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

.........................1 .......3 IL Predicates as Inherently Generic Predicates.............................. Chierchia (1995) ....39 3..........................15 2............................................................................59 3...............Table of Contents Acknowledgments...2......40 3....................................................................51 3......1 The IL/SL Distinction as a Semantic Distinction...........ix Foreword... Milsark (1974) and Carlson (1977)........................................................ by Tim Stowell ...1..................................................38 Chapter 3 Event Classes and Individual-Level Predicates ...................1..1................................................................................4 The IL/SL Contrast as a Categorical/Thetic Distinction.............12 2............... Raposo and Uriagereka (1995)......................................................................................1 Inner Aspect............................1.........................................................25 2....3 Summary of Section 2...........32 2............1.............................2 .................................2 Agents in Event Structure ...... A Brief Stop at “Agentivity”..........................................53 3............................ Kratzer (1988.......................2..............14 2.......2..2. 1995) .......................2 ...........2........................................2 The IL/SL Dichotomy as a Syntactic Distinction............5 2....................3 Aspectual Differences among Individual-Level Predicates .........................1 Distribution of Copular Verbs in Spanish .............................2 Event Types and Event Structure.2.........8 2............................... Inner Aspect and Event Types ........2.....................39 3.............xi Chapter 1 Presentation of the Study ....1...........................................................................3 Summary of Section 3............61 ...........................5 2.............................. When There Is More Than One Copular Verb: Spanish Ser and Estar...33 2.1 Proposals about the Individual-Level/Stage-Level Dichotomy .....11 2....................1......................5 Summary of Section 2...............4 Summary of the Chapter ................53 3...................16 2........................1 Chapter 2 Individual-Level Predicates .. The Structure of Copular Constructions ..............................39 3..61 3.....1....1...........1 A Cluster of Notions .........5 2........................3 Summary of Section 3..............1 .2 Ser and Estar as an Aspectual Differentiation ..... Higginbotham and Ramchand (1996) ...........................2... Introduction to the Tests Discerning the IL/SL Status............3........

......3................................................................ 69 3.................................110 4............................4 ........143 4..........7 ...........1 Lexicalist and Logico-Semantic Approaches .137 4...........................84 4..........76 3.......85 4.....2 Differences between Activities and States ....... Summary of the Chapter ............ The Cruel-type as a Small Clause Integrant........90 4..........................116 4........................2 Summary of Section 3..................3..6............91 4...................4 The Issue: Aspectual Alternation in IL Copular Clauses ...............................................107 4..........2.........................4..................3.3 Adjectives Predicated of an Implicit Event Argument .............4...135 4.............1.1 Prepositions as Aspect Encoders .............................................................72 3.3.....80 3............3 ...............2.4 Summary of Section 3..7..1 Previous Explanations of Nonstative Copular Clauses ....................108 4.................................6 .......................3 Some Confusing Arguments in Defense of a State/Activity Distinction................................2 Summary of Section 4........................................................3 Relational Mental Properties..............83 4...................3...............................81 Chapter 4 Aspectual Alternations in Individual-Level Predicates .....7......................................5 Justifying the Approach .........2 Syntactic Approaches ..............................................5....................98 4......1..................................133 4..........................108 4......................................117 4....6............7...........................................2 Adjectival Predicates Showing an Activity-like Behavior ...........96 4..........2 On the Optionality of the (Affected) Goal PP ..............6...................69 3......... Summary of the Chapter .3 ....................................................138 4.3............2 The Preposition Introducing the “(Affected) Goal”: An Activity Inductor123 4.3 The DP Subject of Relational MPs .............................................84 4..............4....................6 An Account Based on the Relational PP Complement ...................5..............6.........1 Cruel-type Small Clauses Taken by Verbs Other Than the Copula ...................1 Similarities between Activities and States ................................3 The Relational PP with Other APs ...............5 ........1 Aspectual Tests on IL Predicate .4 Summary of Section 4.143 .........................62 3............1 On the Interpretation of the Relational PP ..95 4............................................................3 Summary of Section 4.................................105 4...........8.........................................................1 Lexical Semantic Classes of Adjectives .........4..5......................................................2 Summary of Section 4...............96 4.4 Summary of Section 4...86 4...............5.................117 4.............................1...... The Relational PP Complement.........................106 4.......................... States and Activities: A Grammatically Relevant Difference?............1 The Two-Copulas Hypothesis .................68 3........vi Individuals in Time 3............4........2 ...........2 Be as a Copula Shifting a State into an Activity ..............

