I cannot enter in deeper detail here, observe the following examples in contrast with the ones cited in the text (73) and (76)-(77):

(i) ??/*Habitualmente María ha venido al seminario este año habitually Maria has attended the seminar this year en la piscina rara vez (ii) ??/*Pablo nadaba Pablo swim-PRET-IMPF-3SG in the pool rarely en la piscina (iii) Muy a menudo Pablo nadaba very often Pablo swim-PRET-IMPF-3SG in the pool

I give the Spanish translation with a gloss. habituality emerges from the existence of more than one TT rather than from the existence of a (separate) quantifier over occasions. Estaba muerto find-PRET-PERF-3SG John in the bathtub estar-PRET-IMPF-3SG dead As Klein claims. Imperfect continuous AspP 2 TT Asp′ 2 Asp Q<occ> (within) 2 ∃ be blond This conception of the relationship between a quantifier over occasions and the TT is not universally shared. Progressive AspP 2 TT Asp′ 2 Asp Q<occ> (within) 2 1 walk c. rather than a single one” (Klein 1994:47). he wants to make an assertion about some time in the past—namely. Although the situation becomes fuzzy sometimes. an imperfect habitual refers to more than one TT. For instance. Let me illustrate this with an example from Klein (1994:22). I am going to argue that the number of occasions an eventuality takes place is not encoded in the TT (the interval the speaker refers to) but is a result from an independent quantifier ranging over ETs. the time at which John was . Below the example. Rather. From this perspective. (79) They found John in the bathtub. He was dead (80) Encontraron a John en la bañera. then. the quantifier over occasions I described earlier looks erroneous or unnecessary. Imperfect habitual AspP 2 TT Asp′ 2 Asp Q<occ> (within) 2 >1 walk b. For him. Perfective AspP 2 TT Asp′ 2 Asp Q<occ> (after) 2 1 walk d. when the speaker utters He was dead. he does not want to assert that the time of John’s death precedes the UT. where the quantification over occasions the eventuality holds can be easily distinguished from the number of times the TT refers to.172 Individuals in Time (78) a. Klein (1994) considers habituality as a case where the speaker “chooses to speak about a series of topic times.

” That is. that the number of times an event takes place and the TT are two different notions that should be separated in the formal representation. I argue against the perspective that viewpoint values (perfective. progressive. therefore. we treat all aspect forms and cases uniformly. Depraetere 1995. however. there is no correspondence between the quantificational properties involved in the imperfect form and those of the TT.Outer Aspect and Individual-Level Predicates 173 found. none of them can be taken as the realization of Quantity. If. imperfect habitual. I conclude. we do not have several TTs. The TT of the sentence (the time of “finding John in the bathtub”). 5. among others). (81) He was dead AspP 3 TT Asp′ (the time John 2 was found) Asp Q<occ> (within) 2 ∃ be dead The type of imperfect involved in He was dead corresponds to the “continuous” imperfect. I will argue this is not the case for Spanish since Quantity co-occurs with imperfect-habitual forms. The representation of this would be like (81). in habitual cases. and he asserts that this time is included in the time at which John is dead. as desirable. there are no occasions to enumerate or calculate. I am going to insist on it with the aim of discussing the . which suggests that the quantification over occasions and the occasions designated by the TT are not the same. we can claim that. In He was dead. That is. and nonquantity predicates are compatible with the perfective. can legitimately count as “one occasion. which does not refer to a particular number of occasions. The Spanish data show that these aspectual viewpoint forms contribute information other than inner aspect. imperfect) are aspectual operators that can modify the inner aspect properties of eventualities. In particular. and imperfect nonhabitual or continuous) and Quantity. Borer 2005).3 Inner and Outer Aspect In this section I discuss the relationship between aspectual viewpoints (perfective. Although this has already been noted by several authors (Delfitto & Bertinetto 1995. as has been proposed for the perfective in Slavic languages (Filip 2000. but that the plural quantification over occasions comes from a different component inside Aspect.

which tests the predicate swim as atelic. 5. To begin. and (b) the co-occurrence of perfective viewpoint and homogeneous (nonquantity. atelic) predicates. either in the perfective or imperfect form. prepare the meal and write the report. In contrast. (84) and (85).3. Bearing in mind this telicity proof. let us consider the following sentences: la comida en media hora (82) Pablo preparó Pablo prepare-PRET-PERF-3SG the meal in half an hour “Pablo prepared the meal in half an hour” el informe en quince minutos (83) Juan redactó Juan write-PRET-PERF-3SG the report in fifteen minutes “Juan wrote the report in fifteen minutes” (84) *Pablo nadó en la piscina en doce minutos Pablo swim-PRET-IMPF-3SG in the pool in twelve minutes “Pablo swam in the pool in twelve minutes” (85) *Pablo nadaba en la piscina en doce minutos Pablo swim-PRET-IMPF-3SG in the pool in twelve minutes “Pablo used to swim in the pool in twelve minutes” Sentences (82) and (83) exemplify telic predicates. Their telic nature is tested by the appropriateness of the modifier in x time. and offer a general overview of the relationships in the temporal system. are odd in the presence of such a modifier. which points to the conclusion that the viewpoint does not affect in any sense the nature of the predicate.174 Individuals in Time relation between inner and outer aspect.1 Habitual Heterogeneous Predications The Spanish data show that the imperfect-habitual viewpoint is compatible with telic predicates. I will discuss the following points: (a) the compatibility of telic (quantity) predicates and habituality. consider now the following examples: la comida en media hora (86) Habitualmente Pablo preparaba habitually Pablo prepare-PRET-IMPF-3SG the meal in half an hour “Pablo used to prepare the meal in half an hour” .

Quantity properties. Each occasion will prove to have heterogeneous additivity and subinterval properties. Each occasion in which Pablo prepared the meal is heterogeneous. too. can co-occur. although the fact that the action takes place several times leads to the inference that each occasion meets an endpoint. as well as the habitual suffix. the habitual quantifier in and of itself does not say anything about the actual culmination of the event. put the other way around. . the aspectual component of quantification over occasions is structurally higher than the projection of inner Quantity. as seen in (88). an imperfect habitual leaves the predicate telic. In fact. a telic predicate is compatible with habitual interpretation. as discussed here. which. prepare the meal and write the report are heterogeneous predicates. Nevertheless. correspond to the opposition homogeneous/ heterogeneous.1. habituality implies that the action at stake has been undertaken again. In (89) and (90). Accordingly. such as Chierchia’s (1995) (see chapter 2.3). habitual interpretation is fine with atelic predicates. section 2. Likewise. Each occasion in which Juan wrote a report is heterogeneous. the habitual quantifier is predicted to have scope over the adverbial in x time. these predicates are not turned into homogeneous predicates. activity (homogeneous) predicates such as walk retain their nonquantity properties. as additivity and subinterval properties show: (89) “Prepare the meal” + “prepare the meal” ≠ “prepare the meal” (90) A subinterval of “prepare the meal” is not “prepare the meal” Under the imperfect-habitual quantifier.22 21 As I mentioned before.21 That is. Or. the telic modifier in x time and the habitual adverbial.Outer Aspect and Individual-Level Predicates 175 el informe en quince minutos (87) Normalmente Juan redactaba usually Juan write-PRET-IMPF-3SG the report in 15 minutes “Juan used to write the report in fifteen minutes” en la piscina (*en doce minutos) (88) Normalmente Pablo nadaba usually Pablo swim-PRET-IMPF-3SG in the pool (in 12 minutes) “Pablo used to swim in the pool in twelve minutes” As the well-formedness of (86) and (87) show. are not altered by the occasion-quantifier. I take this evidence to suggest that quantification over occasions (as I take habituality to be) is separate from quantity properties of predicates. Likewise. Inner-aspect properties. 22 The fact that habituality and inner aspect are proved to go separately is troublesome for proposals that emphasize the similarity between habituals (outer aspect) and states (inner aspect).

Bach (1986). As discussed back in chapter 3 (section 3. With the perfective. a perfective form does not make a homogeneous predicate heterogeneous. It is precisely this fact that has led many authors to argue that perfective is an “event type coercer” in the sense that it has the power to turn homogeneous predicates into telic ones. In other words. unlike states. a perfective form does not interfere with the quantity nature of the predicate.4). who conceive the perfective as a modifier that makes an eventuality telic. which leads to the inference of “completeness” and “boundedness. and others.23 Observe these examples: (91) Pablo nadó Pablo swim-PRET-PERF-3SG “Pablo swam” (92) Pablo caminó Pablo walk-PRET-PERF-3SG “Pablo walked” Verbs such as swim and walk test out as homogeneous predicates. it does not mean that he has been walking at every single moment of that interval. However.2 Perfective Homogeneous Predications I turn my attention now to homogeneous predicates appearing in perfective form. at least in Spanish. In the same vein as before. 24 23 . (95) *Pablo nadó en una hora Pablo swim-PRET-PERF-3SG in an hour “Pablo swam in an hour” Other authors. if we say John had that car from 1974 to 1985. In contrast. the TT is located ‘after’ the ET.176 Individuals in Time 5. Piñón (1995). de Swart (1998). If we say John walked from 2 to 3. activities are not 100 percent homogeneous at a micro-level.3. such as Bertinetto (2000).” as mentioned in the previous section.24 as subinterval and additivity proofs show. I will argue that. Contrary to Mourelatos (1978). (93) “Swim” + “swim” = “swim” (94) A subinterval of “swim” is “swim” In (91) and (92) the homogeneous predicates swim and walk appear in the perfective form. The meaning of these sentences is that Pablo was once involved in the task of swimming (91) and walking (92). draw a conclusion in the same direction. I argue that a perfective form does not trigger telicity effects on predicates. sentences such as these show that the presence of the perfective viewpoint does not contribute telicity. he has had that car at every single moment between 1974 and 1985.

Viewpoints may differ among themselves according to these two components: ordering and quantification over occasions. whereas habituality consists of a quantifier over occasions denoting a proportional plurality. However. in turn. I described the most basic inner-aspect properties as quantity properties. it does not convert a homogeneous predicate into a telic one. In conclusion. that progressive and perfective viewpoints refer to a number of occasions whose cardinality is one. the perfective in Spanish does not change the homogeneity of the predicate. by the ordering component of Aspect. Regarding outer aspect. encoding two types of information: the number of instances a predicate takes place. share the ordering component but differ in the quantification over occasions part. Habitual and progressive. I described it as a functional projection. Following Borer (2005). in and of itself. ‘after’ the event. I have been arguing that the (syntactic) presence of the functional projection of Quantity corresponds to heterogeneous predicates.” 5. the combination of the perfective with the adverbial for x time is not problematic. I will take the appropriateness of the adverbial in x time as a telicity test.4 A Brief Summary of Aspect Notions Thus far. as (97) shows. durante una hora (97) Pablo nadó Pablo swim-PRET-PERF-3SG for an hour “Pablo swam for an hour” I take these facts to indicate that. I argued. elaborating on Verkuyl 1999. I have distinguished two aspectual levels: inner aspect and outer aspect. As noted before. whereas its absence gives rise to homogeneous predicates. structurally higher than Quantity.Outer Aspect and Individual-Level Predicates 177 (96) *Pablo caminó en una hora Pablo walk-PRET-PERF-3SG in an hour “Pablo walked in an hour” Following the same line of reasoning as before. share the quantifier part but differ in the ordering component. and the ordering relationship between the interval a particular sentence makes an assertion and the interval during which a certain eventuality holds or takes place. As can be appreciated in (95) and (96). The way I propose to articulate this complexity (ordering and occasion-quantifi- . Technically. which is the origin of the intuition that the action is “complete. With respect to the quantification over occasions. the perfective does not play any role in the inner-aspect properties of a predicate. for example. adding such a modifier turns the sentences bad. with the perfective we just confine our assertion to a particular singular occasion and that occasion is located. The perfective and the progressive. which distinguish between homogeneous and heterogeneous predicates.

Specifically. (98) ZP 2 Zi VP 2 ei VP (99) AspP 2 TT Asp′ 2 Asp0 ZP (after/before/in) 2 VP Z0 (|1|.” perfective or progressive. 26 This aspect form is involved with stative SL predicates (She was sick). |>1|. |>1|. I would like to propose that the quantifier over occasions of the eventuality is the ZP itself (the internal argument for T in Stowell’s work). independently from the kind of predicate. as Kratzer (1988) originally proposed. presented in (3) above (repeated here as (98)) about the internal constituency of ZPs. and makes the prediction that the quantifiers (|1|. I used Spanish habitual and perfective data to show that aspectual viewpoints do not have any impact on the inner-aspect properties of predicates.4.25 The existential quantifier corresponds to the continuous imperfect. This unifies the working of Aspect. where there are no occasions to enumerate (think of stative homogeneous predicates such as Peter is tall).178 Individuals in Time cation) is depicted in (99). I argued that. 1996). which does not make the sentence habitual. in Spanish. ∃) should be possible with any type of predicate. this is borne out.1 for further discussion. As will be shown in section 5. the ordering component takes the quantified eventuality and orders it with respect to the TT interval. I assume to be present with every type of predicate (SL and IL). ∃) 2 e VP Thus. Elaborating on Stowell’s (1993) proposal.5. if “one.” habitual imperfect is at stake. 25 .26 The quantifier over occasions binds the eventive variable (e) which. following Stowell (1993. the perfective does not provide “quantity” to homogeneous predicates and habitual imperfect is not in compleRecall that an action can take place more than once (He hit him five times). but these predicates are also compatible with the habitual quantifier (She was sick usually). See section 5. and not only with SL ones. In the last section. When the quantifier denotes a number roughly describable as “greater than one.

to the behavior of adjectival IL predicates and states. Classical tests show that a well-circumscribed set behave as activities: specifically. as has been widely assumed. which confirms it as an activity and not as an accomplishment or achievement among event types. I established an aspectual division inside adjectival IL predicates. specifically. (101) shows the different behavior in the progressive form of Eskimo and blond. ?Sucedió que Juan fue muy cruel con Pablo happened that Juan ser-PERF-PRET-3SG cruel with Pablo “It happened that Juan was cruel to Pablo” . 5. and cruel. differently from the cruel-type. *Sucedió que Juan fue esquimal happened that Juan ser-PERF-PRET-3SG Eskimo “It happened that Juan was Eskimo” b. on the other. Finally. those that can have a complement introduced by a directional preposition.” which suggests that quantification over occasions and quantity (inner aspect) properties belong to different levels. consider just the following three sets of contrasts. together. (100) As a complement of happen (‘take place’) a. on the one hand.5 Adjectival Individual-Level Predicates and Viewpoint Aspect In chapter 3.Outer Aspect and Individual-Level Predicates 179 mentary distribution with “quantity. (102) shows the (perfect) entailment derivable from the progressive of the cruel-type. The next section continues the discussion of aspectual viewpoints by referring. Adjectives like Eskimo and other qualifying adjectives like blond behave. In (100). The copular sentences in Spanish showed that not all predicates in combination with the IL copula ser behave as states. the suitability as a complement of happen shows that the cruel-type in copular constructions can bear typical eventive properties. As a brief reminder. *Sucedió que Juan era rubio happened that Juan ser-IMPF-PRET-3SG blond “It happened that Juan was blond” c.

Consider the following cases: esquimal (103) Pablo era Pablo ser-IMPF-PRET-3SG Eskimo “Pablo was Eskimo” rubio (104) Pablo era Pablo ser-IMPF-PRET-3SG blond “Pablo was blond” . imperfect. Juan ha sido cruel con Pedro Juan has ser-PAST-PART cruel to Pedro “Juan has been cruel to Pedro” I will concentrate now on the analysis of these adjectival predicates in copular constructions regarding outer aspectual information. *(En aquella ocasión) Juan estaba siendo esquimal that time Juan was ser-ing Eskimo “(That time) Juan was being Eskimo” b.3 (about the relation between quantity and perfective and imperfect viewpoint). I will explore the interpretation that adjectival IL predicates have under the aspect forms mentioned thus far (perfective. *(En aquella ocasión) Juan estaba siendo rubio that time Juan was ser-ing blond “(That time) Juan was being blond” c. (En aquella entrevista) Juan estaba siendo muy cruel in that interview Juan was ser-ing very cruel con el entrevistador to the interviewer “(In that interview) Juan was being very cruel to the interviewer” (102) Perfect entailment from the progressive form a. Juan está siendo cruel con Pedro Juan is ser-ing cruel to Pedro “Juan is being cruel to Pedro” b.180 Individuals in Time (101) Progressive form a.1 The Imperfect and Adjectival IL Predicates In this section I discuss the different interpretations (continuous or habitual) that an imperfect form can have with copular adjectival constructions depending on the adjectival predicate at stake. and progressive). This will lead me to discuss the restriction existing between the progressive and the property of stativity mentioned in chapter 4. This restriction differs from the one discussed in section 5. 5.5.

the absence of habitual interpretation with IL predicates such as Eskimo or blond can be explained by their inability to restart. that the structures corresponding to (103) and (104) involve an existential quantifier. but the property is attributed to the person as a whole. although the habitual paraphrase is not the most salient one when (IL) stative verbs (such as Eskimo or blond) are at stake. whereas (105) can have it: (106) *Normalmente Pablo era esquimal/rubio usually Pablo ser-IMPF-PRET-3SG Eskimo/blond “Pablo used to be Eskimo” cruel (107) Normalmente Juan era usually Juan ser-PRET-IMPF-3SG cruel “Juan used to be cruel” Habituality entails. habituality entails that a given eventuality can hold or take place more than once. That is to say. it seems.Outer Aspect and Individual-Level Predicates 181 cruel (105) Pablo era Pablo ser-IMPF-PRET-3SG cruel “Pablo was cruel” The imperfect form in (103) and (104) does not have a paraphrase expressing habituality. I argued. I want to show that. successive initial points and successive (inferred) ending points. Thus. where we are not counting the number of times an eventuality holds. it is not the case that they are incompatible with a quantifier over occasions by . I propose. It seems. that states are incompatible with such quantification. therefore. at least on a first approximation. that a Q<occ> with a value of greater than one is not present in (103) and (104). (108) AspP 2 TT Asp′ 2 Asp ZP (within) 2 Z0 V (∃) 2 e VP The structure of (108) corresponds to the interpretation of a continuous imperfect. However. then. Since I argued that be Eskimo and be blond pattern with states.

the sentences become less odd. such as blond.” That is. though. since it refers to a cyclic. even predicates covering an entire lifetime can be treated as iterative. I cannot undertake here. are quite natural in contexts such as (i). Li used to be Chinese” In principle. where the adverbial in summer activates a habitual interpretation.27 This special situation of multiple lives is required to license a habitual form because the only way lifetime properties can be iterated is by iterating the lifetime interval of the individual. that this point may deserve more discussion. since we can establish the temporal bounds of each “existence” and. When. that the structure of (108) is not the only one they can have). However. these properties hold. Juan was a very cruel person” I am assuming that every time one is born counts as a different “individual. which suffices to license the habitual paraphrase (107) ‘Juan used to be cruel in his interactions with people’. which. Consider the following scenario: chino habitualmente (109) En sus encarnaciones humanas. Other suitable contexts for (105) can be the following: (una persona) muy cruel (110) Antes de hacerse pacifista. multiple lives are not just different stages of the same individual. this is not the only meaning that the imperfect can have with be cruel. However. time. in (109) there is an individual who has lived more than once. Li era in his incarnations human Li ser-IMPF-PRET-3SG Chinese usually “In his human incarnations. habituality is excluded with permanent predicates (such as Chinese or Eskimo) because these predicates do not denote eventualities that can start several times.182 Individuals in Time definition (that is.28 For a habitual interpretation to be (somewhat) acceptable. unfortunately. Juan era before becoming pacifist Juan ser-PRET-IMPF-3SG a person very cruel “Before becoming a pacifist. the whole big interval containing (or overlapping) with the interval of the eventuality has to be repeated. rubio en verano (i) Juan era Juan ser-IMPF-3SG blond in summer 27 . As can be seen. hence repeatable. they permanently hold. inside the existence of an individual. as a consequence. The predicate in (105) be cruel is not permanent and can restart multiple times. I am aware. If we create a situation where permanent properties can re-hold because their subject can be reborn (in possibly different circumstances). 28 Note that IL predicates not referring to necessarily permanent properties. As I will amplify in the next chapter. of each permanent predicate. the interval over which permanent predicates hold totally overlaps with the interval of the life of the individual they are predicated of.

(115) Pablo paseaba Pablo walk-IMPF-PRET-3SG (116) “Pablo used to walk” . in a similar way as we say John was blond or John was Eskimo. the only reading is the habitual reading (114). as in (113). the habitual reading is the most salient. When the PP complement is present. That is.Outer Aspect and Individual-Level Predicates 183 (una persona) muy cruel. The imperfect form can involve different quantifiers over occasions (Q<occ>): (112) a. The two types of paraphrases (habitual and continuous) are possible with be cruel and each corresponds to a different viewpoint structure. Habitual AspP 2 TT Asp′ 2 Asp Q<occ> (within) 2 >1 b. por eso lo detestaban (111) Juan era Juan ser-PRET-IMPF-3SG a person very cruel for this him hated sus hermanas his sisters “Juan was a very cruel person. when the complement inducing the activity-like properties is entered into the structure. Nonhabitual (continuous) AspP 2 TT Asp′ 2 Asp Q<occ> (within) 2 ∃ Consider now the viewpoint interpretations when the PP complement is present. cruel con Pedro (113) Juan era Juan ser-PRET-IMPF-3SG cruel to Pedro “Juan was cruel to Pedro” cruel con Pedro (114) Normalmente Juan era usually Juan ser-PRET-IMPF-3SG cruel to Pedro “Juan used to be cruel to Pedro” This result dovetails with the fact that the most salient interpretation of activities in imperfect is the habitual reading shown in (116). that is why his sisters hated him” Contexts (110) and (111) represent scenarios where the property is attributed to the person as a whole.

make the habitual reading available. although the progressive form is correct with cruel. (122)). when a relational complement is added (123). the progressive is not a paraphrase saliently available from the imperfect form. 29 . the progressive paraphrase of the imperfect is more natural in these contexts. whereas activities in imperfect (117) can also have a progressive paraphrase (118). see section 5.3. when a complement that can act as a “distributee” is added (125). 30 For the reasons why the progressive is excluded with Eskimo and blond. as noted in section 5. (119) and (120)).184 Individuals in Time It is worth noticing that.2. as noted above. However. which. Likewise. a habitual interpretation emerges. rather than the habitual (cf. the preferred interpretation of adjectives denoting intellectual aptitudes (121) is the continuous reading. this is just marginally available with be cruel (cf. they acquire activity-like properties. estúpido/inteligente (121) Juan era Juan ser-IMPF-PRET-3SG stupid/intelligent “Juan was stupid/intelligent” (122) ??Juan used to be stupid/intelligent muy estúpido con su hermano (123) Juan era Juan ser-IMPF-PRET-3SG very stupid to his brother “Juan was stupid to his brother” (124) “Juan used to be stupid to his brother” The examples are based on temporal subordinate clauses because.30 Finally. in Spanish.5. accordingly.29 (117) Pablo leía mientras María paseaba Pablo read-IMPF-PRET-3SG while Maria walk-IMPF-PRET-3SG (118) Pablo was reading while Maria was walking (119) Pablo era cruel con Pedro mientras este era simpático con él (durante la entrevista) Pablo ser-IMPF-PRET-3SG cruel to Pedro while he ser-IMPF-PRET-3SG nice to him (during the interview) (120) ??Pablo estaba siendo cruel con Pedro mientras este estaba siendo simpático con él (durante la entrevista) Pablo was being cruel to Pedro while he was being nice to him (during the interview) That is.1.

chino habitualmente (127) En sus reencarnaciones. the objects (the house. accomplishments (131). nonstative IL (128). the typo) are subject to referential variability for habitual interpretation to be possible. stative SL (129). and any eventive predicate. or achievements (132). the habitual viewpoint can be said to be compatible with any kind of predicate—stative IL (127). activities (130). Juan era in his reincarnations Juan ser-IMPF-PRET-3SG Chinese habitually “In his reincarnations. Juan used to be Chinese” cruel con Pedro (128) Habitualmente Juan era habitually Juan ser-IMPF-PRET-3SG cruel to Pedro “Juan used to be cruel to Pedro” enfermo habitualmente (129) Juan estaba Juan estar-PRET-IMPF-3SG sick habitually “He used to be sick” (130) Habitualmente paseaba habitually walk-PRET-IMPF-3SG “He used to walk” la casa en diez días (131) Habitualmente construía habitually built-PRET-IMPF-3SG the house in ten days “He used to build the house in ten days” la errata a la primera (132) Habitualmente encontraba habitually find-PRET-IMPF-3SG the typo in his first attempt “He used to find the typo at once” It is important to note that in the case of accomplishments and achievements.Outer Aspect and Individual-Level Predicates 185 ingenioso en sus bromas/inteligente (125) Juan era Juan ser-IMPF-PRET-3Sg cunning at his jokes/ intelligent en los negocios at his business (126) “Juan used to be cunning at his jokes/intelligent at his business” Summarizing. .

Li was a poisonous snake in an hour” (136) *En sus encarnaciones humanas.2 The Perfective and Adjectival IL Predicates I will start by discussing the interpretation of the perfective with permanent predicates. Li fue una serpiente venenosa en una hora in a previous life Li ser-PERF-PRET-3SG a poisonous snake in an hour “In a previous life. Regarding nonpermanent predicates. compatible even with this kind of predicate.5. . as argued before. in principle. Be Eskimo and be a poisonous snake are homogeneous predicates. Li was a poisonous snake” (134) Again. Note that an in + time adverbial is excluded with the perfective as well as with the habitual imperfect: (135) *(En una vida anterior). regardless of the aspectual viewpoint at stake. if an appropriate context is created. Li fue in a previous life Li ser-PERF-PRET-3SG Eskimo “In a previous life. Li fue in a previous life Li ser-PERF-PRET-3SG a poisonous snake “In a previous life. The perfective form does not make a homogeneous predicate heterogeneous or telic. However.186 Individuals in Time 5. or Juan was very intelligent in doing something (139). Permanent predicates are said not to combine with the perfective form. Li era habitualmente chino in his human incarnations Li ser-IMPF-PRET-3SG usually Chinese en una hora in an hour “In his human incarnations. Li was Eskimo” una serpiente venenosa (En una vida anterior). a viewpoint based on a Q<occ> = |1| is. the most salient reading with the perfective (137) is the one where we are referring to a time when Juan was cruel to someone (138). we observe that they are fully natural with the perfective. Li used to be Chinese in an hour” The unacceptability of (135) and (136) provides further evidence that viewpoint aspect (at least habitual and perfective) does not alter the inner aspect of predicates. the following cases sound quite acceptable: (133) esquimal (En una vida anterior). they are true of every subinterval of that time. With respect to be cruel or be intelligent. If they are true of an interval of time. despite the overt absence of the PP.

the progressive is possible with other class of homogeneous predicates. such as processes: (143) Pablo estaba paseando Pablo was walking (144) Pablo estaba siendo muy cruel/estúpido con el entrevistador Pablo was being very cruel/stupid to the interviewer As pointed out extensively in the literature (since Aristotle). whereas processes take time and progress in time (although this progression in time does not mean they advance toward a culminating point). As can be seen. cruel/inteligente toda su vida (140) Juan fue Juan ser-PRET-PERF-3SG cruel/ intelligent all his life “Juan was cruel his entire life” 5. consistent with the description of the perfective above.Outer Aspect and Individual-Level Predicates 187 muy cruel/inteligente (137) Juan fue Juan ser-PRET-PERF-3SG very cruel/intelligent “Juan was very cruel/intelligent” (138) ‘Juan was very cruel to Peter’ (139) ‘Juan was very intelligent in that business’ However. the progressive in English works differently in some contexts.3 The Progressive and Adjectival IL Predicates I now turn my attention to the progressive viewpoint in Spanish. SL (141) and IL (142) statives are excluded with the progressive.31 The progressive viewpoint does not seem to work like the imperfect and the perfective. and therefore reject the expression of progression in time. according to which it is compatible with basically any type of predicate.5. (141) *Pablo estaba estando enfermo Pablo was estar-ing sick “Pablo was being sick” (142) *Pablo estaba siendo esquimal Pablo was ser-ing Eskimo “Pablo was being Eskimo” However. in the sense that the progressive is not compatible with every kind of predicate. . states hold in time but do not take time. it is also possible to have the stative version of these adjectives together with the perfective. 31 As mentioned in chapter 3.