........3 Inner and Outer Aspect .........1 When the Subject Is a QDP ...................3 The Arising of Lifetime Effects .......................................................... and Systematicity ...............147 5......................................................1 Tense.................................... 176 5.. 179 5......... 186 5..2....163 5..........5 The Lifetime Reading in Compound Sentences ..............5....1...... 1996) ....................................................3....6.......158 5.................................194 6..... 187 5................................................................1 Temporal Interpretation as a Consequence of Argument Structure......3 Brief Notes about the Complement of Mental Properties APs .......171 5.......................215 6.197 6............................152 5..........222 ............ Musan (1995........................................2 .............Table of Contents vii Chapter 5 Outer Aspect and Individual-Level Predicates .......1.......2 The Perfective and Adjectival IL Predicates ...............................4.........202 6.........................................193 6............................................................ 174 5....148 5.....................157 5...............................2 The Habitual Interpretation: Iteration.................2.........213 6......208 6.............1 Habitual Heterogeneous Predications........................................220 6...............................2 Context Associated to Individuals ..................3..........................................1 The Imperfect and Adjectival IL Predicates ............................................................210 6......................................................................................205 6.............................4...........................5.............................3..........................................4..............................5 Adjectival Individual-Level Predicates and Viewpoint Aspect .. 1995) .... Summary of the Chapter ...............................................3 The Content of the TT in the Arising of Lifetime Readings .........4 Summary of Section 6.......173 5........1 Necessary Conditions for Lifetime Effects .......2.212 6.......... A Brief Summary of Aspect Notions ................... Stowell (1993...........4 The Determination of the TT Content and Lifetime Effects .............3...........4 .....199 6..............2.................................. Kratzer (1988........205 6.....3 Articulating the Account...3......2..........199 6....5....2 Nonpermanent IL Predicates...............................3 The Progressive and Adjectival IL Predicates .............................. 192 Chapter 6 Tense and Individual-Level Predicates ............. 180 5......4 Summary of Section 6..........2...........4..2.............1 Quantifying over Occasions ............148 5........................... 177 5.........1 Tense and Aspect as Ordering Predicates ...................................................................1 Permanent IL Predicates ................2 The Internal Argument of Tense: The External Argument of Aspect..2 Perfective Homogeneous Predications ...............................................................2 Aspect as a Quantifier over Occasions ................................3 On the Relation between the TT and the Habitual Q<occ> ..............2 Differentiating Temporal Extensions of IL Predicates ..................... 1997) ...........4.... Proportion...........218 6..205 6................................................2 Introducing the Determining Role of Contextual Factors... Aspect as an Ordering Predicate ...

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

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

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

whereas verb phrases and prepositional phrases are (almost) always SL. In this study. 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. 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. Adjectival predicates have constituted the central domain of study in the literature. Noun phrases are (almost) always IL. and the status of the subject is confined to that of a participant in the reported event or situation. because they alone among syntactic categories exhibit a robust internal division defined by the SL/IL divide. but Arche systematically examines the evidence that seems to argue in favor of a lexical distinction between IL and SL adjectives. but no clear consensus has emerged on how it should be formalized. and the sentence serves to provide information about its referent(s). the subject of the IL predicate is the topic of the sentence. and shows that it is ultimately . this claim is contradicted by many direct empirical observations. Sentences containing SL predicates. have been described as thetic. 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. whereas SL predicates involve inessential or transient properties. 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. or even immutable properties.Foreword This book provides a compelling account of the distinction between stage-level (SL) and individual-level (IL) predicates. permanent. At first glance. and a third type (possibly the most numerous of the three) can have either type of interpretation. with greater or lesser degrees of naturalness. on the other hand. adjectives are a diverse crowd. with SL interpretations arising in specific cases through the successive application of syntactic and semantic merger. others are (usually) IL. Sentences containing IL predicates have been described as categorical. This intuition is often associated with a related intuition about the discourse function of sentences whose main predicates belong to either type. combining adjectives with various complements and superordinate aspectual and adverbial categories. The SL/IL distinction is grounded in a core semantic intuition that IL predicates involve essential. some are (usually) SL. they serve to report an event or situation. depending on the surrounding context. In contrast. 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.

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

providing further support for her account of the SL/IL distinction in the process. Arche shows how these effects arise and. 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. TIM STOWELL Professor and Chair Linguistics Department University of California. equally importantly.Foreword xiii In certain types of contexts. explains why they often fail to arise. The account of the SL/IL distinction that emerges is striking and compelling in its simplicity. Her account of the interaction between discourse structure and the interpretative properties of quantifiers. Arche brings to light an enormous body of new data. outer aspect. Los Angeles . involving paradigms of a sort that have been largely overlooked in previous accounts. tense. 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. and inner aspect really feels like it is on the right track.

.

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

that evening). 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. such as (5)–(7). which. as I noted earlier. dark-skinned. respectively). (7) is an instance of a nonstative IL predicate. In the cases with estar (4). outer aspect. or he is in a good mood. none of the predicates are understood as “permanent. the speaker predicates the properties of the subject on a particular occasion. I will study the temporal properties of IL predicates in the three commonly acknowledged domains—inner aspect. the property is predicated of the individual as such: the speaker claims that the subject is a handsome.” In the second place. as well as the nonpermanency exhibited by a large number of IL cases. the sentence in (7) appears in the aspectual form of the progressive. will be analyzed in this work. only combines with nonstative predicates. linked to external reasons (maybe because he is wearing a very nice suit. remain unexplained when descriptions of IL predicates relying on stability and stativity are applied to them. and . he got tanned. such alternations are shown in the following examples. funny person. The dynamic properties observed. When ser is involved (3). (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. That is. I concentrate on those copular clauses where the predicate is an adjective. In other words. when he was little.2 Individuals in Time elucidate what lies behind the IL/SL distinction. stativity and permanency are notions belonging to the temporal realm. the adjectival property in all of the examples is restricted to a concrete period of time located in the past (in her youth. In particular. To examine the properties peculiar to ser-clauses. in Spanish. such as those in (5)–(7). Since. I will take into consideration minimal pairs where only the copula differs.

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- . 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. dynamicity is proposed to correlate with the presence of the relational complement and argued to be rooted in the prepositional phrase. Ramchand 2003. I will show that prepositions can contribute an inner aspect shift from states to activities. Chapter 2 introduces the notion of IL predicate and presents the distribution of the two copular verbs in Spanish. Outer aspect and tense are conceived as dyadic ordering predicates that take two time-denoting arguments (Stowell 1993. 1996). which is described as a viewpoint referring to a plural number of occasions. the dynamic properties of clauses where adjectives referring to mental properties appear (such as cruel. This book is organized as follows. the ideas defended in this monograph are sympathetic to the conception of syntactic structure as event structure. along the lines of Demirdache and Uribe-Etxebarría 2000. my proposal will concern the role of prepositions in event structure. 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 argue that the IL/SL contrast should be rooted in the association to a particular circumstance (SL) or its lack thereof (IL). based on hypotheses that attribute aspectual content to prepositions (Hale 1984). giving the number of times a certain eventuality holds (along the lines of Verkuyl 1999). In particular. this being understood in inner aspect terms (Borer 2005. The discussion of inner aspect is framed in the theoretical debate regarding the effects of syntactic structure on the semantics of predicates. I present some reflections about habituality. In this vein. Thus. In chapter 4. In chapter 3. 1995] with respect to the totality of the Eventuality Time. In this respect. Focusing on copular clauses. 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. 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). 2004) and a quantifier over occasions.Presentation of the Study 3 tense. 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. mean) are proposed to correlate with the presence of the relational complement of these adjectives (cruel to Mary). I will then discuss the theoretical and empirical consequences for the IL/ SL dichotomy. 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). Specifically. kind. Ritter & Rosen 2000).

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

*There are several policemen intelligent b. specifically discussing arguments for semantic. lexically distinguish the two types of predication. I present other two aspects basic for my purposes in this work. since I am chiefly concerned with copular cases. There are several policemen available b.3. I will critically revise those analyses. *There are several policemen insane .Chapter 2 Individual-Level Predicates In this chapter I introduce the concept of the individual-level predicate. I will introduce the syntactic analysis that I will assume for copular sentences. and pragmatic distinctions.1 The IL/SL Distinction as a Semantic Distinction. I concentrate on the most influential proposals about the contrast between individual-level (IL) and stagelevel (SL) predicates. NP). Milsark (1974) and Carlson (1977). Since the data in the work are from Spanish. I present the syntactic analysis I assume for copular cases—namely. ser and estar. arguing for an aspectual distinction. In section 2. but those in (2) are excluded.1. I also introduce proposals by Heycock (1994) and Becker (2000). In section 2. PP. 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. who discuss the constituency of the SCs that copular verbs take (in particular. Section 2. syntactic. a part of this chapter presents the differences between the two Spanish copulas (ser and estar). Likewise. according to most authors.2 is devoted to the differences between the two Spanish copulas.4 summarizes the chapter. This chapter is organized as follows. IL and SL. 2. After that. which.1 Proposals about the Individual-Level/Stage-Level Dichotomy 2. I will first present the main descriptions of this notion found in the literature. There are several policemen in the corner a. the predicates of (1) are allowed. the presence of functional projections) and its relationship with the IL/SL distinction. (1) (2) a. For example.1. 1981). 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. Section 2. and introduce some of the concrete aspects for which this work makes alternative proposals. indicating some of the points on which my proposal differs.

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

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

3. (In the default case. with a knife. e) & (in the bathroom.8 Individuals in Time This view has been adopted by some authors. based on Williams’s (1981) Argument Linking description. whereas SL predicates project this extra argument (which can be understood as a spatiotemporal argument).3 In Kratzer’s proposal. PP. can be modified and quantified over by adjuncts. 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. He proposed that action sentences include an event variable as a primitive element in their logical semantics. (∃e) (buttered (Jones. to mention just a few. 1995) and Diesing (1992) but disputed by others. from Davidson 1967. e) & (slowly. This gives rise to a difference in their argument structure which is the stem of the difference between the two types of predicates. e)).2 The eventive variable is considered a primitive element in the logical semantics of events. at midnight. . it is the DP referring to the individual itself which functions as external argument. As will be shown in section 2.2 The IL/SL Dichotomy as a Syntactic Distinction. 1995) In the spirit of 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 if the IP. The Davidsonian argument is the external argument of the predicate. in the bathroom. denoting actions. (i) Jones buttered the toast slowly. which. When it is not projected.1. Kratzer’s proposal looks like (13).4 2 Davidson (1967) took the classical differentiation from Plato between verbs. like nominal variables. Whereas IL predicates are predicated of the individual directly. the event variable is existentially quantified). such as Kratzer (1988. the toast. 3 4 I will add a few more words about the eventive variable in chapter 4. Kratzer (1988. SL predicates are predicated of something else: the eventive variable proposed by Davidson (1967). Kratzer (1988. Schematically. IL predicates do not. as Diesing 1992 proposes. denoting nonactions but performing actions and proposed that events could be treated as individuals. 1995) claims that IL and SL predicates differ with respect to what they are predicated of. 2. NP). and nouns. e) & (with a knife. The trees in (13) just aim to capture Kratzer’s idea on the location of the external argument in the VP predicate. as in (i). such as Jäger (1999) and Becker (2000). e) & (at midnight.