” Dynamic eventualities progress in time.32. that is. footnote 4. progression in time is not possible for those eventualities that are “dense”—that is. such as Landman (1991). since estar enfermo ‘be sick’ and estar preocupado ‘be worried’ are states and. Kearns (1991) explains the incompatibility between the progressive and states by alluding to their property of essentially lacking onsets and culminations. However. as a result. can be argued to lack any input of energy. In fact. based on Carlson’s (1977) notion of “stage. Aside from the intuitive understanding of the correlation between dynamism and progression in time. does not allow us to use the progressive. the property of being divisive in different stages does not cut the line between the cases where the progressive is licensed and those where it is not. they are excluded in the progressive form. and. where the verb empezar ‘start’ refers to the onset of the predicate remain. strictly speaking. for example. eventualities with such a temporal structure that between every two temporal points. nondynamic eventualities do not. An eventuality can appear in the progressive if it is divisible in stages. Some authors. In a similar vein. examples such as the following. argue that progression in time is not possible for those eventualities lacking internal development.1) for the introduction of this concept. However. 33 32 (i) a estar enfermo con 30 años Juan empezó with 30 years Juan start-PERF-PRET-3SG to estar-INF sick “Juan started being sick when he was 30” Pedro empezó a estar preocupado por su hijo cuando empezó a suspender Pedro start-PERF-PRET-3SG to estar-INF worried about his son when started failing en el colegio at school “Pedro started being worried about his son when he started failing at school” (ii) 34 See chapter 2 (section 2. In some sense. followed by Bertinetto (2000). the idea that states lack onsets argues in the same direction than Comrie’s (1976) description that states are those situations which are not subject to any input of energy to hold.”34 alludes to “temporal stages” to describe the distribution of the progressive. a sickness. if we can point at it and say ‘it’s the same event in a further stage of development. According to these authors.” related to the concept of “movement.33 Landman (1992). unexplained under this view. a third point can be established. . it is difficult to find an explanation in more technical terms. accordingly. nevertheless. an event is a stage of another event if the second can be regarded as a more developed version of the first. The input of energy correlates with the onset of the predicate. as the ungrammaticality of cases like *Juan estaba estando enfermo ‘Juan was being sick’ shows. See also chapter 6. which. nevertheless. According to Landman. we can distinguish different stages in.188 Individuals in Time The property usually invoked to explain the differences regarding progression (advancement) in time is “dynamism.

35 Thus. we can also assert it of a third point between the two formerly designated. where TT contrast exists. if we can assert the activity and the state predicate of two given points. nonetheless. therefore. the combination of states with the predicate within does not seem troublesome. Quantity establishes a distinction among eventualities according to their mereological properties. I consider that a much more precise way to describe it in (technical) temporal terms is pending. as is known. as already mentioned. if absent it is homogeneous. do not impose a particular restriction on the inner properties of predicates: they distinguish neither between heterogeneous and homogeneous predicates nor between types of homogeneous prediIn the terms I am using here. the progressive chooses just the homogeneous dynamic eventuality. distinguishes between activities and states. whereas other aspectual viewpoints (habitual-imperfect and perfective) combine with both types of homogeneous predicates (stative and dynamic). 35 . it seems that it is the combination of the ordering predicate within plus the Q<occ> |1| that cannot apply over states. are also excluded with the progressive (e.. Thus. Klein (1994:42) hypothesizes that the progressive form is excluded when there is no TT contrast possible (i. none of the properties attributed to the progressive here directly explain its oddness with states. It is neither the predicate within nor the Q<occ> |1| that is incompatible by themselves with states. since nonpermanent predicates. the eventuality is heterogeneous.g. since this is the ordering predicate corresponding to imperfective forms. they can be expressed in progress. either. Thus.e. states (IL and SL) can be combined with the perfective. the predicate holds of the subject. States and activities share the same homogeneous structure (–Quantity). whose Q<occ> is also |1| and. If projected. contrary to what Landman and Bertinetto intimate. which is the property that. The notion of dynamicity is not obviously captured by the syntactic projection of Quantity involved (as assumed here) in the event structure. I consider that such mereological characteristics do not discern the property of dynamism. since both can be argued to embrace the subinterval property (see chapter 3). Although I believe that some of these notions (internal granularity. arguably. I will just make the observation that only dynamic eventualities can be expressed in the progressive and briefly discuss how such dynamicity may be encoded in the event structure. However. and. As shown above. from a determined TT we cannot infer that there is any TT’ for which the negated utterance would hold).Outer Aspect and Individual-Level Predicates 189 Furthermore. in the way it is invoked by these authors. That is. internal stages) are involved in the description of dynamic events and. by virtue of which we can predicate the event of every subinterval of a determined period. but activities possess something additional that “dynamizes the eventuality” and licenses the progressive. it is not clear that the notion of density. unfortunately. This is typically the case of permanent predicates: no matter what TT is at stake. activities can be considered “dense” and. However. as viewpoints. Klein’s idea is not explanatory. in their possible expression in progress. be sick). I cannot contribute any clarifying idea at this point. states lack but activities possess. In this sense.. Habitual imperfect and the perfective. Nor can the idea proposed by other authors such as Klein (1994) alluding to the nature of the TT be explained.

Juan estaba teniendo un viaje terrible Juan was having a terrible trip d. mereological properties play no role in licensing them. Whereas the object houses does not refer to anything involving temporal extension. some stative cases allowing for the progressive should be mentioned. may be evidence that have is not a real verb but behaves more like a copula or an auxiliary verb—where the main predicate is the complement of have. It is those referred things that “take place” that can be conceived as involving temporal extension. This property draws the line between (145a) and (145b–d). only in (145a) can we paraphrase have with ‘possess’. where the nature of the object matters.190 Individuals in Time cates. It seems. a trip. either. The restrictions affecting the acceptability of the progressive are not of mereological nature. that it is the nature of the DP object that matters. because they refer to things that develop through time. Pretty much along the lines of the previous examples. a heart attack. *Juan estaba teniendo casas Juan was having houses b. If there were no grammatical difference (of significance) between states and processes. As Tim Stowell (p. 36 . we could not predict that the progressive combines just with one kind of them.) points out. That is. or a day do. By the same token. Juan estaba teniendo un ataque al corazón Juan was having a heart attack c. these cases.36 However. the objects in these examples can be considered some sort of “event objects” (Dowty 1979). it is such temporal extension that can be conceived in progression.c. (145) a. Actually. Juan estaba teniendo un día horrible Juan was having a horrible day Besides the noun directly related to a verb (attack). and only states (SL and IL) are excluded. since states and activities pattern alike in that sense. Their defining characteristic is that they allow the progressive depending on the type of object they have. then. we can say that it is the nature of the restrictive complements (denoting something involving “temporal extension”) that licenses the progressive with adjectives like intelligent or cunning: siendo muy inteligente (146) ?Juan estaba Juan estar-PRET IMPF-3SG ser-ing very intelligent “Juan was being very intelligent” This argues in favor of a distinction between the two types of homogenous predicates.

Outer Aspect and Individual-Level Predicates 191 siendo muy ingenioso (147) ?Juan estaba Juan estar-PRET IMPF-3SG ser-ing very cunning “Juan was being very cunning” siendo muy inteligente en la entrevista (148) Juan estaba Juan estar-PRET-IMPF-3SG ser-ing very intelligent in the interview “Juan was being very intelligent in the interview’ siendo muy ingenioso en sus bromas (149) Juan estaba Juan estar-PRET-IMPF-3SG ser-ing very cunning in his jokes “Juan was being very cunning in his jokes” The interview or his jokes denote “event-objects” (Dowty 1979) that take time. therefore. Landman 1992. it has to be noticed that the progressive interferes in a particular way with the mereological properties of the predicates. Summarizing thus far. (150) *Juan estaba arreglando dos sillas en una hora Juan was fixing two chairs in an hour As a dynamic predicate. the expression of progression and the expression of telicity cannot cooccur. Vlach 1981. makes possible their conception in progress. before getting the endpoint) are incompatible because both . Specifically. Parsons 1990. among others. Although it seems intuitively clear that an adverbial expressing the amount of time invested in the whole event and the expression of the event in its progression (i. to fix two chairs is compatible with the progressive form (151) and. (151) Juan estaba arreglando dos sillas Juan was fixing two chairs (152) Juan arregló dos sillas en una hora Juan fixed two chairs in an hour However. I have argued that the properties enabling the use of the progressive are not of mereological nature and. do not come from the properties encoded in the aspectual functional projection of Quantity. as a result. it is compatible with an in + time adverbial (152). such as in + time. as a telic predicate. Bertinetto 2000. the progressive is incompatible with adverbials denoting telicity. the progressive viewpoint (in Spanish) is not a possible option with states.. which. However. This is known as the “imperfective paradox” (Dowty 1979) or the “progressive paradox” (see. Naumann & Piñón 1997. Kazanina & Phillips 2003). Delfitto & Bertinetto 1995.e. Asher 1991.

if it denotes a proportional plural number of instances.” The perfective viewpoint does not provide telicity by itself to the eventuality. we observed the so-called progressive paradox. I noted that the progressive is a viewpoint that. First. Throughout my discussion of viewpoints and different kinds of IL predicates. we observed that whereas the progressive is compatible with telic predicates (John was fixing two chairs). whereby the progressive viewpoint is incompatible with adverbials denoting telicity (in x time). Second. Simplifying a bit. . seems sensitive to internal properties of the event in two respects. The value of this quantifier gives different aspect interpretations. the interpretation is habitual. In this respect. it is not with adverbials entailing that the endpoint has been reached (*John was fixing two chairs in an hour). In particular. Likewise.192 Individuals in Time notions are contradictory.6 Summary of the Chapter In this chapter I studied some properties of viewpoint aspect. or the progressive alters them. whether. This poses one issue I do not discuss here—namely. This fact explains its compatibility with any kind of event (states and activities included). If it counts just one instance. 1995) and Demirdache and Uribe-Etxebarría (2000). I discussed the relation between inner-aspect properties and outeraspect ones. it can be progressive (if the ordering component has the meaning of ‘within’) or perfective (if the ordering component has the meaning of ‘after’). it does not alter the inner aspect (mereological) properties of predicates. habitual viewpoint can cooccur with the expression of telicity. or ‘within’) the time the speaker makes an assertion about (the TT) and the interval the whole event can extend over. I considered Aspect as a dyadic predicate that orders (‘before’. Inspired by Verkuyl (1999). viewpoint information and inner aspect information do not interfere. The two facts noted (the requirement of dynamicity and the incompatibility with the expression of quantity-telicity) make the progressive different from the (habitual) imperfect and the perfective. unlike the perfective and the (habitual) imperfect. That is. when the progressive is at stake. The differences among the perfective and the imperfect habitual (which do not have any impact on the quantity structure of predicates) from the progressive also leads us to the question as to whether the three of them are viewpoints in the same sense or at the same level. it remains open the question regarding the role that the progressive plays in the quantity (inner aspect) structure. I showed that the imperfect habitual is compatible with Quantity. Following Klein (1994. I proposed that Aspect includes a component quantifying over occasions. we can say that. the predicate actually involves its quantity properties. Again. that is. we saw that it is compatible only with nonstative eventualities. ‘after’. 5. Second. which proves that Quantity and imperfect are not in complementary distribution. I first concluded that Spanish perfective does not turn eventualities into “telic.

The two main points I would like to make here are the following. past. to previous accounts of the temporal interpretation of IL predicates. As I have shown. lifetime effects refer to the interpretation whereby the referent of the DP subject can be interpreted as ‘no longer alive’. those dubbed “lifetime effects” (Musan 1995). formerly discussed by Kratzer (1988. As mentioned in chapter 2. it does not influence the ultimate temporal interpretation of the sentence. in section 6.5 discusses the nature of the TT in greater detail. I assume tenses are syntactically represented in the node of Tense. In the first section I introduce one previous account for the particularities of the temporal interpretation of IL predicates. it is not the case that Tense has the same interpretive impact on all individual-level (IL) sentences.4 I show how contextual factors intervene in the rising of this temporal interpretation. This chapter is organized as follows. In section 6.Chapter 6 Tense and Individual-Level Predicates In this chapter I study another temporal realm—namely.2. The last section summarizes the conclusions. although it is not only the type of predicate that counts. . I will be concerned with the interpretive contribution that tenses (present. Section 6. First. structurally. the domain of Tense. I will make special reference to the work by Musan (1995. Second. and future) have on copular sentences with adjectival predicates. which. IL predicates offer different inner-aspect properties. in particular. whether or not the tense complex of the subordinate sentence is dependent of the main clause one does not influence the content of the TT and. I propose a division among (adjectival) IL predicates that correlates with different temporal interpretations. As it appears. it is the content of the Topic Time (TT) that ultimately decides the temporal interpretation of the sentence. since this tense form gives rise to one of the most important interpretive effects with IL predicates—namely. 1997) and give a specific syntactic treatment to capture the optionality of the mentioned effects even in those cases where the other factors (concerning the type of predicate and aspect form) are met. by exploring the interpretation of IL predicates in compound sentences. In this regard. Likewise. These two features suppose a difference with respect to previous accounts of IL predicates in general and. as a consequence. is located higher than the Aspect node. Throughout the discussion I will be chiefly concerned with the interpretation of IL sentences in the past tense.3 presents the discussion of lifetime effects and describes the properties the predicate has to involve to trigger them. Section 6. which has consequences for their temporal interpretation. 1995).

a spatiotemporal or eventive (<e>) argument (Davidson 1967). IL predicates IP 1 I′ 1 I VP 1 θ-subj Regarding temporal interpretation. temporariness and permanency is encoded in the argument structure of predicates: temporary predicates involve a spatiotemporal argument.2). she proposed that IL predicates applied directly to the subject. That is. 3 As I also said in chapter 2.1 Temporal Interpretation as a Consequence of Argument Structure: Kratzer (1988.1. Kratzer argues that SL predicates involve an extra argument in their argument structure—namely. Kratzer proposes that the spatiotemporal argument is syntactically represented.3 (1) a. just aims to capture Kratzer’s idea on the location of the external argument with respect to the maximal projection of the predicate. In particular. . The configurational difference between IL and SL predicates can be depicted roughly as in (1). SL predicates IP 1 I′ 1 I VP 1 <e> b.”2 Specifically. she argues that the eventive argument is the external argument of (SL) predicates. SL predicates involve an extra semantic operation. Furthermore. as Diesing 1992 proposes. 1995) Kratzer (1988. whereas SL predicates needed something “extra. 1995) is the first in establishing an explicit correlation between the temporal interpretation later dubbed “lifetime effects” and IL predicates. I will mention two points of this proposal. putting in direct correlation the presence of a spatiotemporal 1 2 For an earlier brief discussion on this phenomenon. Kratzer proposes that the differentiation between SL and IL predicates resides in their argument structure. the representation in (1).1). whereas permanent ones lack such an argument. see Anderson 1973. As said before in chapter 2 (section 2.1 I will briefly summarize her line of reasoning and show that there are cases it cannot account for. This author assumes that the factor that draws the line between the two types of predicates is “being temporary” versus “being permanent.1.194 Individuals in Time 6.” See chapter 2 (section 2. the “realization function. First.” Likewise. in the same vein as Carlson (1977). For Carlson (1977). I will not discuss Kratzer’s perspective about where the NP subject should be assumed to generate: either in a specifier of the VP (along the lines of the VP-internal subject hypothesis argued by Kitagawa 1986 and Koopman & Sportiche 1991) or in the specifier of the IP.

understands Tense (realized in I[nflection] in (1)) as a predicate expressing properties of spatiotemporal locations ‘is before now’. Second. (6) Aunt Theresa resembled my mother Although it may be assumed that if A resembles B then B resembles A. She assumes that Tense applies to the external element of the predicate. This way. ‘is after now’. Examples like (6). according to Kratzer. which suggests that it has applied to it. (4) [before now (<e>)] & [<e> (be sick)] Similarly. informally. Kratzer’s temporal account for IL predicates derives from two points: one.Tense and Individual-Level Predicates 195 argument and the type of predicate allows Kratzer to explain. the following contrast: (2) *Henry was French last week (3) Henry was sick last week Whereas IL predicates (be French) cannot be temporally modified. SL ones can. 1995). ‘is now’. that is. as appears in (6). (Kratzer understands resembling as a permanent property). That is. the interpretation obtained is like (5). In sum. (5) [before now (Henry)] By virtue of the direct application of past tense onto the thematic subject. are felicitous just if it is the DP subject the one that is no longer alive. the individual is fully located in the past and the reading that arises is that Henry is no longer alive. the interpretation to be obtained is. for example. Kratzer (1988. tense locates the external argument of the predicate. the temporal interpretation of the arguments points to an asymmetry between the two interpretations. This hypothesis was defended by additionally alluding to cases involving bidirectional predicates like resemble. and two. . like (4). Tense applies to the eventive argument in SL predicates and to the thematic argument in IL ones. by applying past tense to the external argument of an IL predicate. by applying a past tense to the Davidsonian argument of an SL predicate like be sick. the external argument of IL predicates is the thematic subject itself. Assuming with Davidson (1967) that modification of eventualities is made through the eventive-spatiotemporal argument. based on Lemmon 1967. the past tense has located ‘before now’ the thematic argument acting as a subject (the external argument). its absence would account for the impossibility of modification when the predicate is IL.

muy cruel con María (8) Juan fue Juan ser-PERF-PRET-3SG very cruel with María “Juan was very cruel to Maria” muy cruel con María (pero ya no lo es) (9) Juan era Juan ser-IMPF-PRET-3SG very cruel to Maria (but he is not anymore) “Juan was very cruel to Maria. Consider (7) as an example. but he is not anymore” rubio (de pequeño) (10) Juan era Juan ser-IMPF-PRET-3SG blond (when he was little) “Juan was blond when he was little” All of these sentences have the IL copula ser in the past tense. perfective viewpoint does not trigger lifetime effects. On the one hand. but in none of them is a lifetime reading available. Throughout this work I have suggested that not all IL predicates can be considered alike. I advanced that equating permanency to IL-hood is not accurate. Consider the following examples as an illustration. In section 6. I give an account for the way contextual factors can contribute to block the lifetime reading. (7) That day. Harry was from California— does not trigger a reading like ‘he is no longer alive’. That is. In chapters 3 and 4 I focused on the inner aspectual differences among them.196 Individuals in Time I will point out two kinds of facts (regarding temporal interpretation) that remain unaccounted for under Kratzer’s proposal as it stands. these interpretive effects can be easily neutralized. the past tense does not seem to locate in the past either the situation of his being from California or his lifetime. First.3 and section 6. In the next sections I will argue that the reasons for this differ. Intuitively. such as hers. The second point I want to mention is that the lifetime reading does not arise with any type of IL predicate. Harry and I arrived in the USA.4. The reading disappears. which is why they do not appear in (8). a purely syntactic approach. despite being parallel to the one mentioned by Kratzer Henry was French. On the other hand. so he did not have to go through the immigration process. Harry was from California. In chapter 2. In examples like (7). the past tense of Harry was from California refers back to the moment at which the speaker and Harry arrived in the USA. lifetime effects arise more easily when the predicate is just temporally limited . the second part—that is. 1997) noticed. as Musan (1995. However. based on the argument structure. predicts that the temporal reading of lifetime effects always arises.

I dedicate a part of chapter 7 to such discussion. as has classically been upheld. despite the fact that the predicate is not understood as permanent. Second. The following paragraphs show examples of copular sentences with ser and different adjectival predicates. Pablo es muy gracioso Pablo ser-PRES-3SG very funny “Pablo is very funny” muy gracioso b.” Consider. I take it to mean that the predicate is IL. I take the strong native intuitions about the different interpretation of adjectives that can combine with ser and estar to suggest that. Thus. Throughout this work. Regarding the temporal interpretation that gets the existence of the individual implicated (the lifetime reading). I will construct an analysis of the temporal properties of IL predicates. I am not forced to say that the nature of the predicate has mutated from IL into SL. since I argue that permanency is not a necessary characteristic of IL predicates. First. 6. I will consider it is not a forced consequence from IL argument structure but it is a reading possible just when a concrete number of factors meet. a result that proposals equating IL-hood and permanency are pushed to. when an IL predicate comes temporally or spatially restricted (some way or other). among other facts. without being able to explain. each of them involves a different semantics that can properly be described as corresponding to what Carlson (1977) denominated “individual level” and “stage level. the next contrasts: (11) a. I am expected to offer an alternative definition for IL predicates. I have consistently considered that every (copulative) predication with ser is IL.Tense and Individual-Level Predicates 197 by the existential limitation of the referent. This perspective leads me to two outcomes. I propose that being “nonpermanent” does not amount to being “stage level. Since be blond or be cruel does not have to overlap with the whole span of time that an individual’s existence extends over. after having introduced the relevant temporal distinctions. as a reminder. the effects are not borne out. Pablo está Pablo estar-PRES-3SG very funny “Pablo is being very funny” . why the Spanish copular verb does not change. In the remainder of the chapter. permanency or.” In other words. more accurately. in fact. “lifetime permanency” is a property of just a subset of IL predicates. if the predication with ser is all right. As before. As mentioned in chapter 2.2 Differentiating Temporal Extensions of IL Predicates In this section I argue that the span of time over which an IL predicate extends can vary depending on the type of property. I consider ser as the copula marking IL-hood.

198

Individuals in Time

muy guapo (12) a. Pablo es Pablo ser-PRES-3SG very handsome “Pablo is very handsome” muy guapo b. Pablo está Pablo estar-PRES-3SG very handsome “Pablo looks very handsome” moreno (13) a. Pablo es Pablo ser-PRES-3SG dark-skinned “Pablo is dark-skinned” moreno b. Pablo está Pablo estar-PRES-3SG dark-skinned “Pablo is tanned” In (11)–(13) several adjectives appear with ser and estar. All of them are correct with both copulas, although they get different meanings which contrast sharply. In the (a) cases, with ser, the adjectives are predicated of the subject as an individual. The speaker says that Pablo is a funny, handsome, or darkskinned person. In the (b) examples, with estar, the speaker predicates the property of the subject in a particular occasion. Pablo may be saying funny things this evening because he is in a good mood, which may happen very rarely. Pablo may look handsome because he is wearing a nice suit. And Pablo may look dark-skinned because he got a tan. All the (b) cases are perfectly compatible with Pablo being unkind and bitter as a person, or unattractive, or light-skinned. None of the following assertions are contradictory: nada gracioso, pero está muy gracioso (14) Pablo no es Pablo not ser-PRES-3SG at-all funny but estar-PRES-3SG very funny “Pablo is not funny at all, but he is being very funny” guapo, pero está muy guapo (15) Pablo no es Pablo not ser-PRES-3SG handsome but estar-PRES-3SG very handsome “Pablo is not handsome, but he looks very handsome” (16) Pablo es muy pálido, pero está moreno

Pablo ser-PRES-3SG very light-skinned but estar-PRES-3SG dark-skinned

“Pablo is very light-skinned, but he is tan” Taking all this into account, I will consider that the described contrast corresponds to the dichotomy IL versus SL, originally proposed by Carlson (1977). I take it that when ser is involved, I am dealing with an IL predication, no matter whether there appears to be temporal or spatial modification. If that

Tense and Individual-Level Predicates

199

occurs, instead of proposing a mutation from IL to SL predicate, I will try a different definition of IL and SL predicates (see chapter 7). I begin by showing a group of adjectives (and some PPs) in combination with ser that differ among themselves in their temporal extension. Some of them can be accurately considered lifetime properties, whereas others cannot. 6.2.1 Permanent IL Predicates Consider the following set of cases: (17) Pablo es esquimal/gitano/africano/de familia ilustre/ de baja estofa/ del grupo sanguíneo O+/daltónico Pablo ser-PRES-3SG Eskimo/gypsy/African/from an illustrious family/ from poor class/from O+ blood group/colorblind (18) *Pablo ha dejado de ser esquimal/gitano/africano/ de familia ilustre/de baja estofa/del grupo sanguíneo O+/daltónico Pablo has given up being Eskimo/gypsy/African/ from an illustrious family/from poor class/from O+ blood group/ colorblind (19) *En su juventud, Pablo era esquimal/gitano/africano/ de familia ilustre/de baja estofa/del grupo sanguíneo O+/daltónico In his youth Pablo ser-IMPF-PRET-3SG Eskimo/gypsy/African/ from an illustrious family/from poor class/from O+ blood group/ colorblind The predicates in (17), Eskimo, gypsy, African, from an illustrious family, from poor class, and from O+ blood group, or colorblind are properties of the individual that cannot stop holding, as the ungrammaticality of (18) shows, and cannot be restricted to a period of time, as (19) suggests. That is, these predicates hold of an individual from his birth to his death. They can be truthfully predicated of an individual at any time inside his lifetime. More strictly speaking, permanent predicates do have temporal limitations, but they “coincide” with the individual’s initial and final life points. Besides these predicates that totally overlap the lifetime of an individual, others such as those that, once acquired, last forever and cannot stop holding (Ph. D, mother of two children) are also permanent predicates. 6.2.2 Nonpermanent IL Predicates Not all predicates appearing with ser share the lifetime property observed in the previous section. Others, referring to contingent properties, such as physical appearance or attitudes, do not necessarily hold of the individual for his