when(ever)-clauses of the sort of (18)–(20) involve an implicit quantificational adverb. This is explained by the possession or lack of the spatiotemporal argument. she looks happy (19) *When(ever) Mary is a mammal. (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. I discuss the semantics of when(ever)-clauses and. 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. rooted in the argument structure. I argue that they are just an apparent test for SL-hood. it feeds its son According to Kratzer. In (18)–(20). where the absence of indefinites and the absence of a spatiotemporal argument preclude the availability of a variable susceptible to binding. potential variables to be bound are those that indefinites introduce (following Kamp 1981. always. it is predicted to be a stable distinction. the sentence results in ungrammaticality.Individual-Level Predicates 9 (13) a. However. SL predicates (at your disposal) can.5 Let me now introduce another important aspect of Kratzer’s proposal. due to the prohibition against vacuous quantification (Chomsky 1982). Kratzer is forced to readjust such a 5 In section 2. differing from Kratzer’s suggestion. as presented. Stage-level predicates IP 1 I′ 1 I VP 1 <e> VP b. When there is no variable to be bound. Such an adverb unselectively binds all free variables it has in its scope. This is precisely the case of the IL predicate in (19). she feeds her son (20) When(ever) an animal is a mammal. If the IL and SL distinction is. as the ungrammaticality of (15) and (17) shows. Other interesting contrasts are accounted for by this spatiotemporal variable: (18) When(ever) Mary is in Paris. and the spatiotemporal variable (18). Heim 1982) (an animal in (20)).3. .

e. how is it that it can be modified contextually? In principle. in fact. and it appears if some predicate takes it as an argument. that. as Rosen (1999) observes. nothing can be added in the course of the derivation. every configuration has to be built up by the elements present from the initial lexical selection. Recall examples like (23) or (24). Throughout this work I will argue that the opposition temporariness/permanency is not what draws the distinction between SL and IL predicates. although the eventive argument is treated as an argument of the verb. where to add an argument at Logical Form constitutes a violation of the inclusiveness condition. context dependent and vague. . (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. Kratzer assumes that the opposition IL versus SL semantically corresponds to permanency versus temporariness. it does not play any specified semantic role. This is a problem for her proposal. 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. is understood as “altered” (i. in the framework Kratzer inserts her work (the principles and parameters model of Chomsky 1981). Another loose end in this approach is.10 Individuals in Time conclusion for the following reason. argument structure does not vary depending on the interpretation process. in principle permanent. Chapter 6 deals with them in detail.. All these complications arise from the erroneous parallelism between IL/SL and permanent/temporary. (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. “temporary”). At most. According to this restriction. In the view of cases such as (21) or (22). this is a concomitant differentiation that accompanies a vast number of cases. where the predicate is clearly restricted to a delimited period of time but indicates set membership. If the IL/SL distinction resides in the argument structure. The same problem persists from the perspective of minimalism (Chomsky 1995 and subsequent work). where a property. which is considered an IL business. As I have pointed out. Kratzer concludes that the IL/SL distinction is.

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

those referring to the event they introduce. Clauses involving an IL predicate are about a prominent argument—a “category” in Kuroda’s (1972) terms. I consider that this conclusion makes the classification IL/SL lose its appeal since it does not have any predictive force. 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. are about the individual designated by the subject. in some pragmatic sense. if a generic operator were actually in the constitution of IL predicates.1.” Clauses involving SL predicates . we would expect that an overt generic operator could appear. which I have already argued against in the two previous subsections. IL clauses can be. they define IL predicates as those that. simply. can be understood as stable or transient (for example. if we are referring to a chronic illness or an occasional one). Raposo and Uriagereka (1995). Besides. SL predicates are. This strict correlation leads this author to propose that when a predicate (like sick).4 The IL/SL Contrast as a Categorical/Thetic Distinction. Mary is cultivated” 2. 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. therefore. In turn. Specifically.12 Individuals in Time 1974). it can be classified as belonging to both classes of predicates. called “categorical judgments. (33) *Típicamente/*característicamente/*generalmente Juan es ignorante typically/ characteristically/generally Juan ser-PRES-3SG ignorant “Usually. (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. Chierchia argues that the Gen operator is a strong quantifier. Regarding Chierchia’s central proposal that IL predicates are inherent generics. Juan is ignorant” (34) *Típicamente/*característicamente/*generalmente María es María ser-PRES-3SG typically/ characteristically/ generally muy culta very cultivated “Usually. As the following sentences show. that is not the case—another reason why I will not follow Chierchia’s approach on IL predicates. 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.

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

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

In the past-tense cases. those clauses where the topic is an individual are instances of IL predications. In the next section. I have introduced the idea that equating permanency of a property and ILhood is inaccurate. 2. 9 For further discussion on the overtness of Hebrew copula. I take up this issue also in chapter 7. Matushansky 2000). the copula is null in the present tense but overt in the past tense (Kondrashova 1995. Other languages.Individual-Level Predicates 15 are instances of SL predications.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 an NP → copula “don” c. Bambara is one of the officially declared languages of Mali. and as secondary language it is employed to communicate nationwide. 1996. In turn.9 In Russian. (37) a. . ser and estar. such as Bambara (Koopman 1992). I expand the discussion regarding what seems to be an adequate description for the IL/SL dichotomy. the postcopular NP (the predicate) alternates in case (nominative or instrumental) depending on whether the property is permanent (40) or nonpermanent or noninherent (41). When the predicate is an AP → copula “ka” b. see also Rapoport 1987 and Rothstein 1995. 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. Spanish and Portuguese have two lexically different copular verbs. where the semantics of the two Spanish copulas are introduced. found across most of western Africa. The Hebrew copula alternates between overt or null. A large part of the population uses Bambara as its mother tongue. hinging on whether the predicate expresses an inherent or definitional property of the subject (38) or not (39) (Greenberg 1994).8 distinguish three types of copular verbs depending on the category the predicate belongs to (37).

if not a mistake. and accidental properties as temporary. discussed in section 2. Pedro es/*está Pedro ser/estar-PRES-3SG teacher “Pedro is a teacher” When the predicate is a PP. when the predicate is an NP. 2. insufficient to characterize the copular alternation. as . However. 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.11 which. despite being ungrammatical with ser. 12 Although I will not work out this difference here. Este animal es/*está that animal ser/estar-PRES-3SG a goat “That animal is a goat” professor c. there is no general agreement in the literature on how the copula alternation should be characterized.1 Distribution of Copular Verbs in Spanish Some traditional grammarians (Bello 1847. obviously designates a nontemporary property. un ordenador de 1960 (42) a. which led to the claim that ser corresponds to permanent properties and estar to temporary ones. Ese artefacto es/*está that artifact ser/estar-PRES-3SG a computer from 1960 “That artifact is a computer from 1960” una cabra b. if the subject is a physical entity (43). How can the distribution of the two copular verbs be described? As traditionally observed. the copula used can be estar.1.2. other authors observed that the distinction between permanent and temporary was.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. In the case of Spanish.1.12 10 Note that these descriptions fall in line with the ones given by Milsark (1974) for “properties” and “states.1. the copular verb must be ser (42). The classic counterexample in this respect is estar muerto ‘be dead’.” as cited previously (see section 2. if.1).10 Essential properties are understood as permanent. or ser. if the subject denotes an event (44). Roldán 1974) describe the distribution of the two copulas ser and estar on the basis of the Aristotelian dichotomy “essence” versus “accident. whereas estar is used to refer to those properties considered accidental. 11 See also comments by Milsark (1974) and Carlson (1977) about the same predicate.

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

one of the properties that Schmidt (1972) and Bache (1978) provide to distinguish classifying adjectives from qualifying ones is. and Demonte 1999. Bosque and Picallo 1996. see Bosque 1993. Further discussion is presented in chapter 7 (see section 7. However.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. If the subject is a resultative nominal. also.13. 15 13 . it is not gradable. 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. that classifying adjectives are excluded as copulative predicates (The presidential trip vs. For more details about classifying adjectives. precisely. 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. (b) impossible modification by a degree adverb (48). *The trip was presidential). When the adjective is referred to the place where one is born. and those in (51) combine with both copular verbs. These tests are (a) impossible participation in comparatives (47). 14 Note. and (c) impossible participation in binary correlations (49).6). that their acceptability as predicates in copular constructions seems to be confined to those cases where the subject is a physical entity.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.

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

With the copular distinction in Spanish in mind. where the pound symbol indicates when the paraphrases are not appropriate. see section 2. where a perception verb (note) has a SC (NP + AP) as a complement. all of the inflected versions in (55)–(58) that we can construe as paraphrases have to be done by using estar.1. whenever the copular verb is ser. Consider the next group of examples. not ser— that is. the copula estar will be considered a lexical sign of an SL predicate.1. Thus. Interestingly. (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” . Consider (59)–(62). 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 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 1977.1). the copula designing SL-hood. but he looks very handsome” muy pálido. throughout this work. 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 will consider that. pero está moreno (54) Pablo es Pablo ser-PRES-3SG very light-skinned but estar-PRES-3SG dark-skinned “Pablo is very light-skinned. I am dealing with an IL predicate. Correspondingly. (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. as we already know from (51). pero está muy guapo (53) Pablo no es Pablo not ser-PRES-3SG handsome but estar-PRES-3SG very handsome “Pablo is not handsome.20 Individuals in Time nada gracioso.

which only combine with estar. since they combine with estar. as in (65) and (66). as in (63) and (64). are not so. and would therefore be expected to be grammatical as complements of notar. some adjectives. 16 . 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. since the participial forms corresponding to the verbs they are related to are desnudado and descalzado respectively. desnudo and descalzo16 do not.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. (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.17 Adjectives such as desnudo or descalzo are described as “cut-short adjectives” (Bosque 1990). Chapter 7 contains more discussion on these adjectives. whereas those combining with ser are excluded in such a position.

22 Individuals in Time Other perception verbs. inexpensive} We could. short. which. closed} (vi) ??partially/half/completely {long. non-context-dependent) standard of comparison.. it rejects closed-scale ones. completamente ‘completely’. This and other issues concerning the semantics of the predicates allowed by perception verbs need further examination. correlates with a “relative” (i. The following sentences where they are used in comparatives and are modified by proportional degree modifiers such as parcialmente ‘partially’. correspondingly. verbs such as percibir ‘perceive’ or advertir . alterado ‘agitated’ sounds quite natural when combined with a proportional degree modifier. For example.. adjectives compatible with proportional degree modifiers are those gradable predicates with a closed structure scale (with minimal/maximal elements). Notar seems sensitive to other properties of the predicates appearing in the small clause selected by it. then. which. 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). whereas it allows open-scale adjectives.) notes.e. full. correlates with an “absolute” (i. gradable adjectives incompatible with such modifiers involve an open scale structure (without minimal/maximal elements). However.e. it is not the case that any perception verb allows for any type of adjectives as its complement. interesting.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. In turn.c. (vii) and (viii)). open. this does not mean that predicates such as desnudo ‘naked’ or descalzo ‘barefoot’ are not gradable. (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.18 such as ver ‘see’ seem to discriminate IL versus SL at first sight. they argue. 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. context dependent) standard of comparison: (v) partially/half/completely {empty. 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. although cansado ‘tired’ and preocupado ‘worried’ seem to behave as proper closed-scale adjectives (cf.