200

Individuals in Time

entire lifetime. Nevertheless, all combine with ser and can be conceived as properties characterizing people as such. Consider the following cases: (20) Juan es rubio/muy guapo/muy dulce/accesible-de fácil trato/ generoso/altruista/egoísta/atrevido/miedica/valiente/criticón/ retorcido/sensible/soberbio/envidioso/pesado/servicial Juan ser-PRES-3SG blond/very handsome/very sweet/easygoing/ generous/ altruistic/egoistical/daring/fearful/brave/faultfinding/ twisted/sensitive/arrogant/envious/tedious/helpful (21) Juan dejó de ser rubio/muy guapo/muy dulce/accesible-de fácil trato/ generoso/altruista/egoísta/atrevido/miedica/valiente/criticón/ retorcido/sensible/soberbio/envidioso/pesado/servicial cuando se hizo mayor Juan stopped being blond/very handsome/very sweet/easygoing/ generous/altruistic/egoistical/daring/fearful/brave/faultfinding/ twisted/sensitive/arrogant/envious/tedious/helpful when he grew up (22) Cuando era pequeño, Juan era rubio/ muy guapo/muy dulce/accesible-de fácil trato/generoso/altruista/ egoísta/atrevido/miedica/valiente/criticón/retorcido/sensible/ soberbio/envidioso/pesado/servicial When he was a little boy, Juan ser-IMPF-PRET-3SG blond/ very handsome/very sweet/easygoing/generous/altruistic/ egoistical/daring/fearful/brave/faultfinding/twisted/sensitive/ arrogant/envious/tedious/helpful The adjectives in (20) refer to the physical appearance of individuals, such as blond or handsome and to properties of his personality, such as sweet, generous, easy-going, altruist, egoistical, daring, fearful, brave, faultfinding, twisted, sensitive, arrogant, envious, tedious, and helpful. As the good formation of (21) proves, all of them can stop holding of an individual. Also, as (22) shows, they can be asserted of an individual for a concrete and restricted period of his lifetime. That is, although they can hold of an individual for all of his lifetime, they do not have to. Holding of an individual for his entire lifetime is not a defining property of them, which is why lifetime permanency can be overridden easily without triggering oddity. Consider now more examples: (23) Pedro era muy cruel (de pequeño) Pedro ser-IMPF-PRET-3SG (when he was little) “Pedro was very cruel (when he was little)”

Tense and Individual-Level Predicates

201

(24) Pedro estaba siendo muy amable con Juan {durante la entrevista/ mientras le entrevistaba/hasta que consiguió lo que buscaba}, pero luego cambió de actitud Pedro was ser-ing very kind to John {during the interview/ while he was interviewing him/until he got what he was looking for}, but then he changed his attitude “Pedro was being very kind to John…” As noted in chapter 4 (section 4.1.3), adjectives referring to mental properties (such as cruel, mean, or kind) can define either an individual as such or an individual as the property gets manifested in a particular action to another person. I argued in particular that when relational APs (those that can have a PP complement expressing the goal of the action John was cruel to Peter) have such a PP complement present (covertly or overtly), the adjective is interpreted as attributing the property to an individual in a particular action. In the set of examples above, (23) exemplifies attribution of a property to a whole individual. As the temporal complement (when he was a little boy) suggests, an adjective like cruel does not denote a lifetime property, since it can be naturally restricted in time. Example (24), with the PP complement present, also shows how the properties these adjectives denote can be asserted to hold of an individual for a concrete period of time. As can be seen, they combine well with temporal adjuncts introduced by during or while. The temporal restriction in (23), in principle identical to the one in (21) and (22), may seem slightly different from the temporal restriction in (24). Whereas in (21)–(23) we are limiting the property of an individual, to some extent, to a wider period of time (when John was little, etc.), in (24) we are dealing with a more concrete moment (the interview, for instance). However, according to a well-established semantic tradition, it cannot be established that a qualitative relevant difference exists between larger and shorter intervals of time, since temporal relations are established independently of the length of the intervals. Intervals of time are alike in nature, independently of their duration.4,5 I would like to maintain, then, that the only difference among these adjectives resides, as explained in chapters 3 and 4, in their inner-aspect
4

This is attributable to the property of “density,” which, as commonly agreed among semanticists, time involves. The formal definition of this property of ordering relations is as follows: A relation R in a set A is dense if for every ordered pair <x, y>, Є R, x ≠ y, there exists a member z Є A, x ≠ z and y ≠ = z, such that <x, y> Є R and <z, y> Є R. Thus, according to this property, it is always possible to pick out an interval of time between two intervals of time. This amounts to saying that any interval is (sub-)divisible by definition, which gives no way to establish a relevant difference among intervals based on their length. 5 The only difference between “moments” and “intervals” would be quantitative. For discussion, see, among others, Allen 1983 and references therein.

202

Individuals in Time

properties. Those in (21)–(23) behave as states, whereas the ones in (24) as activities, but all of them share the property of not being lifetime properties, contrary to the adjectives in (17). I would also like to propose that all of them should be considered IL predicates. All of these adjectives, even when their application is restricted in time, attribute a property to an individual. The length of the interval the predicate holds over seems independent. 6.2.3 Brief Notes about the Complement of Mental Properties APs Before ending this section, I would like to consider the complement that can appear with the cited mental properties such as cruel or mean, which Stowell (1991) considered as an argument of the adjectival predicate in English. For Stowell, the infinitival clause appearing in examples like (25) and (26) was an additional (optional) argument that such APs could have, besides the nominal argument. (25) Peter was very cruel to ridicule John in front of everyone (26) Pedro was very kind to walk me home There are two differences between English and Spanish regarding these cases. The first is that, in English, the PP goal complement cannot co-occur with the infinitival, whereas in Spanish it can (see (27)). The second is that the semantic and the syntactic status of the clausal complement in Spanish is tricky to characterize. It cannot appear as an infinitive as it does in English but must be introduced by the preposition a ‘to’ plus the definite article el ‘the’ (=al). Contrast (28) and (29). (27) a. Pedro fue muy cruel (con Juan) al ridiculizarlo delante de todo el mundo b. Pedro was very cruel (*to Juan) to ridicule him in front of everyone (28) Pedro fue muy amable (*acompañarme a casa) Pedro was very kind (to walk me home) (29) Pedro fue muy amable al acompañarme a casa Pedro was very kind to-the walk-me home “Pedro was very kind in walking me home” Although I leave the analysis for the infinitival clause in Spanish (whether it can be considered a true argument, etc.) for later research, I bring it up now because, as acknowledged in the literature (Hernanz 1999, García 1999, and references therein), one of the meanings of al + infinitive clauses is temporal

Juan se rompió el pie In rolling down the stairs Juan broke his foot (35) a. at least under this examination. as (35b) shows. (34) a. similar to the during + DP or while-clauses above. However. Hernanz (1999) points out that temporal infinitives can freely appear in either initial or final position (see (34)). Pedro ser-PERF-PRET-3SG very kind From the facts in (31) and (35b) I conclude that the al + infinitive clause of the relational AP cases. if it does not. Pedro fue muy amable al acompañarme a casa Pedro ser-PERF-PRET-3SG very kind in walking me home b. Juan se rompió el pie al rodar por las escaleras Juan broke his foot in rolling down the stairs b. does not work as a temporal adjunct. it is proved it acts as a temporal modifier. one way to test whether an al + infinitive clause has a temporal meaning (concretely. it cannot be concluded. Whereas the property denoted by the adjective can be said to be restricted to the individual as he is involved in the action denoted by the infinitival clause. that of ‘simultaneity’) is by asking “when the eventuality designed by the predicate took place. Pedro fue muy amable In walking me home.Tense and Individual-Level Predicates 203 and I want to discuss whether such “infinitive clauses” act as a temporal adjunct restricting the interval of time the property denoted by the adjective holds of the individual.” If the al + infinitive clause constitutes a satisfactory answer. Al rodar por las escaleras. . as the # symbol in (31) aims to capture. (30) Pedro fue muy amable al acompañarme a casa Pedro ser-PERF-PRET-3SG very kind in walking me home (31) ¿Cuándo fue amable Pedro? #Al acompañarme a casa When ser-PERF-PRET-3SG Pedro kind? In walking me home (32) Juan se rompió el pie al rodar por las escaleras Juan broke his foot in rolling down the stairs (33) ¿Cuándo se rompió el pie Juan? Al rodar por las escaleras When did Juan break his foot? In rolling down the stairs Whereas the al + infinitive clause is a natural answer of a when-clause in (32) and (33). According to García (1999). that the infinitival clause (in Spanish) spells out the interval of time the property holds. it is not so in the adjectival cases. #Al acompañarme a casa. this is not a possibility for al + infinitive clauses in the copular sentences at stake. the status of the clause is not temporal.

It also has an interpretation very close to the one exhibited with ser. this discussion at least shows the nontemporal nature of these clauses. whose paraphrase appears in (ii). it is necessary to point out that this is not the only meaning the al + infinitive clause has with estar. Pedro estuvo muy amable al acompañarme a casa Pedro estar-PERF-PRET-3SG very kind in walking-me home b. Pedro estuvo muy amable In walking-me home. (i) Al ser tan tarde. the al + infinitive clauses can behave as true temporal adjuncts. which would be along the lines of a causal adjunct. According to Hernanz (1999). the cases under discussion cannot be paraphrased by a temporal nor a causal clause: (iii) Pedro fue muy amable al acompañarme a casa Pedro was very kind in walking me home (iv) #Pedro fue muy amable cuando me acompañó a casa Pedro was very kind when he walked me home (v) #Pedro fue muy amable porque me acompañó a casa Pedro was very kind because he walked me home The absence of adverbial paraphrases raises the question whether they are true adjunct clauses. when the copular verb is the SL estar. Pedro estar-PERF-PRET-3SG very kind Both the answer to a when-question and the position-alternation test suggest that al + infinitive clauses work as temporal adjuncts when the copular verb is SL.204 Individuals in Time To the contrary. it is typical of estar. Al acompañarme a casa. 7 The meaning of the al + infinitive clause in these cases is not easy to capture. like that in (i). I did not want to ring you up (ii) No quise llamar porque era muy tarde I did not want to ring you up because it was very late However. the linking to the particular situation is emphasized. 6 . (36) Pedro estuvo muy amable al acompañarme a casa Pedro estar-PERF-PRET-3SG very kind in walking me home (37) ¿Cuándo estuvo amable Pedro? Al acompañarme a casa When estar-PERF-PRET-3SG Pedro kind? In walking-me home (38) a. no quise llamarte In being so late. With estar.7 However.6 Although I cannot investigate further the status and meaning of al + infinitive clauses with ser here. As I intimated in chapter 2. al + infinitive clauses are interpreted as temporal as discussed above or causal clauses.

Second.3.1 Necessary Conditions for Lifetime Effects First. I consider that the fact that some predicates refer to properties that hold for the entire lifetime of an individual is an interpretive outcome from their lexical meaning. Musan (1995) does. such as the presence of another past tense around. do not have to be permanent properties. For the interpretation of the DP subject as ‘no longer alive’ to be an option (in languages like Spanish). as.. The lack of permanency has been proven by the correctness of sentences with temporal adjuncts restricting the span of time the property was said to hold of the individual. 6. As already mentioned. in their lexical entry. literally.2. That is.Tense and Individual-Level Predicates 205 6. In the following section. and (c) the past form should be in the imperfect form. according to independent grounds (such as their combination with a particular lexical choice of the copula in languages like Spanish). First. these three conditions have to be met: (a) the predicate has to be an IL lifetime predicate.2 I have made two main points in this section. I argued that those accounts. 1997) observed that contextual factors. Ph. I am going to argue. compare the following sentences: . for example. 1997).3 The Arising of Lifetime Effects In this section I survey the conditions under which a lifetime reading can arise and introduce the factors that intervene in the arising of this interpretation. I want to briefly enumerate the conditions that are necessary (not sufficient) for the lifetime reading to appear. can neutralize the lifetime effects. I argue that lifetime effects are a salient option in those cases where the predicate.). e.g. 6. More accurately. since there is a large number of predicates that. in the line of Musan (1995. according to which the temporal interpretation known as lifetime effects derives from the argument structure of IL predicates. Differing from Kratzer (1988. I do not consider that those predicates encode. nevertheless. (b) the predicate has to appear in the past form. can be argued to be IL and. 1995).4 Summary of Section 6. Musan (1995. necessarily produce an inaccurate overgeneralization. In support of the first claim. I focus on another restriction for this temporal interpretation. whether they denote a lifetime property.D. holds of the individual for all his lifetime (mainly those referring to the origin and genetic nature of beings and to those properties that once acquired cannot be lost. equating IL-hood to property stability is inadequate. for lexical reasons. that lifetime effects are subject to certain contextual conditions.

The future tense does not yield the interpretation of (43). In the remainder of the discussion I will not make any other allusion to future tense cases. in the present. but can give what can be called a “forward-lifetime effect” (46). However. an interpretation like (43) is available when the predicate is a lifetime IL predicate. either. does not activate the reading in (43). In (40). (43) Pedro is no longer alive The second necessary condition is that the predicate must be tensed in the past. such as be Eskimo. with an IL predicate that does not denote a necessarily lifetime property. This property will hold of its subject all through his life since it refers to his origin itself and cannot be modified or lost. esquimal (44) Pedro es Pedro ser-PRES-3SG Eskimo “Pedro is Eskimo” californiano (45) Pedro será Pedro ser-FUT-3SG Californian “Pedro will be Californian” (46) Pedro is not born yet . A sentence like (44).206 Individuals in Time enfermo (39) Pedro estaba Pedro estar-IMPF-PRET-3SG sick “Pedro was sick” muy cruel (con sus amigos del colegio) (40) Pedro era Pedro ser-IMPF-PRET-3SG very cruel (to his schoolmates) “Pedro was very cruel to his schoolmates” esquimal (41) Pedro era Pedro ser-IMPF-PRET-3SG Eskimo “Pedro was Eskimo” Example (39) involves a typical SL predicate (be sick) and a lifetime reading (42) does not arise: (42) #Pedro is no longer alive The same happens in (40). the predicate is understood as referring to school time.

in consonance with the general agenda of the work. why does the lifetime reading arise with the imperfect form? In chapters 4 and 5 I claimed. Although I have used. but this form is not excluded with them per se. there are three necessary conditions for the lifetime reading to be available. (The slashes represent the TT.” In sum. only examples with the copular verb.Tense and Individual-Level Predicates 207 Finally. I need to first introduce as an assumption something I will explain in more detail in the next section. following Klein (1994. esquimal (47) El príncipe Li era the Prince Li ser-IMPF-PRET-3SG Eskimo “Prince Li was Eskimo” esquimal (en su tercera vida) (48) El príncipe Li fue the Prince Li ser-PERF-PRET-3SG Eskimo (in his third life) “Prince Li was Eskimo (in his third life)” (49) Prince Li is no longer alive To account for the reason why only imperfect forms give rise to the lifetime reading. In (50) and (51) both situations are depicted. the perfective brings about the meaning that the individual has “passed” the span of time the predicate extends over. there are . and will keep on using. that viewpoint aspect is an ordering predicate. I will argue that the lifetime reading arises when the TT at play is the interval that an individual’s lifetime extends over. perfective locates the TT ‘after’ the span of time the predicate can extend over. As mentioned in the previous chapter. The arising (or not) of the lifetime reading hinges on the value that the TT of the sentence has. the span of the individual’s life and the span of the predicate coincide. Whereas imperfect locates the TT ‘within’ the span of time the predicate can extend over. Based on Demirdache and Uribe-Etxebarría’s work. Now. 2004). The interpretation in (49) is available from (47) but not from (48). If he has over-passed it. the individual need not be understood as “dead. lifetime predicates are not usually found in the perfective form.) (50) ------{//////--------------------(51) -----{------}//////----------------Imperfect: within Perfective: after With the imperfect. The first condition is that the predicate must denote an IL lifetime property. the sentences that prototypically trigger the reading come in imperfect form. However. I argued that the difference between imperfect and perfective was in the ordering predicate they denote. bearing this in mind. 1995) and Demirdache and Uribe-Etxebarría (2000.

In the framework she assumes. since it was a consequence of their argument structure. I have already shown that the equation this author establishes between permanency and IL-hood is not accurate and I have suggested that the lifetime reading is expected to be salient when a lifetime property is at stake. In this section I show that even when the predicate denotes a lifetime property. 6. the lifetime reading does not necessarily come out. . This author proposed to consider in contrast sentences like (52).3. Westerståhl (1984) and Lewis (1986) observed that domains of quantification are contextually restricted. This was first noted by Musan (1995. Gregory was from America and Eva-Lotta was from Switzerland. whereas a lifetime reading is available in (52). As Musan observes.8 Thus. as Kratzer’s examples were. Musan (1995. The third condition is that the viewpoint form must be the imperfect.208 Individuals in Time other verbs appearing in lifetime predications. despite the fact that the sentence. To account for this fact. it can be restricted by the context through a variable C. I will show a number of situations where such a reading does not arise. 8 As mentioned in the previous chapter. tense is taken as a sentence operator. (52) Gregory was from America (53) On that day. Then. is exactly the same. 1997). Although I am abstracting away from the mechanics of the semantic formalization. Kratzer’s (1995) account predicts a lifetime reading to be available with IL predicates in all circumstances. The second condition is that the tense form has to be the past. Precisely. it is not in (53). roughly. in the next sections.1. like (53). Musan’s idea can be informally depicted as in (54). Musan argues as follows. 1997) As noted earlier. The sentence Gregory was from America is given a specific temporal context in which it is interpreted temporally. and others with a previous context. Gregory was from America. ‘the day I was introduced to Gregory and Eva-Lotta’. as any other operator. such as have brown eyes or have long legs. I was introduced to Gregory and Eva-Lotta.4. the temporal operator of Gregory was from America in the second case is restricted by the previous context. Both examples here are from Musan (1995).2 Introducing the Determining Role of Contextual Factors. I return to this point in section 6. “out of the blue” cases. I would like to emphasize that these are not sufficient conditions for the lifetime reading to arise.

it is the different value of the contextual restrictor C that gives the different possibilities for lifetime readings. Specifically. Musan assumes that past and present tense differ with regard to their “informativeness. That is. the hearer would react by adding: “…and he still is from America. Regarding what makes the lifetime reading arise. the NP subject itself is able to supply a value for C. Specifically. (56) Utterance: “Gregory was from America” (57) Situation: Gregory is still alive (58) Situation: Gregory is dead If someone utters (56) when the situation is (57).. NP subjects provide the time of existence of the individual they denote as a value for C.e. On my view. In my interpretation of Musan’s proposal.” An utterance such as (56) is appropriate if the situation is the one in (58) but not acceptable if it is as in (57). this could have been enough.” Musan argues that. given the presence of such a contextual restriction. when sentences are uttered out of the blue and the variable C restricting the temporal operator lacks an interval explicitly supplied from the context. Musan does not seem to take this as a definitive explanation and argues that pragmatic considerations play a role in determining the arising of the reading in “out of the blue” cases. Musan assumes that lifetime readings arise when the sentence is in a situation of temporal un-specification. she alludes to the Maxima of Quantity of information proposed by Grice (1975). the sentence is evaluated with respect to that interval. However. it is the value of C that accounts for the arising or neutralization of the lifetime reading. That is. (55) Temp Operator [C] (Gregory was from America) ↓ [Gregory’s time of existence] Since the interval of Gregory’s existence acts as a restrictor.” Musan takes this as a proof . Musan (1995:56) argues that in temporally unspecific contexts (i. Musan says.Tense and Individual-Level Predicates 209 (54) Temp Operator [C] (Gregory was from America) ↓ [the day I was introduced to Gregory and Eva-Lotta] The sentence Gregory was from America is going to be temporally evaluated with respect to the contextual restriction “the day I was introduced to Gregory and Eva-Lotta. the lifetime reading gets neutralized. “out of the blue” sentences).

this approach will have to establish how the TT gets its content. which will be defined as a ZP sensitive to contextual content. Crucially. (59) María y Harry llegaron a EEUU. In sum. then. the speaker is not asserting that ‘Harry had the property of being from California at a past interval’. in the remainder of the chapter I develop a proposal according to which the occurrence of the lifetime effect depends on the content of the TT of the sentence. The different temporal interpretations would ensue. the hearer assumes that the speaker is being cooperative and as informative as necessary.3. as Musan herself proposes it has (the previous context in neutralized cases. Take (59). depending on the value of the TT. From my point of view. in this case. as in any other case. I agree with the essence of one part of Musan’s (1995) account.210 Individuals in Time that. 6. What has to be clarified from this perspective are the factors intervening and deciding the content of the contextual variable. así que no tuvo que pasar la aduana Maria and Harry arrived in the USA. the past tense seems to be referring to the interval of time mentioned in the previous . informative considerations play a role in lifetime arising but not in lifetime neutralization. the individual’s existence interval in the nonneutralized ones).3 The Content of the TT in the Arising of Lifetime Readings As noted. and assuming he is being maximally informative with respect to the temporal duration of Gregory’s being from America. the hearer would understand that Gregory is dead. where it is impossible to combine a present tense with the contextual restriction. Clearly. Thus. In general. I am going to present it now in the theoretical terms I have been making use of thus far with the help of an example. suffices to predict the arising or blocking of the lifetime reading. there is no way of expressing a more informative proposition about the duration of Gregory’s being from America. Harry ser-IMPF-PRET-3SG from California. similar to Musan’s (53). If the speaker is using the past tense instead of the present. This is the line I would like to pursue here. according to Musan. Following the framework introduced previously on Tense and Aspect. Musan acknowledges that informativity considerations do not matter in cases like (54). this supposes an asymmetry in the account for the arising versus the neutralization of the lifetime reading. The different content of the contextual variable C. Harry era de California. given the mismatch between the utterance in (56) and the situation of (57). Rather. so he did not have to go through immigration The crucial question here is to determine what the past tense of Harry was from California refers to. since it contains a past interval. the past tense is less informative than the present tense.

As I introduced in chapter 5 (section 5. Thus. Likewise.2). this interval is called “Topic Time” (TT) (Klein 1994). no lifetime reading is available. but the interval the speaker is referring to. here assumed to be the Utterance Time (UT) (i. From this. since from the perspective I propose here it is not the case that the reading arises and then it becomes neutralized. it can be observed that the value of the TT is decisive in the arising of such readings. In the spirit of Musan (1995).” However. as Stowell (1993) proposes for the other two (see chapter 5. . section 5. in the past). In this case.1. I thank Tim Stowell for this remark. since we have proved that it is the content of the TT that matters.1).Tense and Individual-Level Predicates 211 sentence—namely. 1997) talks about “the neutralization of lifetime effects. I assume that the TT is a time-denoting phrase (ZP). Tense is locating in the past not the whole interval the eventuality (be from California) holds.9 What needs to be defined is which content gives rise to the lifetime interpretation.. the interval at which Maria and Harry arrived in the USA. the temporal structure of Harry was from California of (59) can be depicted as follows: (60) (UT) RT TP 3 T′ 2 T AspP (after) 2 TT Asp′ (arrival in the USA) 2 Asp VP (within) 2 ET VP (being from California) The structure in (60) says that the TT. In other words. I propose that the lifetime reading arises when the TT is defined as the interval of the individual’s existence: 9 Musan (1995. is located within the span of time at which being from California holds and after the RT. I assume that the nature of the TT is not different from other intervals like Reference Time (RT) and Eventuality Time (ET).e. lifetime effects can be reasonably considered a phenomenon of temporal interpretation. it seems more accurate to talk about the arising or not of the interpretation. the interval corresponding to the arrival in the USA.1. The occurrence of the lifetime effect is the opposite side of the coin of their absence.

the lifetime issue is reduced to the working of temporal interpretation in general. The leading idea in this regard will be that the TT is topic sensitive—that is. that referring to the lifetime span of the individual referred to by the DP subject.3). focusing on sentences with lifetime predicates. a lifetime reading arises (or not) depending on the content of the TT. In this sense. I argue that these two cases can be explained uniformly. For “out of the blue” examples. it gets its content from salient elements in the conversation. as I said before. For sentences inserted in a context with a previous past interval mentioned. the reading that arises is a lifetime reading. that the TT takes its content from the lifetime interval of the individual referred to by the DP subject.212 Individuals in Time (61) RT (UT) T′ 2 T AspP (after) 3 TT Asp′ (Harry’s existence interval) 2 Asp VP (within) 2 ET VP (being from California) Since. In sum. based on Musan (1995). The next sections discuss the sources for the TT. I show that the arising (or not) of this reading does not simply hinge on the presence of a “previously mentioned past-tense interval” (as was the case of Musan’s [1995] examples). accounts for them by appealing to different resources (Gricean Maxima of Quantity for “out of the blue” cases.4 The Determination of the TT Content and Lifetime Effects Thus far. I argue that the temporal TP 2 .4. Thus. I claimed. in this way differing from Musan. I will discuss the role of the contextual information conveyed by the DP subject itself in particular. I conceive the lifetime reading as an interpretive outcome derived from the location in the past of a TT with a specific content—namely. also in the line of Musan. in (61). we have seen two types of cases: “out of the blue” cases and sentences inserted in a context. it is the whole span of time of Harry’s existence that gets located in the past. that the TT takes its content from the previous interval mentioned (I will refine this account in section 6. 6. In what follows. who. Put in these terms. and Tense restriction for contextual ones). I proposed.

Kratzer (1977. the content of their TTs is different) because their TTs get their values from distinct “discourse topics”. following von Fintel (1994). lifetime effects do not arise. The common ground determines the options that are live at any point in the conversation. where. In the next two sections.e. 6.4. As defined by Stalnaker (1972. I assume that sentences are uttered against this background. I define “discourse topic” as the subject of immediate concern in the conversation. that it can be described as an anaphoric relationship. Based on Keenan-Ochs and Schieffelin 1976. the topic of the discourse has been proven to be of high relevance in the contribution of possible antecedents for anaphoric elements and for the freevariable restricting quantifiers (von Fintel 1994). I offer a preliminary description of the contextual background that DP subjects themselves can bring onto the sentence and the influence it plays with respect to the TT. 1979). 1981). I assume. a “context” contains a “common ground”—that is.3 I will show how my proposal (to account for all the cases uniformly) works in a systematic way. they are anaphoric to distinct “discourse topics. (62) Todos los chicos eran esquimales Every guy was Eskimo (63) La mayoría de los chicos eran esquimales Most of the guys were Eskimo (64) Varios chicos eran esquimales Several guys were Eskimo (65) Muy pocos chicos eran esquimales Very few guys were Eskimo (66) Muchos chicos eran esquimales Many guys were Eskimo (67) Algunos chicos eran esquimales Some guys were Eskimo .Tense and Individual-Level Predicates 213 interpretation of these two types (with and without context) varies (i. Grice (1975). Among the elements constituting the discursive common ground. I assume that every natural language discourse takes place in a context. and von Fintel (1994).” Following widespread proposals in semantics and pragmatics. Regarding the specific relationship between the discourse topic and the elements influenced by it. Consider the following sentences.1 When the Subject Is a QDP I am going to examine a set of examples without any previous explicit context. among many others. In section 6. I will show that the TT is influenced by the discourse topic. nevertheless..4. a set of propositions that form the shared background of the participants in the conversation.

and von Fintel (1994). However. all together. Longobardi 1994). Partee (1984a). say. but just over the contextually relevant set of plates. and these examples resides in the nature of the subject DPs.214 Individuals in Time (68) Ningún chico era esquimal No guy was Eskimo In none of these examples is the lifetime reading available. which the noun is a predicate of (69a). and this. In (70). In essence. That set is called the “restrictive clause. Then. determiners/quantifiers are interpreted with respect to the set contextually provided. constitutes the quantificational restriction. The context provides a set that the NP intersects with. as mentioned in chapter 5. Higginbotham 1983. D x [ x is N] g restrictive clause Dx [x(N)] Furthermore. among others. the meaning of (70) is that of (71). the set that determiners quantify over is restricted by the context (Westerståhl 1984. in these cases we cannot allude to the presence of a previously mentioned past interval to account for the mitigation of the interpretation. The contextual relevance gets into play through the contextual variable appearing as a subindex of the determiner. Lewis 1986). boys in (69). As is classically argued (Heim 1982. Since the difference between (52) (Gregory was from America). Barwise and Cooper (1981) pointed out that determiners do not quantify over the total domain of entities but just over the set specified by the noun. Thus. the simplest hypothesis is to suppose that the difference in the availability of the reading resides in the different nature of the DPs. where the reading can be defended to exist as a possibility. I am going to argue that the very nature and working of quantifiers is what precludes the arising of the reading. quantifiers bind the open argument (variable x). propose that such a restriction is formalized through a free variable in the determiner. Stump (1981). . in the world. QP/DP 2 Di NP the 2 xi N′ g boy b. every is not quantifying over the whole set of plates.” (69) The boys a. Stowell 1989.