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. and so on that contribute to the acceptability of the sentence. or both types of perception) and the role of modifiers like muy ‘very’. Other adjectives. it is interesting to note the role that overt subjects (pronouns) play in the meaning of this perception verb in Spanish. bastante ‘quite’. emphasize (Quiero que vengas tú ‘I want you [and not other] to come’). Roughly speaking. the verb see distinguishes between states and events. 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. or gracioso. they appear to contrast (Tu hermano puede venir al cine. 50 above) and. as in (ii).Individual-Level Predicates 23 such as (70) and (71) are those that combine with estar (cf. and disambiguate between verbal forms (cantaba ‘sing-IMPF’ can refer to the first and the third person). the semantics of perception verbs deserves a deeper investigation than what I can offer here. Consider the following contrast. 19 With infinitive complements. rather than ser (72). 20 . subject pronouns need not be overt in Spanish. where a state such as know languages is excluded. physical. however. the corresponding paraphrases contain estar.20 When they are overt. (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. rather than IL/SL-hood. pálido. when the adjective is combinable with the two copulas. In this regard. moreno. improve the sentence. 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. the sense of the verb is that of sub- ‘notice’ are odd with predicates such as guapo. not all the results are so clear cut. but you stay’). pero tú te quedas ‘Your brother can come to the theater.

22 (74) a. Veo que estás muy guapa I-see that you-estar-PRES-3SG very pretty “I see that you look very pretty” b. whereas without the overt subject (73a) the sentence should be paraphrased by using the copula estar (74a). (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. with an overt subject the sentence is paraphrased by using ser (74b). Consider the following contrast: (73) a. 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. Whereas. Te veo muy guapa (hoy. you look very good in those pants) ‘I see you very pretty (today. 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. those adjectives that go with ser are allowed (74c) . the past tense (ii) favors an SL interpretation. 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). for example. . when subjects are overt. te sientan muy bien esos pantalones) you-CL I-see very pretty (today. with estar.24 Individuals in Time jective perception and the predicate of the SC is understood as IL. 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.21 Consistent with this fact. other factors can disambiguate between the two readings of ver ‘see’. you look very good in those pants)’ b. maybe something along the lines of Mario te vio muy guapa en la fiesta ‘Mario saw you very pretty at the party’ meaning (iii).

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

PPs (when the subject is a mobile object) Spanish NPs. Since estar refers to a result state. However. happy).1. It is not a state. ser has no inherent temporal structure. PPs (when the subject is an event (party)) APs (tired. 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. PPs (when the subject is a mobile and immobile object) Estar + Table 2. (More on this in chapter 7. round.2). . her reasoning is as follows.2 Schmitt (1992).) 2. its wider flexibility in combining with adjectives indicates that it does not involve internal temporal characterization. round. swim). I take it that her conclusions about the topic can be applied to Spanish. scenarios incompatible with a result state are excluded for estar. old). and processes or activities are those events with duration but without an inherent culminating point (walk. 24 All these concepts will be adequately presented in chapter 3 (see section 3. the description of SL predicates as perfective (in the sense of “delimited”) does not capture the intuition about the whole class of SL predicates.2. (Schmitt works on Portuguese.) Copula Ser + Portuguese NPs. for this author. States are eventualities with no duration or endpoint (be tall). since the distribution of Portuguese copulas nearly parallels those of Spanish [see Table 2. Estar corresponds to a result state of an accomplishment verb. PPs (when the subject is an event (party) or an immobile object) APs (tired. Predicates like be in the garden are not covered by Luján’s generalization. happy). Accomplishments are those predicates that refer to a process with a inherent endpoint (write a letter or build a house). its underspecification in aspect. Among the facts that Schmitt cites to demonstrate the internal structure of estar and the underspecification of ser are those enumerated in (a)–(d).24 In a nutshell. since there is no reason to consider in the garden as a delimited predicate. In the spirit of Luján. female. whereas ser is underspecified with respect to aspect. That is. APs (human. In contrast.26 Individuals in Time (76) Imperfective predicate A(x) at times tj… tj+k However. Schmitt (1992) argues that the distinction between ser and estar is aspectual in nature. nor is it an event or a process.2. ser manifests a wider flexibility. old). female. Portuguese and Spanish copular verbs The concrete claim of Schmitt (1992) is that estar involves aspectual properties. APs (human. which suggests.1]. but.1.

With the progressive. according to Schmitt. since it would be the aspectual head that licensed the aspectual form. In this respect. 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. which looks like a circular explanation. However. Schmitt continues. In sentences like John built a house. nice) can appear in the progressive. (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. kind. then. 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. there is no result reading available. a house has been built but in John was building a house. it seems that the incorporation of ser into the aspectual head (progressive) is what licenses the progressive. I will make two brief remarks. if the reason for the grammaticality of (80) is the incorporation of ser into the . Since the progressive favors subjects that have some control over the predicate. mean. the house has not been built yet. Schmitt (1992) argues that ser incorporates into an aspectual head (the progressive in (80)) and becomes a predicate with internal temporal structure. (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). estar predicates cannot.Individual-Level Predicates 27 (a) Possibility of appearing in pseudo-clefts and equative constructions Only ser can appear in pseudo-clefts and equative constructions. Second. where the copula does not play any role in terms of selection and does not add any semantic aspectual content to the construction. ser + some APs (such as cruel.