Along similar lines as before. where the preferred interpretation is (ii). the availability of the lifetime interpretation does not arise. Let me explain what I mean with an example. the value of the TT is not a lifetime span. 10 . the lack of a (free) restricting variable can be argued to be the difference between generic and nongeneric DPs.Tense and Individual-Level Predicates 215 (70) Every plate (71) Everyc x [x(plate)] The point I want to make here is that it is because this universal property of quantifiers of being contextually restricted that a lifetime reading is absent in the sentences in (62)–(68). no lifetime effect arises. as a result. Dinosaurs were from the North Pole b. during which these examples arose. Compare these two situations. a lifetime reading does not arise. These examples show that the factor triggering the nonappearance of the reading can be other than the very presence of a previously mentioned past interval. and. too. I want to argue that this contextual information establishes the link with the background where the TT finds its content. a background is built up. I turn to more subtle situations.10 This hypothesis establishes the previous context as the context that counts for the sentence in the conversation. (i) a. that lifetime effects arise in sentences such as (i). As a consequence. The contextual restriction of the quantifiers directs us to a previous context where the free variable C finds its antecedent. as explained by von Fintel (1994) and mentioned before. and. and makes it the common ground where the TT finds its antecedent. even though we have a lifetime predicate in past tense. In essence. 6. Actually.4. This absence of a link to a previous context has an outcome in the temporal realm—namely. If the TT content is other than the lifetime span of the individual. since the reference to a previous scenario obtains.11 It is interesting to note that English bare plurals (ia) or generic Romance DPs (ib) do not seem to be contextually restricted.2 Context Associated to Individuals At this point. The point I would like to introduce here is that DPs referring to individuals also have contextual information associated with them. I would like to argue that via such an obligatory contextual restriction. it concurs with Musan’s (1995) observation that an element from previous context may make the reading not to appear. Los dinosaurios eran del Polo Norte (ii) Dinosaurs are extinct 11 I thank Dominique Sportiche and Daniel Büring for the discussions around these topics.

the use of a past tense sounds natural (He was from Portugal). “is a predicate. and.12 A situation where it can be seen that context information matters in the interpretation of a proper name (i. This morning Eva ran into my office and told me: —Eva: You remember João.” 13 I thank Olga Fernández Soriano for this discussion . really? So he is (#was) from Portugal! (that is why we did not notice so much difference between his accent and ours…) The first observation that can be made is that. We went to Brazil three years ago and got a tourist guide for our excursion through the jungle. we used to hang out with him very often. the basic idea is that (72) and (73) are different because the contextual information associated with the individuals referred to by the subject DPs is different. to know who we are referring to by that name) is the following. a name. the tourist guide we got in Brazil? (It turns out that) he grew up in Lisbon. We went to Brazil three years ago and got a tourist guide named João. This kind of information (“the Amàlia from work” vs. “the Amàlia from the For Raposo and Uriagereka (1995:191). since we got along with him. Amàlia. the other is a friend of theirs from the town where they usually spend the summer.13 Suppose the speaker and the hearer know two people with the same name. 12 . I propose that individuals also have contextual information associated and this becomes available through the contextual variable of the determiner. Following Higginbotham (1988) and Raposo and Uriagereka (1995). He happened to be moving to Lisbon.. Fisher. —Me: Oh! So he was from Portugal! (that is why we did not notice so much difference between his accent and ours…) (73) Situation 2: My friend Eva and I are Portuguese. In few words. and so what makes Fisher Fisher is (perceived or imagined) ‘Fisherhood’. the most likely and plausible interpretation will be that the girl they are talking about is “the Amàlia from work. whereas in the first situation. it does not seem so in the second situation. I suggest that even proper names may be subject to contextual information. I want to claim that the reason why this is so is the same reason as why a lifetime reading does not arise in (72). If the speaker and the hearer are once talking about an issue from work and someone called “Amàlia” arises in the conversation.216 Individuals in Time (72) Situation 1: My friend Eva and I are Portuguese.e. whereas one of the individuals named Amàlia is a coworker. This morning Eva came into my office and told me: —Eva: You know what? João grew up right here in Palmela! —Me: Oh. Suppose further that.” not the other one. for example.

Since the contextual sources contain different intervals (past in the first case. In the first case. at a party. In (72). In (73). that is. The explanation I would like to propose to account for it is that. following Musan’s (1995) suggestions could have sufficed to license a past interval as the content of the TT in the lifetime sentence. —Felipe: Oh. Whereas in (74) the background would be the trip we did three years ago. this is due to the contextual background associated to the DP subject. and. were you able to handle it with the language over there? —Me: Of course! Don’t you realize João is (#was) from Portugal?! In this scenario. I suggest that the TT gets its content from the context built up by the contextual information conveyed by the individual referred to by the DP. the contextual information associated with he (=João) is the situation where we had him as our tourist guide some years ago.Tense and Individual-Level Predicates 217 summer town” would be conveyed by the contextual variable C. if it is a proper name or if it is not. We are telling Felipe about a trip that João and I made some years ago to Brazil. So. present in the second) the TT has a different content. in (75) the background is a situation that includes the present. (74) João c= the tourist guide we got in Brazil three years ago (75) João c=a friend of ours in the present Each piece of contextual information builds up a different context for the conversation. and in (73) the past tense sounds awkward. to a previous context. In the second case (73). the contextual information associated with the DP subjects of (72) and (73) matters in the temporal interpretation. the contextual information is a situation such that he is a current friend of ours. there is a past tense mentioned (were able to). by contrast. However. which. the same as before. Felipe. the use of a past tense in the copular sentence (João was from Portugal) is not an option. (76) further supports this view. These examples show that something other than just a previously mentioned past interval (as suggested by Musan 1995 with sentences such as (53)) that can intervene in the determination of the antecedent for the TT content. which is a subindex of the determiner in the DP. but it extends to include the present moment. The contextual information that the DP Joao involves points to the utterance . the lifetime predicate does not have a lifetime reading in (72). the contextual info that may be associated to the individual referred to by the DP is not limited to a situation in the past. (76) Situation: my Portuguese friend João and I are talking to another guy. the scenario directs the hearer to a situation located in the past. As I proposed.

where the individual referred to by João is present. I propose that the mechanical interpretive procedure determining whether a lifetime effect arises is the same: the content of the TT is determined by the elements in the salient discourse. I want to show now. following von Fintel (1994). and. 5. 6. The occurrence of lifetime effects depends on the content of TT of the sentence. Elaborating on Musan 1995.3 Articulating the Account In the last two sections I have shown that the context that counts for the TT content can be that introduced by the DP subject itself. more systematically. as a consequence. lifetime effects do not arise when the interval relevant in the sentence is other than the lifetime span of the individual referred to by the DP subject. I also argued that proper name DPs involve a similar context variable. 3. 2. the context acting as background constitutes the set where anaphoric elements and free variables find their antecedents. how this is articulated. The relationship between the contextual background and the TT or the free variable C has been taken as anaphoric. I have argued that this is brought about through the context variable restricting quantifiers (free variable C). the utterance situation is the most salient context. which contains contextual information associated to the individual referred to by the DP.4. Beghelli & Stowell 1996).218 Individuals in Time situation. along the same lines that specific DPs are generally understood (Pesetsky 1987. The leading ideas introduced thus far are the following: 1. it becomes the source for antecedents. The context relevant to find the “antecedent” for the TT does not have to be another past interval necessarily. Conversely. so that all the surveyed cases can be accounted for in a uniform way. Enç 1991a. In other words. I claimed that lifetime effects arise when the TT refers to an interval that can be described as the lifetime span of the individual referred to by the DP subject. 4. (77) Salient context " TT . The TT refers to a specific interval. The relevant context can be brought to the scenario by the DP subject itself. This seems to suffice to convert the utterance situation as the background that counts as relevant for the antecedent of the TT. as Musan (1995) suggests. That is. A specific interval is an interval that is linked to a previous context.

three years ago. so that a past form becomes excluded. what definitely matters for the arising or not of lifetime effects is found in the contextual information of the DP subject. Since there is no past form. Then. This is because.Tense and Individual-Level Predicates 219 What varies in each case is the salient context and its properties. in the copular sentences I am analyzing. no lifetime effect arises. so he did not have to go through the immigration process. the DPs are surface subjects. (79) João c = a friend of ours in the present TT → looks at the context and finds a situation where João is relevant in the present. Result: a past form is allowed and. while others are simply explained by alluding to a previous TT. this creates an asymmetry in the explanation of (59) and (72) and (73). In the proposal I have sketched. Let us see how this works example by example: (78) João c = the tourist guide we got in Brazil three years ago TT → looks at the context and finds a (past) interval at which João was our tourist guide. they are contain the information relative to previous contextual information. Result: a past form is not allowed. no lifetime effect can arise. This explains why a lifetime reading does not arise or why a past tense is not allowed in certain cases: because the context involved contains a past interval such that the TT of the lifetime sentence does not refer to the lifetime span of the DP subject. the contextual information associated to the DP subject becomes of crucial importance in determining the context where the TT gets its antecedent. As topical elements. which makes them sentence topics. I would like to argue that this kind of procedure (looking at the discursive topic element and its properties) is also behind the interpretation of examples such as (59) above. since it refers to the time of the trip. It could be the case that some cases have to be accounted for by appealing to the contextual information of the DP subject. and because the context involved refers to an interval that includes the UT. which resembles Musan’s (1995) examples. These cases were explained by Musan by arguing that the lifetime reading was not an option because the past interval of the IL predicate referred to the interval at which Maria and Harry arrived in the USA. (59) Maria and Harry arrived in the USA. Harry was from California. In other words. However. the temporal interpretations available vary because the contexts the TT interval is linked to vary. where the information conveyed by the DP proves to be of .

then.4. Gregory himself TT → the whole interval overlapping Gregory’s lifetime span. a lifetime effect does not arise.220 Individuals in Time crucial importance in the licensing and interpretation of a past form. Bearing in mind that the TT is a ZP context sensitive. (81) Harry c= the guy who arrived in the USA with Maria TT → looks at the context and finds a (past) interval at which Harry and Maria arrived in the USA. these typically are out of the blue sentences.” they establish the relevant context. 1997). no lifetime effect arises. it is the information associated to the DP subject that counts. Once a context is fixed. by default.4 In determining the context relevant for the specification of the TT content (which is what finally decides the temporal interpretation of a sentence). Result: a past form is allowed in the copular clause and. Since the contextual variable directs us to a previous context containing a past interval. there is no reason to consider (59) different from (72) and (73). The interpretive procedure is always the same and works the following way. Result: lifetime arises With Musan (1995. Actually. (…) Topic of the sentence that may establish the connections to a previous background = Harry. I have argued that the establishment of the relevant context is brought about through the free variable DPs (more specifically. it finds its antecedents in the most salient context. let us consider a sentence where a lifetime reading can arise. I propose. that in (59) it is the contextual information associated to the DP Harry that establishes the context that counts. 1995) and Musan (1995) assume. the ZP .4 Summary of Section 6. (80) Harry was from California. (82) Gregory was from America (83) Topic of the sentence that may establish the connections to a previous background = Gregory (84) Gregory c= undefined. As Kratzer (1988. when there is no specific context the free variable is to be linked to. the value it takes by default is the individual himself and the interval associated to him. Finally. 6. the determiners) have. since they are the same in all relevant respects—all of them are copular sentences in context. Since the DP subjects of copular sentences can be considered “topical elements. since it refers to the time of the arrival. I assume that.

Result: lifetime arises (86) a. or salient. the possible contents of the TT are not only either a contextual interval or the interval corresponding to the lifetime span of the individual. but that the TT gets its content from the context set up by the DP subject. etc. They arise when the TT is the whole interval an individual’s existence extends over. whose ordering value is ‘within’). the TT of an utterance is the span of time corresponding to an individual’s existence. gypsy. a salient one. and it gets ordered (in the past) within the event time of a predicate. The length of the ET of lifetime predicates can be predicted from the length of the span of time an individual’s existence. because that interval strictly corresponds to the span of time the eventuality lasts. When such an interval is the interval that overlaps the predicate. This makes the interval corresponding to the lifetime span just one option among others. Gregory era rubio/cruel Gregory ser-PAST-IMPF-3SG blond/cruel b. or. Thus. I am not proposing that the TT “copies” the information from the DP subject. When the IL predicate refers to properties that do not (necessarily) hold for the whole lifetime span of an individual (blond. but any interval shorter than the lifetime span (when he was thirteen. the lifetime reading is obtained. (85) TT → the whole interval overlapping Gregory’s lifetime span. in these cases. the interval corresponding to an individual’s existence is particularly accessible. rather than to plainly syntactic ones. . In sum. When a lifetime predicate (Eskimo. in other words. when. (recall that the aspectual form involved in this temporal interpretation is the imperfect. lifetime effects arise depending on the content of the TT. By the same token. which is why the lifetime reading is not. for some reason or other. cruel). before becoming a pacifist. we get a lifetime effect. Gregory veraneaba en Cádiz/ leía el periódico después de desayunar Gregory spend-PAST-IMPF-3SG the summer in Cádiz/ read-PAST-IMPF-3SG the paper after breakfast “Gregory used to spend the summer in Cádiz/read the paper after breakfast” In the next section I work further on the TT property of being highly sensitive to discursive factors.Tense and Individual-Level Predicates 221 TT gets its content from it. from Africa) is at stake. it is the interval contained in the interval the predicate occupies (the ET).). even though the predicate is not a lifetime predicate.

the content of the RT is. the embedded TT would become ordered with respect to the main ET. As mentioned in chapter 5. 1996) work. and its eventuality will be located with respect to the eventuality of the main clause. My purpose in doing this is to explore the temporal interpretation of these predicates when dependence with respect to another tense complex is syntactically established. Following the logic of Stowell’s proposal discussed above. as in (i). have noted. in compound sentences.M.15 However. I focus now on the interpretation of lifetime IL predicates in another type of syntactic environment. starting by the former ones. by default. The time adverb at 3 can modify the TT.1) I described the way temporal interpretation is accounted for following Stowell’s (1993.1 Complement Clauses In chapter 5 (section 5. Peter’s leaving occurs prior to 3 P. the prototypical case of this is the modification of the past perfect. As Hornstein (1990) and Demirdache and Uribe-Etxebarría (2004). whose nature I argue to be also a ZP. According to him. Thus. this could lead us to a situation where the TTs are not ordered between each other strictly speaking. the UT. yielding the interpretation in (iii). Peter’s leaving occurs at 3 P. (Thus far. temporal interpretation is derived from the ordering relationship Tense establishes between the RT and the ET. namely. 6. the subordinate clause will be temporally dependent on the main one. Specifically. the ordering between the TTs should depend on the specification of the embedded RT. giving rise to different temporal readings. (i) (ii) (iii) 15 14 Peter had left at 3 P.5. In simple sentences.1. This way. as the subindexes of the ZPs in (i) suggest. In this work. instead of with respect to the TT. I have just dealt with IL sentences which establish a link with a previous one paratactically.M. if the RT of the subordinate clause is in the c-command domain of the ET. this does not mean that ETs always remain in the dark for interpretation.222 Individuals in Time 6.14 but rather the TT. I argued that what Tense locates is not the ET as such. subordinate clauses (complement and relative). yielding different interpretations. TT and ET can be modified by time adverbs. yielding the reading in (ii) or the ET.M. among others. given that the ET is the closest ZP.5 The Lifetime Reading in Compound Sentences For the sake of a more complete survey of cases. . so that it can be controlled by the ET of the main clause.) I will deal with complement clauses and relatives in turns. since the RT would be controlled by the main ET rather than by the main TT. However. to understand the temporal interpretation of a compound sentence is to understand the relationship holding between the two (or more) TTs. the subordinate RT can be affected by the ET of the main clause.

it seems that. in effect. John told me that Mary had said at 5 P. Consider (ii).Tense and Individual-Level Predicates 223 (i) TP 2 RTi T′ 2 T0 AspP 2 TT Asp′ 2 Asp0 VP 2 ETk VP 2 e CP 2 TP 2 RTk T′ 2 T0 AspP 2 TTm However. and at 3 P. Second. it could also be the case that the embedded RT could be controlled by either of those yielding different readings. unfortunately. we get the following: . only marking the event by an <e>.M. Let us first assume that the time adverbs at 5 and at 3 modify the TTs. This would yield the following ordering in (iii).M. independently from the syntactic position it occupies with respect to other TTs. Even in cases such as (i).-------tell------------UTT Assuming now that the time adverbs at 5 P. (iii) ------wash at 3 P. this fact raises the question whether an analysis in terms of closeness and control is adequate to fully capture the interpretation of the data. that Peter had washed the car at 3 P. temporal ordering is not affected by the choice of TT or ET as RT controllers. where the TT and the ET can be proved to be disjoint in reference. it is the TT the ZP counting for interpretation. a thorough investigation of the precise way how the RT in embedded clauses gets its value has to be left for future research. First. Actually. (i) (ii) Peter had left at 3 P. Since. there do not seem to be differences in the interpretation. as I will show through the study of IL predicates in relative clauses. Just as time adverbs can modify the ET or the TT (see footnote 14).M. Nevertheless.M. in what follows I assume that the time-denoting argument counting for temporal interpretation is the TT for two reasons. modify the ETs. it seems that the temporal ordering between the events is not directly affected by the choice of the ET or the TT as controllers of the embedded RT. for the time being I will just ignore the ET in the syntactic representation of the data.--------say at 5 P.M.M.M.

-----say-----5 P. (89) --------------washing-------------saying--------------UT Their ordering can be explained by the syntactic placement of both tense complexes. to the fact that this is the form that can potentially interfere in the interpretation of the past tense in the IL predicate. Since the subordinate eventuality is located in the past with respect to the main one. The RT of the subordinate complement clause is in the ccommanding domain of the main TT. which is the case we are interested in. For a more complete survey of cases in Spanish. Thus. In what follows I will restrict the main and the subordinate clause to the past tense.M.16 el coche (87) María dijo que Juan lavó Maria said that Juan wash-PERF-PRET-3SG the car enfermo (88) María dijo que Juan estaba Maria said that Juan estar-IMPF-PRET-3SG sick The situation compatible with the reading of (87) is the one described in (89).------tell--------------UTT As can be seen. the tenses of both clauses. and second. Consider as a first illustration the contrast between (87) and (88). I restrict the examples to those where the main verb is intensional (in particular. this reading is usually known as the past-shifted reading.224 Individuals in Time Regarding the temporal interpretation of complement clauses. 16 Also. I am not going to discuss the analyses themselves about the temporal interpretation of complement clauses. a saying verb). Both eventualities are located before the UT but they are ordered with respect to each other. . the nature of the subordinate predicate. as the subindex i indicates. it is controlled by it and gets its value. as the reader may have figured out already. two parameters have to be taken into account to analyze the relationship between the two TTs—first. the reader is referred to Carrasco 1998. the RT of the embedded clause. since that is the form giving rise to the lifetime interpretation under debate. where the event of washing is interpreted as preceding the event of saying. is ordered ‘after’ (because it is a past form) the embedded TT (with a subindex j).M. As will be shown shortly. (iv) -------wash-----3 P. and saying is interpreted before the UT. I will simply take the guidelines already given in the literature regarding eventive verbs and (SL) stative verbs and investigate what we get when the predicate is of lifetime nature. The decision to focus on the past tense also in the main clause is due. this type of verbs trigger ambiguities when both tenses come in past. whose content is the same as the internal ZP (the TT) from the main clause. the temporal ordering between the events is unaffected.

Enç 1987. the situation is more complicated. As widely noted in the literature (Ladusaw 1977. In Stowell’s (1993. respectively. such an ordering is derived from the control of the subordinate RT by the main clause TT. among many others). 1996) terms. Abusch 1988.Tense and Individual-Level Predicates 225 (90) RT TP 2 (UT) T0 T′ 2 AspP (after) 2 TTi Asp′ 1 Asp0 VP 1 e VP 2 María V′ 2 say CP 1 TP 2 RTi T′ 2 T0 AspP (after) 2 TTj Asp′ 2 VP Asp0 1 e VP 2 Juan V′ 5 wash the car The TT of the main clause (TTi) and the TT of the subordinate one (TTj) refer to the concrete intervals at which María was involved in saying and Juan was involved in washing the car. Ogihara 1996. sentences like (88) have two . When a stative predicate is at stake. These two TTs are further ordered between themselves in the way specified in (89). Stowell 1993.

though. In (88). Some authors assume a Sequence of Tense rule. share a similar underlying idea: that the tense in the subordinate clause is not a regular “past”. That is. Just when a past in the ET ZP is under a past in T. Simultaneous interpretation is then explained by arguing that the Tense head itself is null. Basically all. Tense can locate the eventuality in the past. as roughly represented in (92). One of them is a past-shifted reading (see the adverbial complements appearing in parentheses in (91)). Stowell (1993) claims that the past and present morphology is not a realization of the category Tense. it has to be c-commanded by a past in T. A rough representation of the idea is in (93). . the interval of John’s being sick overlaps with the interval of Maria’s saying. From a different perspective. which sort of deletes the content of the subordinate tense (Ogihara 1996 would be along these lines). a reading where the subordinate predicate is understood as temporally overlapping the main predicate. but it originates in the ET ZP. stative predicates have another reading: the so-called simultaneous reading. as it appears from the outside.17 17 For a fuller description of this account. sick sick ? ? (92) ----------------------//////X/////////-----------------UT---------↑ said The explanations to account for this phenomenon. However.226 Individuals in Time possible temporal readings. please see Stowell 1993. (91) María dijo (en el juicio) que Juan estaba enfermo durante el interrogatorio (el cual había tenido lugar tres semanas antes) Maria said (in the trial) that Juan was sick during the questioning (which had taken place three weeks before) This is the predicted reading by accounts like Stowell’s as described thus far. For a past form to mean ‘past’ (which would give rise to a past-shifted reading). while the (visible) morphological past is explained due to the presence of past in the ET ZP. where a past tense does not locate the subordinate eventuality ‘before’ the main one (traditionally called the Sequence of Tense phenomenon) slightly vary among themselves.

Tense and Individual-Level Predicates 227 (93) a. The first one is to assume that the subordinate T has the meaning of a present tense. however that happens. Past-shifted reading TP 2 RT T′ 2 T ZP (past) 2 Z VP (past) b. as it was in the first formulations of the Sequence of Tense phenomenon. representing (88). . Both alternatives are in (94). Simultaneous reading TP 2 RT T′ 2 T ZP (∅) 2 Z VP (past) I will consider now the two mentioned options and show the advantages and disadvantages of each. The second one is to take the intuition that T in the subordinate clause may be null.

As the subindexes gloss. what we do is to order such an interval. we can think that a concrete interval (TTj). corresponds to be sick. (TTi). controlled by the upper TTi. either present (corresponding to the ordering predicate ‘within’) or null. . Let me spell out the two options. María dijo que Juan estaba enfermo ‘Maria said that Juan was sick’ b. First.228 Individuals in Time (94) a. TP 3 RT T′ (UT) 2 T AspP (after) 2 TTi Asp′ 2 Asp VP 2 e VP 2 María V′ 2 say CP 2 TP 2 RTi T′ 2 T AspP (within/Ø) 2 TTj /TTi Asp′ 2 Asp VP (within) 2 e VP 2 Juan V′ 2 be sick Aside from the two options for the content of T. Then. with respect to the (subordinate) RT. there is something else different in the tree I have drawn: the content of the subordinate TT. the subordinate TT can be either an interval different from the main clause one. TTj. different from the TT of saying. or the same one.

the interval the subordinate clause is referring to is the interval at which Maria said something. in contrast. That is. then. Since there is no content in T. which seems to accurately capture the intuition about the interpretation of (88). with no further independent evidence. This way. the interpretation can be worded as follows: the TTi corresponding to the interval of “saying” is included in the interval of “being sick”. therefore. there is no ordering predicate. the RT binds the TT and. The second possibility represented in (94b) is to consider (a) that Tense does not have the meaning of ‘present’. (96) María dijo que Juan estuvo enfermo Maria said that Juan estar-PERF-PRET-3SG sick The unique temporal interpretation is a past shifted reading. this is not fully appealing since we are considering a past form (recall that what we read is was sick) as if its semantic import were that of a present (“within”). which seems desirable since simultaneity is only possible with such an aspectual form. In other words. a past shifted reading? The content of T. but it has no content. sick sick ↓ ↓ (95) ----------------------//////X/////////-----------------UT---------↑ saying However. by virtue of the content of the aspectual head. but ‘within’. and (b) that the subordinate TT is not an interval different from the main clause (TTj). where the main clause interval gets located “after” the subordinate interval. Consider. as a consequence. which is not ‘after’. but the same one of the main clause (TTi). then. crucial use of the content of the aspectual head. their temporal values coincide. then. the possibility that an interval becomes ordered with respect to itself does not actually arise. Is this a technical inconvenience? I would like to argue that it need not. since the content of T is null. If we follow interpreting the tree. the same sentence in perfective. (97) -----------be sick-------------said---------------UT----------I will thus assume that an analysis along the lines sketched in (94) (T as null and the subordinate TT being the same as the main clause’s one) is appropriate .Tense and Individual-Level Predicates 229 as usual. it is null. what we get is that the interval (TTi) is included (‘within’) the eventuality of being sick. This analysis makes. What precludes.

the simultaneous one. (99) *Cristina dijo que Eva había sido de Albania18 Cristina said that Eva had ser-PAST part from Albania (100) Cristina dijo que Eva había estado enferma Cristina said that Eva had estar-PAST part sick For analyzing the simultaneous interpretation. with a stative SL predicate. namely. since it emphasizes the relevance of the aspectual head content. . specifically. I will focus my attention on sentences containing a lifetime predicate. ‘Eva serfrom Albania in a previous life’. 18 PERF-PRET-3SG At most. The past shifted reading is absent. Consider in contrast (100). this could be a paraphrase for Eva fue de Albania en una vida anterior. (98) Cristina dijo que Eva era de Albania Cristina said that Eva ser-IMPF-PRET-3SG from Albania The first characteristic to be observed is that there seems to be just one temporal reading.230 Individuals in Time to account for the simultaneous reading. will examine whether a lifetime reading is available. as can be tested by the unavailability of paraphrases such as (99). I will follow what I said above. Once I have sketched the mechanics of temporal interpretation in complement clauses with stative subordinate verbs. I will examine the temporal interpretation of (98) and.

Given that a lifetime reading arises when . according to what I said above (a lifetime reading arises when the TT is the whole span of time over which an individual’s lifetime extends over). because the content of the subordinate T is assumed to be null. simply. That is. which is “included” in the span of time over which Eva is from Albania. Now. it cannot shift any TT into the past. in (98) it does not arise. The other reason is.Tense and Individual-Level Predicates 231 (101) TP 3 RT T′ (UT) 2 T AspP (after) 2 TTi Asp′ 2 Asp0 VP (after) 2 e VP 2 Cristina V′ 2 say CP 2 TP 2 RTi T′ 2 T AspP (Ø) 2 TTi Asp′ 2 Asp (within) VP 2 e VP 6 Eva be from Albania In words: the UT is ‘after’ the interval at which Cristina was involved in saying. First. The reasons are two. a lifetime reading is not expected in (98) since the TT of the lifetime predicate refers to the time at which Cristina was involved in saying. does this state of affairs trigger a lifetime reading? According to natives’ intuitions.

a lifetime interpretation is not expected from a syntactic situation such as (101). when the DP is indefinite and acts as a complement of an intensional verb. Secondly. Summarizing very much.5. or Juan was looking for a particular girl. I will make some considerations about the interpretation of lifetime predicates in another syntactic environment: Relative Clauses (RCs). it can have two interpretations. I will divide the task in two steps. The reading roughly described as [–specific] would correspond to a narrow scope position with respect to the intensional verb. we see that the interpretation of the DP a girl. as I have been doing thus far. Firstly.232 Individuals in Time the TT is located in the past. 1996) work. (104) Juan besó a una niña Juan kissed a girl Stowell (1993) argues that the tense interpretation of a relative clause can be predicted from its scopal position. (103) a. I will pose what this analysis predicts with respect to the lifetime predicates and show whether it is borne out or not. can be paraphrased in two different ways: either Juan was looking for any girl. Rivero 1975. the DP can be considered as unambiguously [+specific]. which may be determined by that of the DP . Stowell (1993. which can be described as a [–specific] and a [+specific] (Russell 1905. Donnellan 1966. whoever she might be. among others). 1996) builds on Ladusaw (1977) and Abusch (1988) to elaborate his proposal about the temporal interpretation of RCs.2 Relative Clauses In this section. namely. a girl > look for When the main verb is extensional. and the [+specific] interpretation to a wide scope position. look for > a girl b. If we take an example like (102). Abusch demonstrated that the different temporal readings of RCs correlate with the interpretation of the DP the relative depends on. 6. (102) Juan estaba buscando (a) una niña John was looking for a girl The two interpretations of such DPs have been traditionally accounted for in scopal terms. Adriana. following Stowell’s (1993. I will summarize general points concerning the temporal interpretation of RCs.

as well as the RC TT follows the main TT. which guarantees the interpretation of the DP as specific. A clause is considered temporally independent when its TT is ordered with respect to the UT. the tense complex of its RC would become dependent on the main clause one. but. Juan kissed a girl who danced at the party earlier These two temporal readings available are represented schematically below: (107) a. the subjunctive mood correlates with a [–specific] one. that the two past tenses are locating their respective TTs just with respect to the UT. This amounts to saying that (105) can be true if the RC TT precedes the main TT. As Enç (1987) and Stowell (1993) point out. the adjunct clause would be temporally independent. This way. The possible temporal orderings between them are in (106). The idea that both of the TTs are “independent” with respect to each other is represented in (108).Tense and Individual-Level Predicates 233 containing it. Whereas the indicative mood correlates with a [+specific] interpretation for the DP. (107b) equivalent to a past-shifted reading of the 19 As Rivero (1975) noted. (106) a. For further discussion about these issues. in Spanish. see Brugger and D’Angelo 1994 and Quer 1998. rather. if the DP is scoped out to a higher position from where its RT cannot be controlled by the matrix ET (or the TT in the terms of this work). it is not the case that a past form can locate an ET either ‘before’ or ‘after’. -----------dance----------kiss------------------------UT As (107) depicts. instead of with respect to the TT of the upper clause. if the DP remains in a narrow scope position. . Sentences under discussion have indicative in the relative. Juan kissed a girl who danced at the party later b. there are two possible and apparently antithetical temporal orderings. 2001. the specific versus nonspecific interpretation of the DP has a correspondence in the mood that the RC headed by the indefinite DP adopts. Take as an example the interpretation of the RC below:19 (105) Juan besó a una niña que bailó en la fiesta Juan kissed a girl who danced at the party The main and the subordinate clause both carry a past tense form. (108) ------------kiss---------------------------UT ------------dance-------------------------UT Since the matrix and the RC ETs are not ordered with respect to each other. ----------kiss--------------dance---------------------UT b. any ordering of their ETs. However.

following Stowell’s suggestions. can truthfully be captured by (105). However. the indefinite DP a guy is interpreted as [+specific] and can be supposed to take wide scope (becoming an adjunct of TP). . we can suppose that the RC moves along the DP. since the content of the subordinate TT is not in the domain of the higher one. consider now the following example with a lifetime predicate in the RC. and (107a). this case does not look different from (59) above and repeated below. Furthermore. where a simultaneous reading would be an option (making the lifetime reading unavailable as a result). This would seem to pave the way for a lifetime reading to arise. Bearing all this in mind. (109) Selene besó a un chico que era de California Selene kissed a guy who was from California As a complement of an extensional verb (kiss). In fact. equivalent to a forward-shifted reading. where the TT of the IL predicate refers to the arrival time.234 Individuals in Time relative. the lifetime reading does not arise either. (110) CP 2 DPi TP 5 2 RC RT T′ (UT) 2 TP 2 T AspP (after) 2 RT T′ (UT) 2 TT Asp′ T AspP 2 (after) 2 Asp0 VP TT Asp′ 5 2 kiss ti 0 VP Asp 2 e VP 5 from California The subordinate TT is not in a position where its content can be influenced by the upper TT.

As surveyed in chapter 5. I suggest that. Stowell (1993) proposed that the content of the RT (PRO) was defined by the “closest” c-commanding ZP.6 Summary of the Chapter In this chapter I dealt with the temporal interpretation of IL predicates and made two points. independently from their c-commanding relation. I argued that equating IL-hood with permanency of the property is inadequate.2. This leads us to the conclusion that the content of the TT is not determined in the same terms proposed for the RT. 6. considered as a “temporal PRO” affected by control. In section 6. Opositive. I analyzed the temporal interpretation of IL predicates in several contexts. but to conditions related to discourse topicality. by virtue of its properties as “specific. However. but by the discursively most prominent ZP. I then argued that the TT is an interval (arguably. etc. This makes the kissing time the most likely antecedent for the TT in the RC.” reaches a topical position by being scoped out. a ZP in nature) whose content is not determined by syntactic distance factors. TT is an interval sensitive to discourse prominence. therefore.Tense and Individual-Level Predicates 235 (59 ) María y Harry llegaron a EEUU. and (b) those denoting properties that are not necessarily permanent. as Stowell (1993) proposes for the RT. I divided IL predicates into two groups: (a) those denoting permanent or lifetime properties. The analysis of lifetime predicates in RCs suggests that the TT is not determined by the closest ZP. so he did not have to go through immigration As in (59). I argued against the idea that lifetime effects arise with any kind of IL predicate and under any circumstance.1. that one where the guy was kissed. differing from Kratzer (1995). Thus. data like the RCs discussed here show that the content of the subordinate TT can refer to the main TT. Harry era de California. the TT of the lifetime predicate in the RC refers to the context containing the TT of the kissing time. In support of these two points. In section 6. coming from the antecedent DP a guy. the lifetime reading does not arise. First. as native intuitions confirm. the context acting as relevant background is the one the DP is linked to. así que no tuvo que pasar la aduana Maria and Harry arrived in the USA. The TT is not sensitive to syntactic parameters (such as closeness) but to the salience an interval has in discourse. namely. the relevant background (where the subordinate TT finds its content) is built up through the contextual properties present in the relative pronoun. Harry ser-IMPF-PRET-3SG from California. Since. the content of the TT is not primarily subject to syntactic conditions. which. the TT is not the interval over which the lifetime of the individual extends over. in the RC of (109). That is.) coincide with the initial and final temporal bounds of the . The temporal bounds of lifetime properties (Eskimo.

) can hold of an individual just for a determined period of time shorter than the individual’s lifetime. if the TT refers to the interval corresponding to the span of time an individual’s lifetime extends over. Consider (113) in relation to (111). I showed that a lifetime reading is not available in all of the cases where a lifetime predicate appears in the past tense.4. when it refers to the whole interval of an individual’s lifetime. 1997) insight that lifetime effects can be neutralized by the presence of another past interval in the context. Actually. Among the issues I have not discussed in this work is the relationship between the TT content and time adverbials. in section 6. I assimilated her observation to the theoretical terms assumed here. In sections 6. although it has become clear that the TT of an IL lifetime predicate can refer to an interval given by the discourse.236 Individuals in Time individual’s lifetime that is their subject. Others (kind.3 and 6. he was from California . are not necessarily lifetime properties. After introducing Musan’s (1995. who defends the position that the lifetime characteristic comes specified in the lexical entry of the predicate. therefore. I argued that the contextual restriction of DPs can intervene in building up a previous background. (111) Harry and Maria arrived in the USA. I also offered a number of examples of lifetime predicates in the past tense that. Thus. blond. I differ from Musan (1995. 1997). etc. which prevents a lifetime reading from arising. their good combination with restrictive temporal adjuncts proves. as. I have shown that a good number of predicates that can be considered IL. I defended the lifetime reading as an interpretive outcome due to a specific content of the TT—namely. that interval cannot be “spelled out” by an adverbial. Harry was from California. for example. and I argued that the arising of the lifetime reading depends on the content of the TT. In this respect. I argued that being a predicate that overlaps with the lifetime span of individuals (or not) is an interpretive outcome from the lexical properties of such a predicate. I concluded. For example. do not have a lifetime interpretation. that the content of the TT is not sensitive to syntactic distance but to discourse prominence. based on independent grounds. In both cases the lifetime reading does not arise. I examined the arising of the lifetime reading in syntactic composition—specifically.5. so he did not have to go through immigration (112) TT refers to the time of Harry’s (and Maria’s) arrival (113) *When Harry arrived in the USA. Finally. complement and relative clauses. even without the overt presence of another past tense. although the syntactic dependence between the matrix and the subordinate tense complexes was different. a lifetime reading will be available. A large number of questions remain to which I have not even suggested a possible answer. most of the properties are like the latter ones.

Whereas the eventuality “keeps the same. which remains for future work. when a TT is maintained “anaphorically” (as is the case in (111)). at least. Likewise.” how can a topical element work as contrastive at the same time? On my view. That is. two reasons. this contrast is not possible with predicates of the kind of be from California. adverbials are excluded in sentences such as (113) because of. the possible variability of the TT content. the fact that the eventuality keeps the same does not explain. The assertion is not explicitly limited down to that interval in contrast to some other interval. Second. . If the TT is supposed to be a “topical element. According to Klein. it contrasts the TT to some other possible TT. First. where the eventuality keeps the same no matter which TT we choose.” as Klein claims. That is. two things can be pointed out. as we have been able to observe throughout this chapter.Tense and Individual-Level Predicates 237 Klein (1994:164) suggests that when an initial adverbial narrows down the TT to a certain time span does so “in contrast” to some other time span for which the same assertion could be made. the temporal interpretation of the sentences varies according to the TT content. one may wonder whether a TT can count as “contrastive” at all. in and of itself. Among other questions. this is not entirely clear in Klein’s explanation and deserves more investigation. no such contrast can be involved.

.

When ser is involved (1). and. In particular. First. correspondingly. or funny person. and tense. In the following pages. contrary to widespread belief. any instance of estar yields an SL one. The length of the interval an IL property extends over can be restricted in time. dark-skinned. 7. (1) guapo/ moreno/ gracioso Pablo es Pablo ser-PRES-3SG handsome/ dark-skinned/ funny “Pablo is handsome/dark-skinned/funny” guapo/ moreno/ gracioso Pablo está Pablo estar-PRES-3SG handsome/ dark-skinned/ funny “Pablo looks handsome/is tanned/is being funny” (2) Based on these minimal pairs. by studying the temporal properties of IL predicates in copular sentences at three levels—inner aspect. I summarize the main claims I have made and sketch some of the consequences they may have for the discussion of the IL/SL dichotomy in general. got tanned. linked to external reasons (maybe because he is wearing a nice suit. In the cases with estar (2). I examined whether this distinction is rooted in temporal terms. I have consistently taken it that any instance of ser yields an IL predication. the speaker predicates the properties of the subject on a particular occasion. in this book I sought to explore what lies behind the dichotomy traditionally known as individual-level (IL)/stage-level (SL). most of them denoting properties related to the origin (ethnic or genetic) of the . ser and estar. the property is predicated of the individual as such: the speaker claims that the subject is a handsome. outer aspect.1 Summary of the Conclusions The detailed study of temporal properties of ser/estar copular clauses has shown the inaccuracy of many of the traditional tenets concerning the IL/SL contrast and enables us to make the following four conclusions. Only a subset of IL predicates are permanent predicates. or is in a good mood). IL predicates are not necessarily permanent properties.Chapter 7 Conclusions and Final Remarks As set out in the presentation. Spanish distinguishes between two copulas. whose semantic characterization I have argued corresponds to the IL/SL distinction.

mean) have been proposed to correlate with the presence of the relational complement of these adjectives (cruel to Mary). Rather. which clearly pattern with activities in an inner-aspect classification.240 Individuals in Time individual. in Spanish. such as Eskimo. the aspectual characteristics of the copular constructions are derived from their syntax. the dynamic properties of clauses where adjectives referring to mental properties appear (such as cruel. the temporal interpretation of IL predicates is derived. this study is a contribution to a better understanding of the role of prepositions in the properties of event structure. as is the case with any other type of predicate in Spanish (e. among others). however. Thus. The restrictions between a ser-predicate and aspectual forms like the perfective or progressive are derived from the inneraspect properties of the predicate. in the case of IL predicates. differing from previous influential proposals like Kratzer 1988 where. Topic Time has been argued to be a time interval taking its value from the salient discursive context. I have concluded that those proposals arguing that ser-predicates should be considered SL under aspectual forms like the perfective and the progressive cannot be correct. kind. Ramchand 2003. dynamicity is proposed to correlate with the presence of the relational complement and argued to be rooted in the Prepositional Phrase. no stative verb can appear in the progressive). Finally. Demirdache & Uribe-Etxebarría 2000)..g. IL predicates are not all stative. Specifically. Second. I have shown that ser-predicates can appear in any aspectual form. or color-blind. are properties that can be true just of a limited period of time. gypsy. Crucially also. such as blond or young. contrary to general belief. Ritter & Rosen 2000. there are two types of IL predicates regarding inner aspect: stative and dynamic predicates. ser and estar. this book gives further empirical support to recent hypotheses according to which aspectual properties are syntactically encoded (Borer 2005. As a result. but it is independent from that: cruel mean/kind to someone can appear with both copulas. based on hypotheses that attribute aspectual content to prepositions (Hale 1984. The investigation of Lifetime Effects (the interpretive phenomenon whereby the DP subject can be understood as ‘no longer alive’) in chapter 6 deepens our understanding of the concept of Topic Time. I have accounted for a variety of contextual situations illustrating the sources from . In particular. I have proposed that the dynamic and active properties classically observed in certain copular constructions are due to the presence of a complement. Others. from the content of the so-called Topic Time (Klein 1994). I have argued that the dynamic nature of copular constructions does not correlate with an SL classification. as is the case with any other type of predicate. The book proposes a uniform treatment of tense. unlike what is commonly assumed. In this respect. tense takes and orders the nominal argument (the NP subject). Third.

most authors dealing with the ser/estar contrast concur with the idea that estar-predicates put the property in relation to a situation. 7. As mentioned in chapter 2. more in general. As was surveyed. and the semantic domain they belong to. Previous claims regarding IL predicates Throughout this work I have analyzed IL predicates in their relation with the temporal realms cited. One of the crucial questions is how the association to a situation is brought about. we will be proposing something about the terms in which the IL/SL dichotomy is represented in the grammar. I will introduce some discussion in the following sections. most .1. tense orders the individual referred to by the DP subject Perfective and progressive viewpoints make an IL predicate SL IL predicates are stative predicates Notion involved Length of the interval the property holds Temporal ordering Semantic domain Tense Tense Delimitation and progress in time Internal temporal characteristic of the property Outer aspect Inner aspect Table 7. Although I will not be able to offer a definitive answer to this. and I have given an account for the different temporal interpretations that IL sentences can have without yielding inaccurate generalizations. the notion involved.1 summarizes the claims previously made.Conclusions and Final Remarks 241 which the Topic Time can take its value. Ser simply classifies the subject into the category denoted by the predicate that appears in combination with it. As noted in the beginning of the book. In examining this question we address the key difference between IL/SL predicates. By answering this question. Table 7. these notions. One of the important questions here is whether the difference between ser/estar clauses (or. These conclusions differ from most previous literature. are temporal concepts in nature.2 The IL/SL Dichotomy and the Ser/Estar Contrast With major or minor differences. while ser is more “innocuous. between IL/SL predicates) can be cast in temporal terms. where the definition of IL predicates has been mainly associated to the notions of permanent property and stativity. a number of authors have proposed that the IL/SL distinction should be cast in “aspectual” terms. Claim (in previous literature) IL predicates express permanent properties When an IL predicate is at stake.” leaving the property without any association to a determined situation. classically alluded to in the description of IL predicates.

In this work. I understand that the difference consists of something describable along the lines of the rightmost column. outer aspect.).” or “temporally anchored” (in contrast to IL predicates. nonstable predicates. if we say that SL and IL predicates differ in certain temporal terms. Table 7. among others). 7. tense). Temporal unit descriptions and definitions In what follows I summarize the conclusions I have drawn from the study of each temporal realm and discuss whether the IL/SL dichotomy can be ultimately established in temporal terms. and gave concrete definitions to each. SL predicates are conceived as episodic. Chierchia 1995. if we . tense) and have described the working of each of these heads as well as the differentiations we can establish among predicates depending on the distinctions made by each temporal head. I have argued that to attribute permanency to a property amounts to assigning it a determined temporal interval that the property extends over. their definitions. Different number of occasions eventualities occur and different orderings between the Eventuality Time (ET) and the Topic Time (TT) intervals. which lack all such characteristics). I have considered that each of them is syntactically represented (quantity. 1995. To understand a property as permanent means either that the property holds for the entire lifetime span of the individual (intelligent) or that. Temporal unit Tense Description Predicate that orders the Topic Time (TT) with respect to a Reference Time (RT) Complex projection containing a quantifier giving the number of occasions an eventuality occurs and a predicate ordering the Eventuality Time (ET) and the Topic Time (TT) It refers to the algebraic mereological properties of predicates Distinctions Depending on the content of tense (present/past/future) and the content of the Topic Time (TT).242 Individuals in Time of them describe SL predicates as “involving aspectual properties. Differentiation between homogeneous and heterogeneous predicates Outer aspect Inner aspect Table 7. In contrast.2. aspect. One of the difficulties in these descriptions is the vagueness as to what exactly the authors mean by such terms.3 The IL/SL Dichotomy Is Not a Permanent/Episodic Distinction IL predicates have been defined in the literature as permanent predicates (Kratzer 1988. Thus. I distinguished three temporal levels (inner aspect.” “temporally bound. However.” it lasts for the remainder of the individual’s lifetime (Ph.2 summarizes the temporal units. once “acquired. and the distinctions that can be made according to each one.D.

but then he changed his attitude (4) (5) In conclusion. In (4). (3) Juan dejó de ser rubio/guapo/dulce/generoso/altruista/ egoísta/atrevido/miedica/valiente/criticón/retorcido/sensible/ soberbio/envidioso/pesado/servicial cuando se hizo mayor Juan stopped ser-ing blond/handsome/sweet/generous/altruist/egotistical/daring/fearful/brave/faultfinding/twisted/sensitive/arrogant/ envious/tedious/helpful when he grew up Cuando era pequeño. since the interval IL predicates hold over can be restricted. . Juan era rubio/muy guapo/muy dulce/ generoso/altruista/egoísta/atrevido/miedica/valiente/criticón/retorcid o/sensible/soberbio/envidioso/pesado/servicial when ser-IMPF-PRET-3SG little Juan ser-IMPF-PRET-3SG blond/very handsome/very sweet/generous/altruistic/egoistical/daring/fearful/ brave/faultfinding/twisted/sensitive/arrogant/envious/tedious/helpful Pedro estaba siendo muy amable con Juan {durante la entrevista/mientras le entrevistaba/hasta que consiguió lo que buscaba}. In (3). all the following cases (discussed in chapters 2 and 6) become unexplained. 1 For a recent proposal on the IL/SL distinction in temporal terms. pero luego cambió de actitud Pedro was ser-ing very kind to John {during the interview/while he was interviewing him/until he got what he was looking for}. in direct relation to this. and in (5) the copular ser-clause appears with another number of adverbial complements restricting the interval the property holds. the IL/SL dichotomy is not rooted in temporal terms. which can be tested by the nonvariation of the copular verb in Spanish. which is unexpected if IL predicates are forced to apply for the entire lifetime span of an individual. the copular clause appears as a complement of dejar de ‘stop’. at least in temporal terms regarding the length of the interval.1 Another important respect in which the proposal defended here differs from earlier accounts (particularly Kratzer 1988.Conclusions and Final Remarks 243 endorse permanency of a property as a definitional characteristic of IL predicates. Since the length of the span of time properties hold does not define IL predicates. an adverbial clause restricts the period the property holds. in correlation with the wa/ga marking of subjects in Japanese. the distinction between IL/SL cannot be described in terms of permanent versus episodic. 1995) concerns the structure of IL and SL predicates and. In other words. the argument that Tense takes. see Torii 2000. The interval a certain property lasts seems independent from its nature as SL or IL. permanency cannot be a defining characteristic of IL predicates.

Since. In Kratzer’s proposal. whereas the external argument of an SL predicate is a spatiotemporal variable. which is different in (6) and (7). within) with respect to the TT interval. the (spatio)temporal variable is bound by the quantifier over occasions (Q<occ>) discussed in chapter 5. Crucially. the differentiation between IL/SL predicates cannot be set on the basis of the presence or absence of the spatiotemporal variable. namely the TT. as I have argued. whereby it takes an interval. if an appropriate context is built up). I followed Stowell (1993. Tense takes and locates (with respect to a reference interval) the spatiotemporal variable in the case of SL predicates. in contrast. |>1|. does not affect the choice of copula. One advantage of acknowledging a similar argument structure for both IL and SL predicates is that Tense is argued to work uniformly. Harry estaba Harry and Maria arrived in the USA. temporal interpretation depends on the content of the TT and its relation with respect to the reference time. which is not accurate. this view has enabled me to capture the temporal interpretations potentially available with IL predicates without yielding any overgeneralization. In Kratzer’s account. before. The TT can vary without triggering any variation on the copula choice. In contrast. as shown in chapter 6. In these examples. and orders it with respect to another interval taken as a reference. Harry estar-PAST-IMPF-3SG enfermo y pidió ayuda para recoger la maleta sick and asked for help to get the suitcase “Harry and Maria arrived in the USA. As suggested there.244 Individuals in Time As cited in chapters 2 and 6. This has the limitation of predicting Lifetime Effects for all appearances of IL predicates. and their external argument is the DP subject itself. which is supposed to keep stable. In my account. Q<occ> gives the number of times an eventuality occurs and can be present with any kind of predicate (including those such as Eskimo or gypsy. Instead. I have argued for a uniform treatment of Tense. whereas it takes the DP subject referring to the individual in the case of IL ones. the TT in the copular clause is exactly the same (8). (6) Harry y María llegaron a los EEUU. The TT is the interval the speaker refers to and for which a particular predicate is asserted to hold. Temporal interpretation does not hinge on argument structure but depends on the content of the TT. IL predicates lack a spatiotemporal variable. Harry was sick and asked for help to get the suitcase” . This. temporal interpretation was derived from argument structure. however. ∃). which is discourse sensitive. The number of occasions obtained. thus. Kratzer argues that the difference between IL and SL predicates resides in their argument structure. is ordered by the Aspect predicate (after. depending on the quantifier (|1|. 1996) in arguing for the presence of this variable in all types of predicates.

Although these authors are not entirely explicit about the temporal realm they are referring to (or about the definition of some terms). but on the verb plus its complement(s). On my view. and Fernández Leborans (1999). 7. which is why he did not have to go through immigration” TT = the time of the arrival (8) This constitutes another piece of evidence that the IL/SL dichotomy is not rooted in temporal terms of the realm of Tense.Conclusions and Final Remarks 245 (7) de Harry y María llegaron a los EEUU. Whereas estar possesses internal temporal structure. Schmitt (1992). In this sense. when they establish the IL/SL distinction. it is not a state. Schmitt (1992) argues that estar involves aspectual properties. That is. Luján (1981) establishes a parallelism between estar and predicates like write a letter. Likewise. In contrast. an event or a process. in this respect. it is not accurate to say that ser is not a state. the predicate may pattern with either states or activities. ser lacks any inherent temporality and is described as “inert with respect to aspect. the notions they allude to. Harry was from California. in the works by Luján (1981). Tense has been defined as a predicate ordering the TT with respect to a reference time. both undelimited predicates. the inner-aspect type of one predicate does not depend only on the verb alone. I defined inner aspect in a . In the first place. a delimited process. an event. whereas ser is underspecified with respect to aspect. but the copular verb plus its complement can be asserted to behave either as a state or as an activity. The copular verb itself may be none of those. Specifically. Harry era Harry and Maria arrived in the USA Harry ser-PAST-IMPF-3SG from California. or a process. and Fernández Leborans (1999) what the term “aspectual” refers to or how inner aspectual properties are tested is not made explicit. for Fernández Leborans (1999) the IL/SL dichotomy is founded on aspectual properties. then. Second. As mentioned in chapter 3. I have largely shown in chapter 4 that ser-clauses differ with respect to their inner-aspect properties. I argued that depending on the adjective. In turn. it makes no difference between an SL and an IL predicate. seem to belong to the domain of inner aspect.” I would like to make two remarks regarding these proposals. ser has no inherent temporal structure. the copular verb ser would not be very different from regular predicates. Schmitt (1992).4 The IL/SL Distinction Is Not a Matter of Inner Aspect…Completely In chapter 2 I surveyed the accounts by Luján (1981). por lo que no tuvo que pasar la aduana California for it that not have to go through immigration “Harry and Maria arrived in the USA. ser is conceived as a predicate parallel to write or admire. In a similar vein.