1995). the reason why estar and the perfect are excluded as complements of perception verbs may be because both behave as states. it is not clear that tests along the lines of (82) and (83) show that estar designates a result state. (81)) remains unaccounted for.1. such as when(ever)-clauses: . 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. they may just show that estar predicates behave as states.2. In other words.28 Individuals in Time aspectual head.) (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. 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.” (I will introduce these notions properly in chapter 3). Schmitt also notes that estar. is not allowed as a perception verb complement. like all verbs in the perfect. and (infinitive) states are not accepted as perception verb complements. (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. (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. (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. However. as I already suggested. (See section 2. the fact (also noted by Schmitt) that not all adjectival predicates combine with the progressive (cf.

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

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). Durative aspect would be opposed to “result state” in her proposal. Chierchia (1992). and Jäger (1999) argue that whenever is correct with predicates that are “naturally iterable. mean. Consider (89). Although. the presence of Eskimo. for example.26 (For the moment this is sufficient to introduce my point. ser can be ruled out neither in when(ever)-scenarios nor in progressive ones. note that. I argue that if an appropriate context is built up (think. even adjectival predicates of the type of Eskimo fit.) As a last remark regarding when(ever)-clauses. kind. This is what licenses. . of a tale). If it finds a variable susceptible of referential variability. siempre que Juan es esquimal. he bears a life full of difficulties and shortages” Here is a context where the existences of Juan can be iterated. nice) are possible in these contexts. se enfada whenever Juan estar-PRES-3SG dissatisfied with his work gets-angry “Whenever Juan is dissatisfied with his work. this may seem to be the case. according to Schmitt (1992). a permanent predicate all through someone’s life. at first sight. they select for some sort of durative aspect.” Instead. 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.30 Individuals in Time (87) Siempre que Juan es cruel. whenever Juan is Eskimo. expands on this account about iteration. he gets angry” In sum. (89) (En todas sus reencarnaciones). which Schmitt treats in 26 De Hoop and de Swart (1989). which would lead to the conclusion that estar is excluded in such contexts. se arrepiente después whenever Juan ser-PRES-3SG cruel. which is dedicated to Aspect. in this case. it is licensed. he regrets right-after “Whenever Juan is cruel. I have formulated the distribution of whenever (which I consider a temporal distributor quantifier) as a subcase of distributivity in general. Chapter 5. Schmitt argues that ser is allowed in when(ever)-clauses by virtue of its underspecification with respect to aspect. under whenever. contrary to fact: (90) Siempre que Juan está descontento con su trabajo.

2. 2.4 Summary of Section 2. which do involve internal temporal structure (estar in her own proposal and many others). which. I consider any appearance of ser to denote an instance of an IL predicate. are also perfectly grammatical in such scenarios. Ser predicates are considered parallel to write or admire.” As chapter 7 discusses in greater detail. and estar as a predicate that is true of an individual for a delimited period of time. Likewise. Luján conceives of ser as a predicate that applies to an individual during a stretch of time with no beginning or end assumed. ser lacks any inherent temporality and is described as “inert with respect to aspect. the opposition. where no . Another author arguing in similar terms as Luján (1981) and Schmitt (1992) is Fernández Leborans (1999).” As I understand her proposal.2.” she means that serpredicates are not tied to temporal limits—the reason why she argues such predicates can be described as “durative.2. I argue that the oddity (or not) of such aspectual forms is not related to the IL/SL distinction.2. to say that ser is allowed because it does not involve internal temporal structure is. Fernández Leborans considers that in contexts such as the progressive (91) or temporal forms suggesting that the property does not hold any longer (92). in her view. Luján (1981) and Schmitt (1992) defend the idea that the distinction between ser and estar should be understood in aspectual terms. In sum.Individual-Level Predicates 31 parallel to when(ever)-cases. (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. For this reason. she argues that the ser/estar distinction can be described as an IL/SL dichotomy. ser predicates work as SL predicates. Although Luján refers to this differentiation as an imperfective/perfective contrast. which are conveyed by a type of adjectives in a specific way. since other predicates. at best.2. I argue that the availability of the progressive is related to the internal temporal properties of the construction. alludes to the lexical content of the predicates. as she explicitly points out (Luján 1981:206).2. 2. Whereas estar possesses internal temporal structure. just a partial answer. by “inert with respect to aspect. I will not have to propose that when the copula ser combines with them it becomes an SL predicate. In her study on copulative clauses of Spanish. I show that the perfective preterit can be licensed under the appropriate contextual conditions. However.3 Fernández Leborans (1999). is founded on aspectual properties.

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

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

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

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

in the presence of a functional projection (Aspect) and an Event Phrase (in the spirit of Kratzer 1988.36 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. The evidence adduced to by Felser for the presence of an eventive argument is that perception verbs complements have to be “eventive.28 (107) a.1 for a discussion in which opposition (eventive/stative or IL/SL) is at stake. 28 See section 2. besides hosting the event argument. can host an argument (a subject) in its specifier. as in (107). Felser argues for the existence of an eventive argument and for the existence of a functional projection that. *I saw John know the answer As for the existence of an extra projection. 1995) or its lack thereof.” rather than stative. Felser presents evidence of a filled specifier position. . I saw John draw a circle b. 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.