I will briefly discuss four points. Once we have established a concrete definition of inner aspect. First. I showed that estar (SL) and ser (IL) predicates are both homogeneous predicates. which indicates their impossibility to construe heterogeneous predications. that is. subject to empirical verification. Nevertheless. we can see that the differentiation between SL and IL predicates cannot be put in such terms so neatly. relying on the subinterval and additivity tests (chapter 3). A subinterval of “ser de California” is “ser de California” b. In this respect.). “Estar enfermo” + “estar enfermo” = “estar enfermo” (12) a. A subinterval of “estar enfermo” is “estar enfermo” b. etc. (9) Pedro estaba Pedro estar-PAST-IMPF-3SG “Pedro was sick” enfermo sick de California (10) Pedro era Pedro ser-PAST-IMPF-3SG from California “Pedro was from California” As (11) and (12) show. it is true that estar predicates can constitute heterogeneous predications. It seems clear that Inner aspect does not give us the key to establishing the difference between the minimal pairs I am alluding to (ser/estar guapo ‘be/look handsome’. which we can test by the acceptability of adverbials such as in + x time. I have described inner aspect properties as mereological properties. According to their mereological properties. predicates can be distinguished between heterogeneous and homogeneous.246 Individuals in Time concrete way. ser-clauses do not tolerate in-adverbials under any circumstance. “Ser de California” + “ser de California” = “ser de California” These examples suggest that the IL/SL dichotomy cannot be described in inner aspect terms. I have made inner-aspect distinctions in terms of the algebraic notion of part. as (9) and (10) exemplify here. (11) a. this conclusion deserves some remarks. both types of predicates give same results in the proofs testing mereological properties (subinterval and additivity tests). as I will show in short. However. .

only if the adverbial in +x time is present. the verb they are related to has to be of heterogeneous nature too. the predicate gains a telic interpretation: morena *(en dos semanas) (16) Con esta crema. Bosque proposes that they are not adjectives derived from participles.. but rather adjectives yielding verbs. you get tanned in two weeks” Finally. Regarding cut-short adjectives.3 2 For a detailed discussion about participles.Conclusions and Final Remarks 247 guapo en una hora (13) *Juan es Juan ser-PRES-3SG handsome in an hour “Juan is handsome in an hour” Second. ‘get sick’) which. and cut-short adjectives. The projection of AspQMAX gets justified by the properties of the participles themselves (or the adjectives derived or related to them). rather than llenar > llenado > lleno (‘fill > filled > full’). which come from heterogeneous verbs. do not give rise to telic constructions with the copular verb: *Pedro estuvo enfermo en dos horas ‘Pedro estar-PRET-PERF-3SG sick in two hours’. There are other adjectives (enfermo ‘sick’) related to a verb (enfermar. AspQMAX is projected. behaving. ‘sick-INF’. as canonical telic predicates: terminado en una hora (14) El artículo estaba the paper estar-PRET-IMPF-3SG finished in an hour “The paper was finished in an hour” llena en tres horas (15) La piscina estaba the pool estar-PRET-IMPF-3SG full in three hours “The pool was full in three hours” Third. 3 For participial and cut-short adjectives to yield a heterogeneous construction. therefore. estás with this cream estar-PRES-3SG dark-skinned in two weeks “With this cream. estar-clauses admit in-adverbials with certain predicates (participles and perfective adjectives that have been described as “cut-short”2). due to their atelic nature. That is: lleno > llenar > llenado (‘full > fill > filled’). see Bosque 1990. I would like to suggest the idea that in the cases of participles (14) and cut-short adjectives (15). with predicates other than participles and cut-short adjectives (i. from which participles derive. . the ones of our minimal pairs in (1) and (2)). the meaning of the adjectives (of the minimal pairs in (1) and (2)) with the in +x time adverbial is different from the meaning that ensues from the contrast in such minimal pairs. participial adjectives. Following the theoretical framework I have worked with here.e.

the [+Q]/[–Q] estar contrast. etc. funny. heterogeneous/homogeneous. where the SC is an AP (sentences such as (16)). dark-skinned. that is. the property (morena) does not hold at every interval. I will label such estar-constructions ‘[+Q] constructions. where the unique difference is that the property is or is not linked to a particular circumstance. In both cases. this seems possible only if the adverbial in + x time is made explicit. 5 In section 7.’ As mentioned before. 4 I will not deal with the question whether the subject of the SC moves to the specifier of AspQ to check quantity features. in conclusion. (13)5).). in (16). you get dark-skinned (‘tanned’) in two weeks’. One possible way to capture this dependence is proposed in (18). as I pointed out. too. Borer 2005). ser cannot appear with the adverbials in +x time (cf. does not capture the contrast of the minimal pairs ser/estar handsome. darkskinned. we are dealing with a homogeneous predicate. In contrast.248 Individuals in Time (17) estar 2 estar AspQ 2 SC (= Participle Phrase) 2 el trabajo terminado la piscina llena In the cases of “regular” adjectives (handsome. the property holds of Pablo uniformly at every subinterval at which the property (and the circumstance) holds.7 I return to the impossibility of the combination of participles with ser. . However. Nevertheless. only estar can constitute heterogeneous predicates.4 (18) estar 2 estar AspQ 2 en dos semanas SC (= AP) 2 tú morena The meaning of examples such as (16) can be paraphrased as follows: ‘(by using that cream). the presence of AspQMAX can be defended. The truthful predication of the property depends (with estar) or not (with ser) on a specific circumstance. but only when the process is completed. as it is the case of objects in transitive sentences (cf. If we say Pablo está muy guapo ‘Pablo looks very handsome’.

it cannot be captured in temporal terms. Fernández Leborans 1999). we have to explain independently why the copular verb does not change. More specifically. or the ordering between the TT and the ET (outer aspect). 7. if we accept the idea that outer-aspect forms cause a switch from IL to SL.Conclusions and Final Remarks 249 In sum.6 Recasting the IL/SL Distinction The distinction we want to capture is that represented by the minimal pairs (1) and (2). the property is understood as linked to a distinct situation. or the ordering between the TT and the RT (tense) does not make any difference between ser and estar clauses. muy guapa en su juventud (19) María fue Maria ser-PAST-PERF-3SG very pretty in her youth The reason commonly adduced is that.5 Outer Aspect Does Not Affect the IL/SL Distinction Some authors (e. So. Perfective aspect is argued to be one of those. as I concluded earlier. we accept that copular verbs have more than one value. which leaves the admitted copular differentiation unexplained. In view of the minimal pairs (1) and (2).. have argued that certain (outer) aspectual forms induce an IL predicate to “work” as an SL one. I argued that the distinctions that can be established according to the properties of the different temporal domains do not make any clear-cut difference between ser and estar clauses. ser can be either IL or SL. The native intuitions about the interpretation of those examples suggest that when the copula is estar. I have shown that when we apply temporal distinctions in a strict way. for the IL/SL contrast. we are speaking of an individual. intuitively appealing. The opposition between homogeneity and heterogeneity (inner aspect). the property results understood as ‘delimited in time’. If we endorse the idea that outer-aspect forms cause a “switch” from IL to SL.) are. inner aspect is not responsible for the contrast registered in the minimal pairs (1) and (2)—that is. where delimitation in time does not yield any change regarding the copula choice. and the association to a particular circumstance is absent. at first. etc. temporal anchoring. Although the descriptions of this dichotomy in temporal terms (aspect involving. with the perfective.g. they lack the projection of AspQMAX. who agree with the description of the alternation between ser and estar as an IL/SL dichotomy. Likewise. it seems natural to argue that the IL/SL differentiation does not purely reside in the adjectives themselves. Both types of constructions are [–Q]. However. (3)– (5). 7. we are talking about a concrete situation. since . I have shown several examples. sight. no revealing difference ensues regarding the ser/estar distinction. when the copula is ser.

the semantics of SL-hood would consist. I think that the properties of SL-hood consist of the ability to link a property to an external situation. The next natural question is what such properties are. the external situational variable (s) exists in addition to the eventive variable 6 I leave aside here those adjectives that combine only with estar. whereas the semantics of IL-hood would consist in the absence of such an association. Therefore.250 Individuals in Time they can go with either of the copulas. the copular verb that appears in the paraphrases of the adjectives inside the DPs is ser and not estar. copular verb estar. Crucially. Following Demonte (1999). I want to consider that adjectives are IL by default. Along the lines of Higginbotham and Ramchand (1996). and restrict my attention to those that can appear with either of the copular verbs. According to Higginbotham and Ramchand. I argue. Categorical judgments are predications about an individual (x). I consider that copular verb ser leaves the character of the predicate unchanged and simply ascribes the subject to the category denoted by the predicate. As I showed. Higginbotham and Ramchand (1996) conceive the IL/SL contrast as a distinction pertaining to the Logical Form. and the other is due to the intervention of one of the copular verbs. outer aspect. describable in terms of categorical versus thetic judgment (Kuroda 1972). It is estar that introduces the properties typically associated to SL-hood. whereas thetic ones are predications about a situation (s). both options are quite close to each other). the interpretive status of those adjectives without the presence of any copula is the same as when they appear with ser. . there are no differences between ser and estar predicates regarding inner aspect.6 Demonte shows that adjectives have an IL interpretation when used as modifiers inside a DP. Consider the following examples with the adjectives mentioned above: (20) Los niños guapos/morenos/graciosos ganaron el concurso the handsome/dark-skinned/funny won the contest “The handsome/dark-skinned/funny guys won the contest” (21) las niños que eran/*estaban guapos/ morenos/graciosos the guys who ser/estar-PRET-IMPF-3PL handsome/dark-skinned/funny “the guys who were handsome/dark-skinned/funny” As shown in (21). From this perspective. there are two possibilities: to propose that adjectives are not IL or SL. This way. makes the predicate SL. or to propose that adjectives are of one kind by default. but it is the entire predicate together with the copula that is IL or SL. or tense. That is. temporal heads are not capable of contributing this external situation. (In a sense. however. As introduced in chapter 2. in the association to a particular situation. among many others. precisely.

intelligent). 8 Obviously. I believe that a description of the IL/SL (ser/estar) dichotomy along these lines captures the Spanish native intuitions.. Whereas ser does not have any impact on the interpretation of the adjective—it leaves it as it appears as a modifier inside a DP (20)—estar does have an impact and associates the property to a concrete situation. the perspective suggested here shares the intuition underlying Schmitt’s (1992) and Fernández Leborans’s (1999) hypotheses that ser is. the predicate cannot lexically refer to an external situation in particular. instead of ascribing myself to the view that the difference belongs to an abstract predicational level. These authors argue that estar relativizes any property to a personal evaluation or perception. With ser predicates.Conclusions and Final Remarks 251 (which. whereas. in the case of Spanish copular alternation. The point of (22) is that the predicate includes as its meaning the necessary reference to an external situation. there is. since it is the presence of the eventive variable that. coercion is a process that eliminates the meaning conflicts between two elements of the same construction. This view is congenial to other recent proposals such as Escandell-Vidal and Leonetti (2002).7 The external situational variable (s) is the ‘subject’ that the ‘event property’ is predicated of. the ser/estar contrast is ultimately rooted in the lexicon and the interpretive differences between clauses containing ser or estar are due to such distinct lexical properties in the copulas. even when it combines with adjectives typically understood as IL (e. there is no predication of an external situation. As a consequence. for them. the association to an external situation. Fernald (1999). by the properties in its lexical entry. more “vacuous” than estar. to be defined in context. distinguishes between IL and SL predicates. Escandell-Vidal and Leonetti conceive this fact as a pragmatic reinterpretation process syntactically induced by the copula estar. . the adjective is interpreted in relation to a particular perception. it is the copula estar itself that conveys the linking to an external situation lexically. 7 Note that they differ from Kratzer (1995) in this respect. I would like to argue that. Coercion produces a conceptual adjustment that restores the acceptability of the utterance. among others. somehow. as a coercion process. I propose that it is the very copular verb estar that provides. is present in all types of predicates). and Escandell-Vidal and Leonetti (2002). for her. by definition. According to this hypothesis. with estar predicates. Thus. However. I propose that the lexical entry of estar includes the following content: (22) Estar: predicate that refers to a circumstance in which an individual is8 In a sense. That is. According to Pustejovsky (1995).g. in technical terms. every combination of an adjective with estar will be an SL predicate.

these cases differ from (1) and (2). Juan is taking the Americans’ side” muy demócratico (24) Juan está Juan estar-PRES-3SG very democratic “Juan is behaving very democratically” As the glosses show. Nevertheless. This explains their interpretation as IL when they act as modifiers inside a DP. then. since.hood has been proposed to derive from the properties of a lexical head. In principle. muy americano (últimamente) (23) Juan está Juan estar-PRES-3SG very American (lately) “Juan is behaving in the way Americans usually do” “In his opinions. they will be interpreted that way.. as soon as we add a degree quantifier such as muy ‘very’. strictly compositional. As suggested previously. whereas SL. inner aspect). .g. inside them. The cases I have in mind are those like (23) or (24). in a similar vein as I have conceived other grammatical properties in this work (e. Juan usually takes the Americans’ side” 9 This view of SL-hood is. there is no anomaly (regarding the adjectives’ meaning) to be restored in the cases of the minimal pairs (1) and (2). The main difference between inner-aspect properties and SL ones. I do not consider it necessary to talk about a pragmatic reinterpretation process when the verb estar enters into the picture. muy americano (25) Juan es Juan ser-PRES-3SG very American “Juan usually behaves in the way Americans do” “In his opinions. is that the former are built up out of a functional projection (Quantity). In this respect. I would reserve coercion to describe those cases where the adjective gets a different interpretation in order to make sense of its combination with the copula estar. coercion converts the SC containing the adjective from IL to SL in order to meet the copula requirements.9 However.” in the sense that in the absence of anything else in the structure.252 Individuals in Time Escandell-Vidal and Leonetti (2002) consider that estar selects for a small clause (SC) with certain aspectual properties. We have similar consequences with the copula ser. the sense of the adjective has undergone a slight change. as I have sketched here. I consider that adjectival predicates are conceived to be IL “by default. That is to say. When these properties are not met. note that this is not totally attributable to estar. there is nothing producing the linking to a circumstance. coercion occurs and converts an IL SC (lacking the necessary aspectual properties) into an SL SC. Strictly speaking.

quite. it becomes a qualifying adjective. Observe below that it is only in the presence of the degree quantifier that the adjective can be understood as ‘usually behaves in the way Americans do’ or ‘he is proAmerican’. if the adverb disappears. Once we are dealing with a qualifying adjective (however this is brought about). whereas ser + APs without very or the equivalent does not yield any of the interpretations in (28). Still. etc. therefore. or that adjectives are neither qualifying nor nonqualifying but it is the combination in syntax with very that makes them qualifying (again. whereby the adjective has undergone a reinterpretation process to make sense of its combination with very. The difference in meaning can be. the status of the adjective can be considered on a par to those in (1) and (2). along the lines I have developed inner aspect in this work). its rendering as SL with estar is explained the same as before: because of the lexical properties of the copula estar.) or in a comparative (26).Conclusions and Final Remarks 253 más americano que Pedro (26) Juan es Juan ser-PRES-3SG more American than Pedro “Juan usually behaves in the way Americans do more than Pedro” “In his opinions. Juan usually takes the Americans’ side more than Pedro” When a nonqualifying (classifying) adjective appears with a degree quantifier (very. . A possible shortcoming of attributing the change in meaning to the degree quantifier is that. Juan usually takes the Americans’ side” We can say either that a coercion process has taken place in the AP. we have a previous step where a nonqualifying adjective “has become” a qualifying one. there exists a contrast in acceptability between (29) and (30). the qualitative interpretation and the related readings are gone: americano (27) Juan es Juan ser-PRES-3SG American “Juan is American” (28) #“Juan usually behaves in the way Americans do” #“In his opinions. At that point. with estar the sentence without a degree quantifier can get them. That is. attributed not to its combination with estar but to the process the adjective undergoes by its combination with muy ‘very’.

10 Likewise. basically every adjective can appear in combination with estar. contento. 11 Recall from footnote 2 that Bosque (1990) argues that it is these adjectives that produce verbs and such verbs yield participles.254 Individuals in Time (29) ??Juan está americano Juan estar-PRES-3SG American muy americano (30) Juan está Juan estar-PRES-3SG very American 7. There are some adjectives that combine only with estar and reject ser.7 Some Remaining Questions The perspective I sketched in the previous section accounts for those adjectival predicates that can appear with either of the two copulas (ser and estar). according to them. contentado. which in Spanish is formed by ser + past participle. according to what I said in section 7.12 (31) Estar/*ser descalzo. which selects for [+perfect] SCs. where a [+perfect] feature is difficult to defend. Being conceivable as an object of perception is. falso ‘false’. cortado. Escandell-Vidal and Leonetti (2002) argue that the reason is because these predicates take a propositional argument as their subject. would otherwise become un- Bosque (1999) points out that not all adjectives can combine with estar—among them. (The [past] participle ending in Spanish is -ado/-ido. When these participles are part of a passive form. harto13 full fed up estar/ser barefoot glad (32) Estar/*ser destrozado. necesario ‘necessary’. 12 This remark is restricted to (adjectival) copular constructions. this is not the case. a necessary previous step for estar to induce coercion. hartado. lleno. in contrast to other authors who consider that adjectives of the type cited in (31) derive from participles. Nevertheless. More specifically. llenado. any adjective is predicted to be able to combine with it. conmovido. which is not an object of perception. As Demonte (1999) points out. the mentioned participles are allowed in combination with ser. It is the copular choice that makes the predicate either IL or SL. despedazado cut torn estar/ser destroyed moved The rejection of ser has been explained by arguing that the aspectual value (perfect) of the adjectives is only compatible with estar. given the innocuous characteristics attributed to ser.6. I leave aside here the study of the differences between the passive forms and the ser-copular constructions. That estar (obligatorily) selects for [+perfect] SCs cannot be the case since its combination with adjectives such as handsome or pale. most of the adjectival predicates that only combine with estar are cut-short adjectives (31)11 and adjectival participles (32).) 10 . and evidente ‘evident’. 13 The corresponding participles are: descalzado.

cansado Juan está/*es Juan estar/ser. As many authors have pointed out. Nominal predicates do not combine with estar (40). the impossibility of participial adjectives to combine with ser is complex and seems subject to exceptions. leaves unexplained other copular combinations. this does not affect the discussion about the combination with the copulas.14 muy seco (33) Pedro es Pedro ser-PRES-3SG very dry muy madura (34) María es María ser-PRES-3SG very mature muy abierta (35) María es María ser-PRES-3SG very open muy limpio (36) Su periodo presidencial fue his presidential period ser-PST-PERF-3SG very clean muy frío (37) El océano Atlántico es the Atlantic Ocean ser-PRES-3SG very cold muy caliente (38) El mar Muerto es the Dead Sea ser-PRES-3SG very hot Although. depending on their combination with ser or estar. However. the metaphorical reading proves that the hearer tries to give an interpretation in consonance with the structure. I will not investigate this issue here. in principle. 15 There are other types of adjectives derived from participles that.PRES-3SG tired cansado b. which only combine with estar (39). respectively. as described in chapter 2. note that there are adjectives of the type of (31) that can appear with ser. Rather.) (i) a. combinable with ser. El trabajo en la mina es/*está the work in the mine ser/estar-PRES-3SG tiring aburrido (ii) Juan es Juan ser-PRES-3SG boring aburrido (iii) Juan está Juan estar-PRES-3SG bored 14 . as Fernández Leborans (1995) notes.Conclusions and Final Remarks 255 accounted for. the necessary combination of (perfect) participles with estar looks consistent. either. in cases such as (33)–(36). (Note that they correspond to different participial forms in languages such as English.15 The idea suggested previously that any predicate is. Sometimes this alternation appears in correlation with the (in-)animate properties of the subject (i). the idea that estar contributes the meaning of linking to a circumstance with any type of predicate is not accurate. Also. Likewise. the interpretation of the adjective with ser gets. have an active or stative reading. such as copula + locative PP. a metaphorical reading. then.

the cruel-type. it seems that both notions can be conceived apart from each other. Thus. The definition of ser given explains why nouns combine with ser (40) (since nouns denote classes). but does not account for why they can appear just with ser and cannot be conceived as linked to a particular situation. are retained in their combination with estar: cruel con Pedro a propósito (43) Juan estuvo Juan estar-PST-PERF-3SG cruel to Pedro on purpose . I proposed that the peculiar properties to the constructions these APs participate in emanate from the particular properties the APs themselves involve. although episodicity can be one of the consequences of the linking to a particular situation. the description of estar as ‘episodic’. the obligatory predication with estar is explained. In chapter 4. I would like to add a few words about the combination of estar and the type of APs I have devoted special attention to in this work. such as agency.256 Individuals in Time en Brasil (39) Juan está/*es Juan estar/ser-PRES-3SG in Brazil ‘Juan is in Brazil’ profesor (40) Pedro es/*está Pedro ser/estar-PRES-3SG teacher ‘Pedro is a teacher’ We can make sense of (39) by alluding to the description of ser given. but something external to it. the linking to a particular situation is independent from the restriction of a predicate to a determined temporal interval: (42) Pedro fue camarero dos meses Pedro was a-waiter two months Finally. Since a location is not a class. which would leave (41) without explanation since in Spain is not a temporary property of Madrid. Ser classifies the subject by categorizing it into the class denoted by the predicate following the copula. As we already know. those peculiar properties. unlike adjectives. the ungrammaticality of Juan es en Brasil is explained. Consistently with this hypothesis. This perspective also allows us to avoid. en España (41) Madrid está Madrid estar-PRES-3SG in Spain Since a location is not a property of any individual. among other things.

since this does not happen with other adjectives: guapa (49) María fue Maria ser-PERF-PRET-3SG beautiful “Maria was a beautiful woman (in a period of her life)” guapa (50) María estuvo Maria estar-PERF-PRET-3SG beautiful “Maria was beautiful (as soon as she made herself up)’ . With cruel-type APs. There are other issues regarding these APs that deserve to be explored in deeper detail in future research. One of them directly refers to the copula alternation in Spanish. Whereas there is a neat contrast between (45) and (46). the interpretive difference in (47) and (48) decreases significantly. which makes the meanings of ser and estar converge. where the contrast between the two copulas looks mitigated. This is left unanswered here. waiting to be studied with the progressive form itself. the fact that the progressive is incompatible with estar is not accounted for. contrary to ser. as I mentioned in chapter 5. (44) *Juan estaba estando cruel con Pedro Juan was estar-ing cruel to Pedro If the dynamic properties are kept no matter what copula is at stake.Conclusions and Final Remarks 257 One important difference between the construction of these APs with ser or estar is the impossibility of estar to appear in the progressive. poses some issues still unexplained. The closeness in interpretation cannot be attributed to an effect of the perfective. since it appears that estar behaves in this respect as if it were a stative predicate. the adjectival property is interpreted restricted in time and tied to a particular occasion. amable (45) María era Maria ser-IMPF-PRET-3SG kind amable (46) María estaba Maria estar-IMPF-PRET-3SG kind amable (47) María fue Maria ser-PERF-PRET-3SG kind amable (48) María estuvo Maria estar-PERF-PRET-3SG kind In (47) and (48). which. the difference between ser and estar seems slightly neutralized.

regarding the minimal pairs where the IL/SL opposition is detected.258 Individuals in Time 7. I have shown that. Therefore. (1) and (2). I have shown that both estar and ser clauses can appear with the same outer aspect forms. The copula estar. 16 . outer aspect (by alluding to the completion or perfection of the property). or inner aspect (mereological properties).8 Summary I have proposed a definition of the ser (IL)/estar (SL) dichotomy that differs from the traditional one based on the contrast permanent/episodic. both predicates ser and estar are homogeneous predicates. Second. whereas estar is the only copula capable to constitute heterogeneous predicates with certain adjectives (participles and cut-short adjectives). by virtue of its lexical characteristics. links the property to a situation. then. in the first place. would locate the individual subject at the circumstance expressed by the predicate following the copula. Finally. One difference that exists between ser and estar regarding inner-aspect properties is that estar cannot appear in the progressive under any circumstance. that the length of the interval a property holds does not differentiate estar from ser clauses.3 summarizes all of these points. so that no distinction can be made between them in this sense either. inner-aspect properties do not seem to be able to differentiate between the two copulas. either. Restricting myself to the contrasts argued to correspond to the IL/SL dichotomy of (1) and (2). regarding inner aspect. I have contributed the idea that the association to a particular situation cannot be recast in terms of tense (the length of the interval). I have argued that the lexical entry of estar includes as a part of its meaning ‘circumstance at which an individual is found’. I have considered ser to be a copula that classifies an individual into a category (independently from how long the individual belongs to such a category) and estar to be a copula that. I have argued so by showing. since we can construct both types of sentences with adverbials that restrict the duration of the property.16 Table 7.

hinging upon the type of the predicate following the copula and the telic inducing adverbial in + x time. stative and dynamic Can appear in any outer aspect form The property is referred to the interval specified in the Topic Time Estar Situates individuals in a determined circumstance • Homogeneous constructions (stative and dynamic) and heterogeneous ones. Differences between ser and estar . • The heterogeneous cases of estar have a meaning different than the one registered in the minimal pairs ser/estar.Conclusions and Final Remarks 259 Ser Classifies individuals Homogeneous constructions. Can appear in any outer aspect form The property is referred to the interval specified in the Topic Time Table 7.3.

.

Allen. In Análisis del discurso oral: Anejos de Oralia. M. A. H. M. Bache. Los Angeles. The Development of the Copula in Child English: The Lightness of Be. and Scope. Bañón. M. Bello. Cambridge. University of California. dissertation. Madrid: Arco Libros. A default. MIT. 1996. 1988. Indiana University Linguistics Club. and T. 84–113. University of California. Anagnostopoulou and M.D. M. Toward the logic of tense and aspect in English. F. Chicago: University of Chicago Press. A. Linguistics and Philosophy 4:159–219. . M. Communications of the ACM 26(11):832–843. Barwise. P. Linguistics and Philosophy 15:463–508. Denmark: Odense University Press. The Phrase Structure of Quantifier Scope.). dissertation. 2000. The Order of Premodifying Adjectives in Present-day English. Stowell.D. 1978. 1986.L. and R. E. Bennis.References Abney. 1972. Asher. Szabolcsi (ed. Alexiadou. Becker. Cortés. truth conditional semantics for the progressive. N. Intensionality. Unergative adjectives and psych verbs. D. Sequence of Tense. The algebra of events. Cooper (1981). Ms. Los Angeles. 1983.). Incorporation: A Theory of Grammatical Function Changing. Oxford: Oxford University Press. Gramática de la Lengua Castellana Destinada al Uso de los Americanos. 1991. Abusch. and J. Baker. dissertation. S. J. Dordrecht: Kluwer. Partee. Espejo. J.D. A. E. Everaert (eds. Linguistics and Philosophy 9:5–16. To appear. 2004. F. Proceedings of WCCFL 7:1–14. A. 71–107. J. Foundations of Language 9:481– 491. Santiago de Chile.. Buenos Aires: Espasa Calpe. F. Muñío (eds. Ph. Bennet. Bach. La interpretación temporal de la enumeración de eventos en el discurso. 1995. Distributivity and negation: The syntax of each and every. In The Unaccusativity Puzzle. C. The English Noun Phrase in Its Sentential Aspect. Generalized quantifiers and natural language. 1847. cited by the edition of 1945. The Ghost of Times Past. and B. J. 1973. Beghelli. 1987. Ph. In Ways of Scope Taking. Arche. MA.). Beghelli. Ph. Maintaining knowledge about temporal intervals. Odense. 1988. L. M. Anderson.

Ph. Borer. and H. 1977. Bosque and V. In Quaderni del Laboratorio di Linguistica 1. J. Picallo. 1996. I. Stanford.). Demonte (eds. Almeida and J. S. and M. Bertinetto. Florence. I. Athens. Dorta (eds. Bosque. Bosque. 217–311. 1999. Preposición tras preposición. Tempo. 1998. Amherst. Farrell and D. Spain: Montesinos. University Complutense. Oxford: Oxford University Press. Aspect and quantification. and V. I. R. CA: CSLI Publications. In Tiempo y Aspecto en Español. J. Madrid: Cátedra. 2000. 1999. Paradisi (eds. L.. Amherst.262 Individuals in Time Benua. Occasional Papers in Linguistics 17. I.). Bertinetto. In Gramática Descriptiva de la Lengua Española.). Brugger. Chicago: University of Chicago Press. Zaenen (eds. Madrid: Espasa Calpe. Brentari (eds. Cioni. H. M. P. Paper presented at GLOW. La correlación de tiempos en español. Movement at LF triggered by mood and tense. Berardo. El sintagma adjetival: Modificadores y complementos del adjetivo: Adjetivo y participio. Bosque (ed. dissertation. 1986. Pisa: Scuola Normale Superiore. P. Journal of Linguistics 32:349–385. 1996. M. 1981. Madrid. 1993. 1998. I. 1994. The projection of arguments. dissertation. Agonini. On a frequent misunderstanding in the temporalaspectual domain: The perfective = “telic” confusion. Gramática Descriptiva de la Lengua Española. Ph. 1994. D’Angelo. G. I.). Academic Press. C. University of Venice Working Papers. 1996. H.D. Deriving passive without theta roles. Carlson. Bosque. New York. M.). Sobre el aspecto en los adjetivos y en los participios. Borer. 2005. Bosque. In Contribuciones al estudio de la lingüistica hispánica: Homenaje a Ramón Trujillo. M. In Tense and Aspect [Syntax and Semantics 14]. Aspetto e Azione nel Verbo Italiano: Il Sistema dell'Indicativo. Reference to Kinds in English.).D. The Evolution of the Grammar: Tense. dissertation. L. . University of Kansas. Madrid: Espasa Calpe. Italy: Accademia Della Crusca. P. Tedeschi and A. Borer. Tenerife.). Borer. and E. 177–214. The passive/anti-passive alternation. H. 1994. Bybee. In Morphology and Its Relations to Phonology and Syntax. A. I. 9–42. Ph. M. Demonte (eds. 60–99. Bosque.D. April 1996. Revista Argentina de Lingüística 9:9–48. 1999. MA: GLSA Publications. Perkins and W. G and M. Aspect. Postnominal adjectives in Spanish. P. Carrasco. Structuring Sense. Sobre las diferencias entre los adjetivos relacionales y los calificativos. Carlson. 31–64. 1990. University of Massachusetts. Lapointe. I. Pagliuca. Animacy and Shawnee verbal inflection. and Modality in the Languages of the World. Bosque. L. 114–145.

Lectures on Government and Binding. Oxford: Oxford University Press. Chomsky. The Logical Form of action sentences. H. D. MA: MITWPL. Knowledge of Language: Its Nature. University of British Columbia. Derivation by phase. Predication times in ST'at'Imcets (Lillooet Salish). Illinois State University. Cambridge. G. Cambridge. Chierchia. Chomsky. Cinque. Bertinetto.). Squartini (eds. Linguistics and Philosophy 15:11–183. MA: MIT Press. The Linguistic Forum 38:73–88. 2001a. 176–223. Origin.). In Ken Hale: A Life in Language. Occasional Papers in Linguistics 20. Italy: Rosenberg and Sellier. N. Higginbotham and M. 1999. D. Towards a unified theory of tense and aspect. In The Generic Book. and P. C. 1981. (ed. H. PA: University of Pittsburgh Press. Chomsky. 1997. 1992. Davis. and M. Delfitto. Comrie. Aspect: An Introduction to the Study of Verbal Aspect and Related Problems.). MA: MIT Press. Chomsky. J. 1967. Pittsburgh. Cambridge. Some Concepts and Consequences of the Theory of Government and Binding: Cambridge. 1979. M. Minimalist inquiries: The framework. . G. In prep. Pragmatics [Syntax and Semantics 9]. In Temporal Reference. Carlson and F. Martin. Bianchi. Lake Arrowhead. N. 1999. Bertinetto. N. G. Uribe-Etxebarría. Kenstowicz (ed.). Cambridge. 1–52. and Actionality: Semantic and Syntactic Perspectives. MA: MIT Press. Corpus of Spanish. Paper presented at the Conference of Tense and Aspect. Available at http://www. G. Michaels and J. D. In Step by Step: Essays on Minimalist Syntax in Honor of Howard Lasnik. P.corpusdelespanol. N.). 1997. Pelletier (eds. MA: MIT Press. 1976. New York: Academic Press.). N. Anaphora and dynamic binding. In The Logic of Decision and Action. 1995. H. 1997. Chomsky. Cambridge: Cambridge University Press. Beyond explanatory adequacy. and Use. A Teaching Grammar of St’át’imcets. 1986. Cambridge. N. Aspect. Demirdache. Cole. Adverbs and Functional Heads: A Cross-Linguistic Perspective. CA.org. Uriagereka (eds. N. New York: Praeger. Davies. Chomsky. Collins. Turin. Rescher (ed. 1982. Individual-level predicates as inherent generics.References 263 Chierchia. N. Dordrecht: Foris. Demirdache. A case study in the interaction of aspect and actionality: The imperfect in Italian. M. P. Chicago: University of Chicago Press. The Minimalist Program. 125–142. Davidson. M. B. 89–155. 1995. 1995. 2001b. Local Economy. Chomsky. 81–95. R. V. MA: MIT Press. 2000. M.

R. D. Linguistics and Philosophy 18:1–19. Revista Española de Lingüística 9:133–171. Madrid: Espasa Calpe. 1979. Madrid: Espasa Calpe. Structure-Preserving. 1992. In Step by Step: Essays on Minimalist Syntax in Honor of Howard Lasnik. In The Syntax of Time. Uribe-Etxebarría. 143–179. Donnellan. 1991. Ms. Cambridge. 1979. V. H and M.). Demonte. H. D. Bosque and V. Demirdache. Demonte (eds.264 Individuals in Time Demirdache. I. 1966. 1995. The primitives of temporal relations. On the absence of the present tense morpheme in English. 2461–2524. Clases de adjetivos: La posición del adjetivo en el SN. Egg. Lecarme (eds. I. and J. M. Dowty. Demonte. The semantics of specificity.). H. The effects of aspectual class on the temporal structure of discourse: Semantics or pragmatics? Linguistics and Philosophy 9:37–61. La predicación: Los complementos predicativos. Towards a theory of the diversity of temporal systems. and M. In Gramática Descriptiva de la Lengua Española. and M. Word Meaning and Montague Grammar. Demonte. I. Sintaxis y semántica de las construcciones con ser y estar. Enç. Bosque and V. Ms. . and Local Transformation. D. Reference and definite descriptions. Linguistic Inquiry 6:579–588. 1976. Martin. M.). 1986. 1991b. A Transformational Approach to English Syntax: Root. 2004. Depraetere. Dowty. Diesing. Uribe-Etxebarría. 1987. M. and Masullo. 1977.. 1991a. Linguistic Inquiry 18:633–657. Demirdache. 1999. 1999. 129–215. J. Thematic proto-roles and argument selection. M. Madison. MA: MIT Press. The ingressive as a new category of verbal Aktionsart. 157–186. 1995. Dordrecht: Reidel. The stative in the progressive and other essence/accident contrasts. Where have all the adjectives gone? Studies in Language 1:19–80. M. Uribe-Etxebarría. In Gramática Descriptiva de la Lengua Española. Dixon. Emonds. 1975. Indefinites. Anchoring conditions for tense. M. D. Enç. 1999. Linguistic Inquiry 22:1–25. J. Uriagereka (eds.. The syntax of time adverbs. Philosophical Review 75:281–304. University of Wisconsin. V.). University of Nantes and University of the Basque Country/Basque Center for Language Research. Dowty. Guéron and J. M. Dowty. Language 67:547–619. J. 2000. D. New York: Academic Press. Cambridge. MA: MIT Press. Demonte (eds. Michaels. Enç. K. V. Journal of Semantics 12:311–356. MA: MIT Press. R. Cambridge. On the necessity of distinguishing between (un)boundedness and (a)telicity.

Stanford. In Features and Projections. Greenberg. In Gramática Descriptiva de la Lengua Española. Tense-aspect-modality: The creole prototype and beyond. A. 1999. Ramchand. The quantization puzzle. C. 1999.D. H. Y. J. S. 1981. Evidential coercion: Using individual-level predicates in stage-level environments. 3129–3208. Prepositions and results in Italian and English: An analysis from event decomposition. Verkuyl. 43–58. Hebrew Nominal Sentences and the Stage/IndividualLevel Distinction. 1961. Amsterdam: Elsevier. Muysken and H. In Gramática Descriptiva de la Lengua Española. H. Madrid: Espasa Calpe. In From Words to Discourse. Gili Gaya. Amsterdam: John Benjamins. Ph. and T. New York: Academic Press. Fernald. The temporal interpretation of predication.). Leonetti. Grimshaw. Foris: Dordrecht. CA: CSLI Publications. Felser. Demonte (eds. 2002. T. New York: Academic Press. 233–254. T. Fernández Leborans. P. P. von. T.References 265 Escandell-Vidal. Hoper (ed.-J. Madrid: Espasa Calpe. Logic and conversation. van Riemsdijk (eds. J. H. K. L. Folli.). Israel. Bar-Ilan University. 1994. Fintel. J. 159– 179. 1995.). Los complementos adverbiales temporales: La subordinación temporal. J. University of Massachusetts. and A. Curso Superior de Sintaxis Española. H.). R. van Hout (eds. M.-J. 77–107. MA: MIT Press. V. 1975.). 1982. 1995. Cardinaletti and M. and G. P. Formal Methods in the Study of Language Amsterdam: Mathematical Centre Tracts. 2000. I. MA thesis.). Notes on world view and semantic categories: Some Warlpiri examples. Janssen and M. Amsterdam: John Benjamins. Fernández Leborans. and M. . de Swart. dissertation.). 1984. In Events as Grammatical Objects. Gutierrez Rexach (ed. Hoekstra. L.. In Perspectives on Aspect. Hale.). Coercion and the stage/individual distinction. Bosque and V. 115– 163. Dordrecht: Springer. J. García. 2005. M. Stokhof (eds. P. I. Guasti (eds. Verba 22:253–284. Restrictions on Quantifier Domains. 1994. K.). Verbal Complement Clauses: A Minimalist Study of Direct Perception Constructions. Bosque and V. Amherst. 1999. In Small Clauses [Syntax and Semantics 28]. Givón. Guéron. Argument Structure. Barcelona: Spes. Filip. In Speech Acts [Syntax and Semantics 3]. C.). La predicación: Las oraciones copulativas. 1990. 2357–2461. 1999. Las construcciones con el verbo estar: Aspectos sintácticos y semánticos. Studies in the Linguistic Sciences 29:43–63. Tenny and J. Cambridge. 81–105. 39–96. T. Groenendijk. Morgan (eds. Grice. Pustejovsky (eds. In Tense-Aspect: Between Semantics and Pragmatics. Cole and J. Demonte (eds.

Jäger. Herweg. H. Janssen. 1994. Glot 12:19–35. A Compositional Semantics for Aktionsarten and NP Reference in English. 1999. In Gramática Descriptiva de la Lengua Española. Madrid: Espasa Calpe. Kamp. and meanings. Higginbotham. Cambridge. Amsterdam: Mathematical Centre Tracts. Toward an explanatory semantic representation. 1995. E. Heycock. Demonte (eds. dissertation. 1972. R. Ohio State University. 277–322. Semantic Interpretation in Generative Grammar. K. 1991. 53–109.). Amherst.D. J. and the semantics of the copula. Cambridge. MA: MIT Press. Heycock. Higginbotham.-L. N. El infinitivo. Keyser (eds. H. Hinrichs. A theory of truth and semantic representation. and perhaps even quantification in English. 1983. Cambridge. 2197–2356. What Counts. 29–48. J. 1993. New York: Garland. T. The proper treatment of measuring out. 1982. 1976. 1990. Heim. Ramchand. Kempson (ed. The logic of perceptual reports: An extensional alternative to situation semantics. de Swart. models. In Mental Representations. Higginbotham. and G. Berlin: Zentrum für Allgemeine Sprachwissenschaft. Keyser. Cambridge: Cambridge University Press.). Stokhof (eds. Higginbotham. and S. Linguistics 29:969–1010. Jackendoff. 1996. Columbus. 1981.D. .). Accomplishments. 1985. states. 1988. 1989. The internal structure of small clauses: New evidence for inversion. North East Linguistic Society 25:223–238. de and H. J. 1996. E. C. 1999. R. MA: MIT Press. On semantics. Over indefiniete objecten en de relatie tussen syntaxis en semantiek. G. Linguistics Inquiry 7(1):89–150. Hale and S.. dissertation. The Journal of Philosophy 80:100–127. R. Higginbotham. Jackendoff. J. Perfective and imperfective aspect and the theory of events and states. K. MA. telicity. University of Massachusetts. Hornstein. MA: MIT Press. The stage-level/individual-level distinction and the Mapping Hypothesis. M. Stage levels. and M. M. Ph. Contexts. J.). As Time Goes by: Tense and Universal Grammar. Layers of Predication: The Non-Lexical Syntax of Clauses. In Formal Methods in the Study of Language. I. Hernanz. R. J. Natural Language & Linguistic Theory 14:305–354. Herburger. Jackendoff. Proceedings of GLOW in Asia 2:72–82. Cambridge. J. Hoop. University of Oxford. Bosque and V. 2000. Ph. C. Linguistic Inquiry 16:547–593. MIT Press. 2000. Papers in Linguistics 14:65–94. J. The Semantics of Definite and Indefinite Noun Phrases. Ms.266 Individuals in Time Hale. In The view from Building 20. 1986. I. On argument structure and the lexical expression of syntactic relations. Groenendijk.

P. On the absence of Case chains in Bambara. W. Koopman. S. Butt (eds. Lingua 85:211–258. Berlin: Mouton de Gruyter.). Klein.). postpositions. 1986. and C. 265–308. In Meaning. Scale structure. London: Routledge. 2000. H. circumpositions. Linguistics and Philosophy 1:337–355. and Interpretation of Language. MA. C. H. Kondrashova. In Subject and Topic. Koopman.).). The notional category of modality.). Keenan-Ochs. Stanford. J. Kratzer. Amherst. University of Massachusetts. MA: Cascadilla Press. and the semantics of gradable predicates. dissertation. E. 2001. Y. In The syntax of specifiers and heads: Collected essays of Hilda Koopman. 1998. June 2001. and L. Phillips. von Stechow (eds. H. 1991. N. 1963. and B. A time-relational analysis of Russian aspect. Subjects in Japanese and English. 1977. In Words. W. Temporal reference frames and the imperfective paradox. N. and D. Ben Gurion University of the Negev. In The Projection of Arguments. and C. Language 81:345–381. The Russian copula: A unified approach. Action. Eikmeyer and H. Worlds. 250–269. H. What must and can must and can mean.D. Kiparsky. Paper presented at the workshop The Syntax of Aspect. Kearns. 1996. Natural Language & Linguistic Theory 10:555–594. Ph. The partitive revisited. C. 1983. 1976. Schwarze and A. Greuder and M. 1995. degree modification. London: Routledge. Tense in texts. MIT. Li (ed. A. Use. Berlin: Mouton de Gruyter. R. C. Ann Arbor: Michigan Slavic Publications.D. Somerville. N. . Proceedings of WCCFL 22:287–300. Language 71:669– 695. Topics as a discourse notion: A study of topic in the conversations of children and adults. Schieffelin. W. The Semantics of the English Progressive. Cambridge. Toman (ed. CA: CSLI Publications. A. Kiparsky. Formal Approaches to Slavic Linguistics: The College Park Meeting 1994. 38–74. 1995. dissertation. 1991. 337–384. Time in Language. 1981. Rohrer. Rieser (eds. Partitive Case and aspect. A. Klein. and particles. Kondrashova. Ph.References 267 Kamp. and Contexts: New Approaches in Word Semantics. Kenny. New York: Academic Press. 2005. P. and Will. Semantic functional projections? ∃P: Evidence from Russian. London: Routledge and Kegan Paul. The position of subjects. Kratzer. Kitagawa. Prepositions. 1994. Baüerle. Sportiche. Proceedings of the Western Conference on Linguistics 24. 2003. J. Kennedy. H. McNally. 1992. Koopman. Kazanina. Emotion.

Telicity and the meaning of objective Case. Linguistic Inquiry 25:609– 666. E. 1976.). Kratzer. 1994. Kratzer. A. 109–137. Kuroda. M. dissertation. G. 1970. J. In The Generic Book. 1991. Irregularity in syntax. Luján. Linguistic Inquiry 19:335– 391. 1996. Rooryck and L. Kratzer. 1992. J.). J. Rinehart and Wilson. R. J. Comments on D. February 2000. 1981. Hensey (eds. 1995. The progressive. Lingua 54:165–210. Oxford: Blackwell. I. 247–284. Los Angeles. and F. Washington.). Structures for semantics. Chicago: University of Chicago Press. 1972. The event argument. University of Massachusetts. 125–175. Szabolcsi (eds. Cambridge. In Genericity in Natural Language [SNS-Bericht]. Zaring (eds. Davidson’s “The Logical Form of action sentences. The Determination of Grammatical Relations in Syntax. Krifka (ed. . N. Carlson and F. Natural Language Semantics 1:1–32. Landman. Pittsburgh. 2004. Some problems with tense in PTQ. A. Reference and proper names. Lewis. Foundations of Language 9:153–185. M. Martin. 2000. Larson. 1988. Cambridge. Ms. Lemmon. Kratzer. M. PA: University of Pittsburgh Press. Paper given at the Berkeley Linguistic Society. MA: MIT Press. Ph. A. Kratzer. G. 1991. Krifka. 1977. Longobardi. Thematic relations as links between nominal reference and temporal constitution.). Rescher (ed. Stanford. S. Stative Verbs and Adjectives in English [Harvard Computational Laboratory Report NSF-17]. Guéron and J. In The Syntax of Time. G. 1994. 389–423. A. Amherst. The categorical and the thetic judgment: Evidence from Japanese syntax. Building statives. 1967. Ladusaw. A. A. Lakoff. Current Studies in Romance Linguistics. 1992. On the Plurality of Worlds. CA: CSLI Publications. Luján.).). Pelletier (eds. New York: Holt. W. Stage-level and individual-level predicates. 29–53. 1966. Tübingen: University of Tübingen.” In The Logic of Decision and Action. Landman. F. M. Severing the external argument from the verb. On the double object construction. F. Dordrecht: Kluwer.. 1986. MA: Harvard University. DC. 1988. Dordrecht: Kluwer. G.D. Texas Linguistic Forum 6:89–102. Stage-level and individual-level predicates. Georgetown University Press. University of California. In Lexical Matters. Lakoff.). Lecarme (eds. In Phrase Structure and The Lexicon [Studies in Natural Language and Linguistic Theory]. Sag and A. 96–103. The Spanish copulas as aspect indicators.268 Individuals in Time Kratzer. A.-Y. D.

. F. Proceedings of WCCFL 19:280–301. Dekker. 1978. de. El aspecto léxico. Madrid: Espasa Calpe. Amsterdam: North Holland. A. Some structural analogies between tenses and pronouns in English. R. S. Ms. 1984a. The instrument of inversion: Instrumental case and verb raising in the Russian copula. Venema (eds. University of California. Y. On the Temporal Interpretation of Noun Phrases. MA: MIT Press. Linguistics and Philosophy 7:243–286. Mourelatos. Existential Sentences in English. 1973. MIT. Journal of Philosophy 70:601–609. 2977–3061. Determiner Systems and the Quantificational Strategies: Evidence from Salish. B. In Gramática Descriptiva de la Lengua Española. Dordrecht: Kluwer. Ph.). P. 2000. 281–312. Compositionality. B. Partee. P. In Proceedings of the 4th Amsterdam Colloquium: Variety of Formal Semantics.D. Somerville. 1984b. Partee. dissertation. MA. Dordrecht: Foris. 1996. Attitudes. Estructura Semántica del Sistema Preposicional del Español Moderno y Sus Campos de Usos. Amsterdam: ILLC/Department of Philosophy. O. 1988. MA. R. L. M.References 269 Matthewson. 1982. MIT. Los Verbos de Movimiento. M. Milsark. and Scope. Puerto del Rosario: Servicio de publicaciones del Excmo. Tense. B. McConnell-Ginet. P. dissertation. Musan. In Proceedings of the Eleventh Amsterdam Colloquium. A Primer in the Semantics of English: “Some Nuts and Bolts” [Course Reader]. Ogihara. 2000. Events.). A. Veltman (eds. T. Ph. In Linguistic Structures Processing. 241–256. R. Munro. Naumann. Cabildo Insular de Fuerteventura. 1991. Nominal and temporal anaphora. 1982. Partee.).. Ph. Natural Language Semantics 5:271–301. Matushansky. and lifetime effects. Syntactic relations in Western Muskogean: A typological perspective. Musan. 1990. Active/agentive Case marking and its motivations. Tense. T.D. and States. MA: Cascadilla Press. John is easy to please. Partee. Stokhof and Y. and C. and L. Linguistics and Philosophy 2:415–434. 1996. 1974. Language 58:81–115.D. Adverbs and Logical Form. I. 1997. E. Mithun. Gordon. 1997. 1999. Miguel. B. University of British Columbia. Landman and F. University of Amsterdam. Parsons. Los Angeles. Bosque and V. Language 67:510–546. Cambridge. Zampolli (ed. dissertation. 281–312. Morera. Cambridge. Demonte (eds. T. Piñón. 2001. M. Decomposing the progressive. Events in the Semantics of English. Parsons.). Language 58:144–184. 1977. 1995. Madrid: Visor Libros. processes. predicates. Morimoto. G. Cambridge.

University of Oxford. 420–433. Stanford. Dordrecht: Kluwer. Rosen. Reuland and A. 1998. MA: MIT Press. and S. 1995. In The Representation of (In)definites. 1976. 1996. A. 179–206. Durative adverbials for result states. 98– 129. C.-Y. In Small Clauses [Syntax and Semantics 28]. and J.). Cardinaletti and M-T. Ph. . 1947. Rosen. E. and the structure of IP. 1995. MA: MIT Press. Pustejovsky. Ph. ter Meulen and R. and Small Clauses: A study of Israeli Hebrew. 1995. The Generative Lexicon. C. CA. J. In The Projection of Arguments. M. 1998. Reinhart. The theta system: syntactic realization of verbal concepts. H. Stanford University. 1993. Rapoport. Ritter. The semantics of copulative constructions in Portuguese. Wall. The geometry of events. Word and Object. T. A. B. Washington. DC: Georgetown University Press. Greuder and M. In Studies in Generative Approaches to Aspect [Lexicon Project Working Papers 24]. C. Piñón. 1987. Piñón. 1960. Cambridge. J. Carnie. Quer. dissertation. New York: Academic Press. Probus 13:81–111. In Proceedings of WCCFL 18. Wh-in-situ: Movement and unselective binding. C. Butt (eds. Cambridge. Pollock. New York: The Free Press.). Copular. In Current Studies in Romance Linguistics. Ph. D. Strong and weak predicates: Reducing the lexical burden. Pesetsky. Berkeley.. Cambridge. In Lexical and Conceptual Structure. University of California.D.. 2000. Quer. 1989. Guasti (eds. Ramchand. ter Meulen (eds. Cambridge. Pustejovsky. First phase syntax. E. 47–81. MA: Cascadilla Press. Querido. A Mereology for Aspectuality. Pinker (eds. Delimiting events in Syntax. 343–366. J. W. J. Haugen and P. Tenny (ed. Piñón. Elements of Symbolic Logic. dissertation. Raposo. G. Linguistic Inquiry 20:365–424. Cambridge: Cambridge University Press. MIT. Reichenbach. 2003. Linguistic Analysis 26:29–62. MA: MIT. UiL OTS Working Papers.).D. 2000.D. T. Two types of small clauses (Toward a syntax of theme/rheme relations. Nominal. Stanford. University of Utrecht. Ritter. Proceedings of the Berkeley Linguistic Society 26:435. Oxford: Blackwell. J. 1999. MA. Mood at the Interface. 1988. 2001.). J. 1991. Verb movement. Universal Grammar. V.). Interpreting mood.270 Individuals in Time Partee. Mathematical Methods in Linguistics. Quine.). Levin and S. J. S. University of Utrecht.). Somerville. B. A. The syntax of event structure. Pustejovsky. CA: CSLI Publications. Luján and F. E. A. 1987. Uriagereka. 135–164. and S. Bird. dissertation. Happening gradually. 19–39. E. Hensey (eds. Ms. W. Norquest (eds.

Small clauses and copular constructions. Natural Language Semantics 7:347–420. Russell. R. The syntactic representation of linguistic events: State of the article. Farkas. Stowell.). 1979. Events and Predication: A New Approach to Syntactic Processing in English and Spanish. In Logical Form and language. Chicago: Chicago Linguistic Society. Lecarme (eds. The Concept of Mind. Ryle. M. Towards a semantic characterization of ser and estar. In Semantics of Natural Language. Preyer and G. 1999. Language 51:32–48.). C. C. Mind 14:479–493. S. Referential properties of Spanish NPs. 1972. Schein. 1905. Cardinaletti and M. Peter (eds. M. Perlmutter and C. . Guasti (eds. New York: Academic Press. Stanford. P. Events and the semantic content of thematic relations.). R. J. 2004.-L. Rothstein. London: Barnes and Noble. Rosen. Pragmatics. Dordrecht: Kluwer. Rivero. February 2003. D. Rosen. 2003. G. R. C. In Pragmatics [Syntax and Semantics 9]. Harman (eds. GLOT International 4(2):3–11. Sanz. T. The Parameter of Aspect. 597–621. 27–48. In Studies in Relational Grammar 2. and S. 1978. Cole (ed. S. W. Rosen. 2000. Stalnaker. Oxford: Clarendon Press. B.).). 1995.References 271 Ritter. 38–77. M. Todrys (eds. Ser and estar: A matter of aspect. The interface between semantic roles and initial grammatical relations. M. 1974. Stalnaker. A. C. E. 458–471. 1999. D. Smith. MA: MIT Press. 380–397. Jacobsen and K. Assertion. In The Syntax of Time. Rosen (eds. What was there before there was there. M. North East Linguistic Society 22:411–426. 2000. In Small Clauses [Syntax and Semantics 28]. anglais et allemande. 187–238. Rosen. Guéron and J. 1975. The domain of tense. Rothstein. University of Southern California. Activities: States or events? Linguistics and Philosophy 22:479–508. 1984. Schmidt. Paper presented at the Workshop on the Philosophy of Events. Event structure and ergativity. 1991. C. L’adjective de relation en français. S. W. Schmitt. Pustejovsky (eds. Los Angeles. Ritter. Hispania 57:68–75. Smith. On denoting. B. 263–344. 2002. Dordrecht: Reidel. 1999. Cambridge. T. Smith. On the aspectual nature of subject splits. and S. Amsterdam: John Benjamins. 315–332. C. Fine-grained structure in the eventuality domain: The semantics of predicative adjective phrases and be. D. Chicago: University of Chicago Press. E. Tenny and J. Göppingen: Alfred Kümmerle. Roldán. 1949. In Papers from the Parasession on the Lexicon. 1992. T.). In Events as Grammatical Objects.). CA: CSLI Publications. New York: Academic Press. 1972. G.). Davidson and G.

D. dissertation. MA. 1993. Stump. Los Angeles. Tungseth. Tenny and J. and J. MIT.).. 145–185. In Universals of Human Language: Part 4: Syntax. C. Grammaticalizing Aspect and Affectedness. 2000. Kempchinsky and R. Proceedings of the 1994 Annual Conference of the Canadian Linguistic Association.). 232–262.).). T. Ways of Scope Taking. Tenny. Tenny. Tenny. Talmy. P. Ms. Stanford. MA. 1981. 277–291. T. S. Aspectual Roles and the Syntax-Semantics Interface. Stowell. T. Pustejovsky (eds. G. In Events as Grammatical Objects. Cambridge.. and X-bar theory. Stowell. 625–649. 1998. In Phrase Structure and The Lexicon [Studies in Natural Language and Linguistic Theory]. 1981. Dordrecht: Kluwer. P. Rothstein (ed.272 Individuals in Time Stowell. C. Subjects. Cambridge.). 147–169. specifiers. PP. C. D. Ph. Stanford. Natural Language & Linguistic Theory 16:347–385. The Origins of Phrase Structure. M. . A. CASTL. Strategies for scope taking. CA: CSLI Publications. Dordrecht: Kluwer. Swart. M. Slabakova (eds. 1996. de. 109–154. CA: CSLI Publications. Toronto: Toronto Working Papers in Linguistics. 2005. A. Ms. In Alternative Conceptions of Phrase Structure. Victoria University of Wellington. 1994. The alignment of arguments in adjective phrases. Event structure in syntax. 105–135. 1978. Ph.D. 1996. Tenny. Syntax of tense. dissertation. 1994. Ph. 2004. Kluwer. MIT. Wa/Ga Subjects and in Japanese and Subdivisions of Tense. Zaring (eds. CA: Stanford University Press. Events as Grammatical Objects. Chicago: Chicago University Press. In Perspectives on Phrase Structure: Heads and Licensing [Syntax and Semantics 25]. Event phrase and a theory of functional categories. The Aspectual Interface Hypothesis [Lexicon Project Working Papers 31]. Rooryck and L.D. dissertation. Svenonius. dissertation. New York: Academic Press. L. Aspect shift and coercion. Baltin and A. Columbus. H.). 2000. Cambridge. Szabolcsi (ed. Dordrecht.). T. Szabolcsi. University of California. In Ways of Scope Taking. Dordrecht: Kluwer. Kroch (eds. 1987. path. University of Tromsø. 1989. Dordrecht: Springer. 1996. Travis. Ohio State University. In Aspectual Inquiries. J.). MA: MIT. (ed. J. S. The phrase-structure of tense. A. Torii.). 1989. T. 1991. Figure and ground in complex sentences. Stowell. Travis. Greenberg (ed. L. 559–570. 2000. Ph. C. Szabolcsi. Spatial P in English. Stanford. L. Pustejovsky (eds. Stowell. The Formal Semantics and Pragmatics of Free Adjuncts and Absolutes in English. and the telic/atelic distinction in Norwegian motion structures. C.

Dordrecht: Foris. ter Meulen (eds. The Philosophical Review 66:143–160. 271–292. Verkuyl. M. Kiss (ed. Verkuyl.References 273 Uriagereka. Cambridge: Cambridge University Press. Higginbotham. The Linguistic Review 1:81–114. J. Williams. 1993. In Topics in South-Slavic Syntax and Semantics. Tense. J.). The semantics of the progressive. Prepositional aspect and the algebra of paths. Tedeschi and A. 1984. Ms. 1981. Zemach. 1957. Synthese 31:509–515. 2005. Z. Events as dividuals: Aspectual composition and event semantics. Dordrecht: Reidel. In Speaking of Events. Argument structure and morphology. DimitrovaVulchanova (eds. In Tense and Aspect [Syntax and Semantics 14]. 1967. G. Westerståhl. A Theory of Aspectuality: The Interaction between Temporal and Atemporal Structure. J.-L. 1987. 45–71.).. Vendler. Vendler. University of Washington. Paths in the semantics of verbs. Verkuyl. H. Zwarts. K. H.). Amsterdam: John Benjamins. 1972. Verkuyl. Zaenen (eds. 1981. Zwarts. Linguistics and Philosophy 28(6):739–779. Vlach. Oxford: Oxford University Press. Amsterdam: John Benjamins. Dordrecht: Foris. New York: Academic Press. Times as temporal argument structure. Z. van. H. 2000. Verkuyl.). Seattle. Ms. Determiners and context sets. L. 169–205. Event Structure. P. 1988. An F position in Western Romance. F. 1999. 2006. F. E. .). Verbs and times. Hellan and M. H. Levels of Representation in the Lexicon and in the Syntax. Cambridge: Cambridge University Press. 1990. J. Radboud University and Utrecht University. Linguistics in Philosophy. 1975. K. D. Voorst. In Discourse Configurational Languages. J. On the Compositional Nature of the Aspects. Aspectual classes and aspectual composition. On the adequacy of a type ontology. 1989. NY: Cornell University Press. H. E. Varzi (eds. 153–175. 123–162. 1994. Linguistics and Philosophy 12:39–94. In Generalized Quantifiers in Natural Language. Pianesi and A. and aspectual composition.. Zagona. Ithaca. van Benthem and A. aspect. J. Zubizarreta.

.

194 Arche. 224 Chierchia. 38. 144. 216. 12. 55 Bertinetto. 153. 14. 30 Kamp. 157. 5–8. J. 36. 117. H. 240. M. 80. K. 150 Abusch. 118. 57. 126. 112–116. M. 97. 35. 225. 106. 176 Bache. 188 Cooper. 213 Grimshaw. 218 Bello. 124 Davis. I. 80. P. 153. 212. 144 Delfitto. 191 Egg. M. R. P. 144. 8. 150. 127. 18. K.-L. 3. 252. 138. 251. de. 25 Givón. 91 Donnellan. 165. 153. J. M. 201 Anderson. 16. D. J. 194 Dixon. 9. 188. M. 156. 98. 109. M. G. D. M. P. S. 29. 116–119. 14. 81. 110. 173 Diesing. 40. 233 Escandell-Vidal. C. J. 32. 214. 71. 251 Fernández Leborans. 136. 173. C. 255 Filip. 122. 14. 222. 222 Jackendoff. 11. H. J. 100. B. 233 Davidson. D. 37 Hoop. 40 Emonds. H. 84. 240 Heim. F. 134. 120 Carlson. 99. 154. 105. L. 137. 32. 130 Jäger. 77 Kazanina. G. 30. E. 9. 12. 192. T. 202. 187 Asher. 37. 22. 191 Bach. M. 119. 127 García. 5. L. 197. 20. 225. 118. 157. 18 Baker. C. 120– 123. 125. 202–204 Herweg. 40. 15 Grice. 10. L. R. 1. 25. M. 70. H. 40. H. von 213–215. 250 Hinrichs. 214 Becker. K. 155. R. 8. J. 189. 176. 247. J. C.-J. 21. T. 209. 108. G. V. 55 Comrie. 251. I. 85. 240 Demonte. 248 Bosque. 150. 198 Carlson. N. 233 Bybee. 84. J. 188. 144. 9. Y. 45. 80 Heycock. 111 Barwise. 121 Gordon. 102. R. 34. 194. A. 195 Davies. 38. 245. N. 109. M. 37. 203 Gili Gaya. 218. 32. 94 Greenberg. 56. 40. 169. 191 . 151. 31. 168 Hernanz. 55. 43. 38. H. 190. G. 254 Brugger. 38. 254 Felser. 191 Demirdache. M. 1. 80. 32. 113 Cinque. H. 207. E. 147. 35–37 Beghelli. 45. 3. 40 Guéron. M.Name Index Abney. 8. 37 Hale. S. 156. 5. 115 Benua. 232 Dowty. 80 Aristotle 40. 80 Bennis. 214 Herburger. 242 Chomsky. 173. 56 Collins. 250. 112 Berardo. 16–18. 145. 1. E. 249. 232 Allen. 135. L. 70. 81. 30 Hornstein. 159. 45. 154. 149. N. 38. 177. 173 Fintel. 83. 218 Folli. 37 Higginbotham. A. 131. T. 41. 3. 11. D. 43. M. 147. H. 126. 254 Depraetere. 37 Fernald. 84. N. 136. 71 Carrasco. 16 Bennet. 8. 194. 173. M. 120. J. 70. 143. G. M. I. 175. F. 80 Hoekstra. 132. V. 37 Enç. 32. 32. 1. 214 D’Angelo. 191 Borer.

94 Musan. M. M. 11. D. 208. 251 Quer. 240. R. 148–150. 178. 152. A. 60. 127 Uriagereka. R. A. 189. A. J. 61. 191 Ogihara. 244 Stump. 113. 243 Travis. 8. 15. 130 Mourelatos. J. 155. 122. 97. 3. 254 Lewis. 176 Szabolcsi. 122. 22 Meulen. 144. 29 Talmy. 29. E. 55 Rosen. 59. M. 86–90. 89. 143. 80 Ritter. 102 Matthewson. 113. 56 Sanz. T. 77 Roldán. 127 Swart. 195 Leonetti. 116. Y. C. 94 Morera. 16. 96. 22 Kenny. 1. 114 Schieffelin. 112 Tungseth. 193. E. 119. 18 Piñón. 38 Mithun. M. 58.-L. 47 Keyser. 86. T. 126. B. 191 Partee. 33. C. 3. 45. G. M. 196. 37. 41. 207. P. G. J. J. S. 55. 194 Klein. 208–212. C. V. 208. 162. 191 Pollock. 240. B. B. 163. 81. 29. 43. 46. 109. S. 74. L. 116. 214 Perkins. 148. 250 Ladusaw. 32. 242–244. 37. 148. H. 37 Pustejovsky. 114 Kuroda. 14. 119. 178. P. T. 145. 32. C. S. 40–44. 61. 194. 135 Kiparsky. 145. 194 Kratzer. 150 Reinhart. 191 Larson. 225. de 48 Milsark. 31. J. 245 Martin. 109. D. 240 Rothstein. 119. 214 Luján. 216 Uribe-Etxebarría. 144. 70. 84. 147. 222. 252. 225. 25–32. 3. 240 Kondrashova. 232. 93. 61. M. 120– 123. 70. 96. O. 236 Naumann. 218 Phillips. 117. 8. 193–196. 150. 25. L. 71. 233 Rohrer. 213. 216 Reichenbach. de 30. 71. 240 Rivero. 144. 78. 226. R. 16 Rosen. 73. 152. 124. 188. 49. S. 213 Stowell. M. ter 148. 215 Stalnaker. 251. 94. 57 Masullo. 120 Pesetsky. L. 61. 75. C. 50 Landman. G. 117. 164. 205. 192. 12–14. 77. 15 McConnell-Ginet. S. M. 38. 233 Querido. 6. 232 Lakoff. J. D. 202. 38. 153. H. 70. G. 152. 122. 85. 127. 14. 176 Munro. 43. 55. 176. 106. 118 Matushansky. 189. 143. S. 217–220. 15 Raposo. E. 3. 36–38. 162. J. 50. M. 10. C. T. 226 Pagliuca. 113 Kitagawa. 250 Rapoport. 12–14. 8. 235. 222. 220. W. P. 80. 34. 71 Ramchand. P. 245. F. 25 Quine. 12. 167. W.276 Individuals in Time Kearns. 164 Miguel. 145. 43. 190. 138. C. 88 Lemmon. 211. G. 113. 214 Svenonius. A. 140. 232–235. 156 Tenny. 191 Picallo. T. R. T. 215. 213 Schmidt. 115. W. B. 28. 116. 33. P. 15 Koopman. 40. 40. 38. 136. 55. 44. W. 154. S. 11. 192. 107. 109 McNally. 156. N. 144. A. 84. J. 172. 232 Ryle. 214 Longobardi. 112 Schein. 8–12. E. 18 Schmitt. 213 Kennedy. 188 Keenan-Ochs. 131. G. 97. C. 40. Y. 120 Parsons. 251 Sportiche. 15. 237. 151. 112. 138. 126 Morimoto. 225. 143 Russell. 147. 5. 56. 240 . 136. 12–14. 98. R. 38. M. 43. 214. M. H. 3. 205. 218. 102. 40–43. 5. 3. 112. L.-Y. 207. 106. 154. 43. 26. 165. 79. 110– 112 Torii. 84. 41. 251 Krifka.-Y. 211. E. 127. R. A.

81 Verkuyl. 73.Name Index 277 Vendler. J. 73. 208. 110. 165 Zubizarreta. 41–43. 148. 75. 81. 157. 191 Voorst. 148. 136 . 129. 164 Westerståhl. van 40. 43. 177. 192 Vlach. G. F. 80. 3.-L. E. H. 70. 39–41. 8 Zagona. D. 163. 120. 50. 56 Zwarts. R. K. 214 Williams. 53. E. Z. 110– 112 Wall. 158. J. 150 Zemach. M. 81. 130. 164. 113.

.

see Temporal interpretation Compositional(ity) (of aspect) 43. 218. 94. 141–143. see also Context (salient) Discourse topic 213. 104. 85. 139. 271 Semantic classification of 91 Affected argument 96–98. 107. 130–135. 217. 85. 83. 129. 84. 117. 81. 89. 215. 236 Bambara 15 Bantu 120 Bare plurals 6. 205. 40. contextual restriction 208–210. 144 Agent 53–57. 235 Controllability. 214–217. 116. 192. 80 Event(ive) argument 9. 51. 73. 128. 86. 142–144. 89. 66. 235. 75. 133–136. 90. see also Spatiotemporal argument Event Correspondence Rule 110 . 72. 59–61. 148. 189. 48. 217. 112. 150. 135. 123. 45. 251–254 Complement clauses temporal interpretation of. 62. 247. 201 Discourse background 213. 247. 208. 117. 136. 62. 22 Participial adjectives 21. 208. 219. 22. 254. 105– 107. 196. 253 Density 188. 18. 253 Cut-short adjectives 21. 212. 144. 258 Dyadic adjectives 86–89 Dynamic adjectives 2. 106. 175. 188. 127. 215. 56. 61. 176. 93–95. 108. 14. 240 Context-dependent interpretation 4. controller 27. 247 Background (knowledge) 213. 103–108. 255 Perfective adjectives 25. 22. 11. 113. 105. 213. 96. 83. 110. 57– 59. 47. 105. 194. 51. 90. 192 Energeia 41. 196. 91. 31. 118. 69. 164–166. 236. 116. 99. 33–37. 108. 84. 112 Argument structure 8–10. 101. 39. 144 Aktionsart 32. 137. 129–131. 18. 118. 123. 235. 105. 235. 141. 236 Control (between temporal heads) 151. 91. 8. 53. dynamic events 3. 220. 220. 84. 135. 76. 186. 209. 108. 137. 139. 197. causer 53–57 C-command domain 222. 189. 93. 144 Algonquian languages 55 Animacy. 94. 246 Adjectives Agentive adjectives 85. 106. 179. 252 Context (salient) 13. 72. 117 Davidsonian argument 8. 235 Domain of quantification 208. 156 Chichesaw 94 Coercion 162. 255 Argument mapping 111. 10. 77. 222–225. 105. 84. 80. 148.Subject Index Additivity property 70. 253. 257 Gradable adjectives 18. 132 Aspectual classification of 93 Classifying/classificatory adjectives 17. 228. 244 Aspectual Interface Hypothesis 111 (A)telic(ity) 40. 147. 35–37. 67. 42. 59–61. 107–110. 195. 38. 39. 191. 94. 205. 133–137. 148. 3. 195. 218–220. 90. see also Event(ive) argument Deadjectival verbs 136 Definiteness Effect see Quantification restriction Degree modifier 18. 135. (in)animate 53–57. 236 Discourse prominence 235. 247 Qualifying adjectives 17. 240. 250 Cause. 220. 224 Central/noncentral coincidence 118–123. 117. 14. 86. 252. 113– 115. 83–85 Case-checking 113 Categorical (judgment) 12. 88. 115. 10. 61. 59–61. 129. 134. 233. 194. 35. 174–177. 124. 215 Basque 120 Be 7. 214 Dutch 121 Dynamicity. 61. 29. 235 Contextual variable. 83. 218. 111–113. 218. 112. 176. 76.

55. 257 Perform(er) 53. 147. see also Utterance Time Stativizer 115. 244. 116. 189. 226. 252 Realization function 6. 35. 229. 94 Pomo 94 Prepositions Directional prepositions 127. 131. 134 There-sentences 5. 83. 137 Subinterval property 70. 175. 158. 153–160. 14. 230. 242. 148–152. 178. 129 Progressive 153–163. 155. 196. 209. 186. 244. 72. 148. 226. 144. 159. 144. 136. 189– 192. 129. 148 Sequence of Tense 226. 158. 105. 183. 67. 123– 126. 241. 64. 134–136. 156 Hawaii-Creole 120 Hebrew 15 Heterogeneous (predicates) 72. 232. 233–235. 158–160. 183. 168 French 121 Generic operator 11. 132. 250 . 158–160. 131. 76. 189. 258 Progressive paradox 191. 57. 195. 194 Reference Time (RT) 3. see Temporal interpretation Russian 15 Salish 118. 187–192. 100. 165. 147. 191 Event(uality) Time (ET) 3. 148. 186. 212 Mereological properties 3. 132. 128–131. 12. 211. 234 Situation aspect 39 Small v 55 Spatiotemporal argument/variable 8–10. 111. 212. 192. 245. 128. see also Event(ive) argument Speech Time 122. 154. 148. 181. 207. 212. 142. 211. 221–223. 72. 178. 155–157. 202 Ground 119. 186. 189. 11. 79. 249 Existential coda 12 Existential quantification 8. 258 Hindi 120 Homogeneous (predicates) 70. 162. 173–180. 207. 106. 160. 113. 221 Habitual 77. 246. 184. 172. 141. 123. 248 Swahili 120 Temporal interpretation of Complement clauses 76. 180 Perfective (preterit) 31. 125. 150. 31 Imperfective paradox see Progressive paradox Initiation. 134–136. 165. 157. 122. 37 Thetic (judgments) 12–14. 181 Existential reading 6. 227. 233 Perfect entailment 47. 227. 156 Focus 13. 123. 258.280 Individuals in Time Event-object 110. 192 Prospective 120–122. 191. initiator 61. 229. 171– 175. 117. 12. 132. 94. 187. 240. 86. 122. 236 Theme 105. 94 Margi 121 Maxim of Quantity 85. 158. 194. 80. 215 Goal(Path) 96–98. 230. 186. 134 Instigator 55. 152. 176–178. 177–180. 246. 161. 35 Figure 119. see also Subinterval property and Additivity property Narration temporal interpretation of. 246. 248. 224. 6. 136 Place/situational prepositions 120. 147. 233. see Temporal interpretation Past-shifted reading 76–78. 179. 14. 227 Simultaneous reading 76–79. 79. 257. 172. 227–229. 180. 129. 201. 184. 148. 192 (Im)perfective contrast 25. 173. 136. 162. 167. 187. 35. 248. 141. 26. 116. 178. 224. 249. 58. 189. 154–157. 249 Relative clauses temporal interpretation of. 184. 240. 172–174. 112 Juba Arabic 120 Kinesis 40 Krio 120 Lakhota 55. 147. 258 Imperfect Continuous 76. 80. 222–225. 242. 72. 175–177. 144. 148–152. 148. 158. 176. 112. 131. 186. 242. 192. 14 Generic reading 6. 135. 115. 176. 222. 118. 88. 189. 174–178. 230 of Narration 77–80 of Relative clauses 223. 208. 130. 226. 11. 231. 246–248. 159 Quantification restriction 11 Quantity 113–116.

see also Speech Time Viewpoint aspect 179. 240–242. 178. 219. 101. 231–237. 154– 157. 67. 217–225. 249. 104–106. 161. 210–213. 151. 226. 235 . 228. 222. 207 Volition(ality) 51. 192. 189. 233. 122. 258 Undergoer 134 Uniformity of Theta Assignment Hypothesis (UTAH) 111 Utterance Time (UT) 76. 210. 152. 4. 77. 150. 150–152. 220. 211.Subject Index 281 Tiwa 55 Topic Time (TT) 3. 176. 192. 229. 224. 186. 159. 151. 234 . 231. 244. 211. 172. 171–173. 61. 212. 109. 222–224. 53–58. 193. 207. 134 Warlpiri 118 Whenever-clauses 28–30 Zeit Phrase (ZP) 149. 178. 133. 215.

71 GELDEREN.): Clitic and Affix Combinations. 2004. 320 pp. 68 BREUL.): The Syntax of Nonsententials. 91 BOECKX. 2005. Copy and Match. Theoretical perspectives. Carsten: Focus Structure in Generative Grammar. A formal approach to the syntax and grammaticalization of verbal inflection. Kleanthes K. Language.): Formal Approaches to Function in Grammar. 2003. 2004. Satu Helena: Small Phrase Layers. A study of Finnish Manner Adverbials. A study of Hungarian. 2005. Marit: Nominal Phrases from a Scandinavian Perspective. 2004. and Henk van RIEMSDIJK (eds.): UG and External Systems. 87 JULIEN. 346 pp. Petra: The Syntax–Discourse Interface. 84 PAFEL. x. 2006042931x.): Minimalist Essays. 2005. Referentiality and Phrase Structure. 64 BOECKX. 378 pp. 366 pp. 81 FUSS. 2005. vi. 2003. Anna Maria (ed. 2006. 251 pp. 222 pp. 2005. 2004. viii. 65 MANNINEN. 2005.): Multiple Wh-Fronting. 259 pp. 89 VELDE. xvi. xii. xviii. 72 FUSS. xii. On the Anti-Locality of movement dependencies. 2006. 2005. 281 pp. Eugenia CASIELLES and Ellen BARTON (eds. 2005. 292 pp. 67 MIŠESKA TOMIĆ. Expected Novermber 2006 94 ARCHE. 372 pp. xvi. On the syntax of verb-initial languages. Heidi: The Distribution of Pronoun Case Forms in English. Cedric: Islands and Chains.): Agreement Systems. www.): The Function of Function Words and Functional Categories. The interplay between meaning. Melita and Arhonto TERZI (eds. Eric and Carola TRIPS (eds. 2003. 353 pp. 2003. brain and computation.benjamins.): Non-definiteness and Plurality.: Prolific Domains. . 73 CARNIE. 93 PROGOVAC. 2005. 434 pp. Representing and interpreting dependency. xvi. semantic and intonational approach. viii. Ljiljana. xvi.: Individuals in Time. vii. Katalin É. 285 pp. xvi. xvi. Walter: The Order of Prepositional Phrases in the Structure of the Clause. Cedric (ed. xii. Multidisciplinary perspectives. 62 CARNIE. x. In honor of Dimitra Theophanopoulou-Kontou. 292 pp. 77 ÖZTÜRK. 2006. Kate PAESANI. 399 pp. viii. aspect and the individual/stage distinction. Balkız: Case. x. 390 pp. xii. xii. predication and equation. 2005. x. In honor of Eloise Jelinek. John R. 2005. 275 pp. x. 66 GROHMANN. Eric: The Rise of Agreement. Line: Copular Clauses. 2006. 336 pp. 70 AUSTIN. Expected October 2006 92 BOECKX. 207 pp. 88 MOHR. 76 STAVROU. and syntactic structure. 74 HEGGIE. 432 pp. 385 pp. GROHMANN (eds. 86 COSTA. 338 pp. Impersonal constructions in the Germanic languages. Andrew. 80 BURKHARDT. German and Dutch. 514 pp. Tanja: Infinitival Syntax.): Diachronic Clues to Synchronic Grammar. Specification. te: Deriving Coordinate Symmetries. + index. Cedric (ed. 312 pp. x. 2005. Infinitivus Pro Participio as a repair strategy. 499 pp.): Verb Clusters. xiv. 85 MIKKELSEN. Jürgen: Quantifier Scope in German. Gréte: The Role of Agreement in Non-Finite Predication. xii. An integrated syntactic. Marcel den and Christina M. viii.): Adverbials. 2006. 2005. Jennifer R.com 95 VOGELEER. viii. 83 SCHWEIKERT. 210 pp. 268 pp.. Andrew. vi. 78 DIKKEN. Heidi HARLEY and Sheila Ann DOOLEY (eds. 346 pp. Olga (ed.): Studies on Agreement. 2004. 2005. 228 pp. 63 BOECKX.): Advances in Greek Generative Syntax. Lorie and Francisco ORDÓÑEZ (eds. 224 pp. xii. 79 SCHMID. A phase-based approach integrating Select. Elly van: Grammaticalization as Economy. Merge. context. María J.Linguistik Aktuell/Linguistics Today A complete list of titles in this series can be found on the publishers’ website. 398 pp. 2005.): Verb First.): Balkan Syntax and Semantics. 75 DI SCIULLO. Sabine: Clausal Architecture and Subject Positions. Svetlana and Liliane TASMOWSKI (eds. Resumption as stranding. 2006. xiv. 2004. 82 QUINN. Stefan ENGELBERG and Gisa RAUH (eds. Cedric and Kleanthes K. 90 DALMI. João and Maria Cristina FIGUEIREDO SILVA (eds. Tense. 409 pp. TORTORA (eds. Heidi HARLEY and MaryAnn WILLIE (eds. xvi. xiii. 2003. 69 KISS. 348 pp.

Sign up to vote on this title
UsefulNot useful