She just needs to assume one copular verb. it cannot be said to have been generated inside the VP but inserted at a higher level within the perception-verb complement. the difference between SL and IL copular constructions resides in the complement of the copula. for Becker. Thus. Furthermore. Thus.1. I argue that IL-hood 29 See also Guéron and Hoekstra 1995. As can be appreciated from (105). In particular. argued for the existence of two homophonous verbs be. as pointed out in section 2. Felser proposes the extra layer to be an Aspect projection. authors such as Schmitt (1992). However. see (105). Becker’s (2000) idea differs with Kratzer’s (1988. in particular to copular sentences. Because there is not a thematic subject. since both establish a relationship between copular distinction and aspectual content. unlike Schmitt.1. perception verbs take Aspect Phrases. Likewise. Whereas SL predicates are linked to aspectual content. differing in this respect from Carlson (1977). which are presented as codependent. Summarizing. other authors have argued for the existence of functional projections inside them too. . the contribution of this book will be twofold. but the projection of two nodes. Chomsky 1986. assuming that to occurs in the T head (Emonds 1976. who. Becker considers that an aspectual distinction does not exclude in and of itself a split conceived as IL versus SL. In the first place. the exclusion of a to-infinitive. The reasons to discard a CP or TP projections come from the ungrammaticality of sentences like (109) and (110). Event Phrase and Aspect Phrase. such an aspectual content is absent in IL predicates. 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. suggests that the perception-verb complement is not a TP either. Becker establishes a correspondence between the presence of Aspect and SL-hood. The proposal defended by Becker (2000) shares the spirit of Schmitt’s (1992). whereas Stowell (1981) originally conceived SCs as lexical projections. Felser concludes. (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.Individual-Level Predicates 37 (108) I wouldn’t like to see [there be so many mistakes] Her argumentation goes as follows. Regarding the internal constitution of the SCs taken by the copular verb. 1995) in that SL predicates do not involve the projection of just an event argument (here an Event Phrase). Heycock (1994). Pollock 1989). and Becker (2000)29 have put forth the presence of a functional node (Aspect).

3 introduced the syntactic analysis of copular constructions I will assume. among others). etc. I start by dealing with their internal temporal characteristics (discussing whether they are stative. I therefore discarded those proposals establishing a strict correlation between IL-hood and the permanency of the property in the subject (differing. etc. and Fernández Leborans (1999). Centered on ser + AP cases. the difference between IL and SL copular clauses.) in chapter 5 and finish by examining their tense properties (i. I will not argue for a plain functional head but.. NP). Therefore. PP. section 2. Because my primary concern is copular IL predicates. I argued that such a distinction is operative in the grammar of natural languages since it captures important contrasts. In this work I defend a finer grained distinction among IL constructions with ser. and the predicate (AP. I will develop a hypothesis about Aspect in predicates combining with ser and. dynamic. which contributes activity properties) comes from a lexical item—namely.) in the next two chapters. based on the different aspectual properties they show. In the second place. I defend IL SCs with aspectual content. as I will argue. As a result. Following Stowell (1978. then. I do not conceive of the IL/SL contrast as a permanent/temporary contrast. based on Hale 1984 and Demirdache and Uribe-Etxebarría 2000. In essence. Chapter 7 takes up the debate about the characterization of the SL/IL distinction. as the minimal pairs of “copula + AP” in Spanish suggest. 1995. I consider that the copular verb takes an SC containing the nominal argument. I continue with an analysis of their outer aspectual properties (their combination with imperfect.38 Individuals in Time and SCs with aspectual content are not incompatible. for a lexical item whose semantic import has a correlation in the aspectual realm. perfective. aspectual content does not amount to SL-hood.. that is. 2. temporal interpretation) in chapter 6.4 Summary of the Chapter This chapter introduced the dichotomy between IL and SL predicates. I do not submit the idea that only estar-predicates involve aspectual content. I develop this hypothesis in chapter 4. 1981). 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. unlike Luján (1981). . such as the distribution of copular verbs in Spanish. whereas SL predicates refer to properties that are presented linked to a situation.e. dynamicity.e. In particular. a preposition. much in the line of Milsark (1974) and Carlson (1977). subject of the predication. I propose that the aspectual import I will be dealing with (i. taking into account the analysis of the temporal domains presented. strictly speaking. Schmitt (1992). I consider IL predicates to be classificatory predicates that apply directly to the individual. and Chierchia 1995. from Kratzer 1988. I will not consider that Aspect is.

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

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

the event of walking ends.g. a point that authors such as Verkuyl (1972. the locative phrase delimits the process.” and “accomplishments. (5) is a resultative construction: Pablo walked until he got dizzy.” It is worthwhile to add something here about the difference between activities and accomplishments. nothing in the predicate (in the sentence) forces an endpoint. In (3). These authors got deeper in the characterization of event types and provided linguistic tests to diagnose the membership of a predicate to a class. When the state of dizziness is reached. The predicates that fill the template described for (3)–(5) are called “accomplishments. Vendler’s work. That is.” “activities. Smith (1991) added an additional class: “semelfactives. The action is delimited and its duration gets also delimited as a consequence. where the verb appears with an internal argument (a cognate object). 1989.” This division has been kept in its basic terms by most of the subsequent literature. and Pustejovsky (1988. the event of “walking” denotes a process that is not internally delimited. Finally. where the adjunct around the park refers to the space where such an event took place.Event Classes and Individual-Level Predicates 41 and those that are ongoing without any definite terminus. Kenny (1963). Predicates that behave this way are known as “activities. named “energeiaverbs” (e. It is the same with (4). This taxonomy of predicates spread in the Anglo-Saxon tradition mainly through the work of Ryle (1949).” Vendler considered predicate classes as a strictly lexical matter of verbs. It has been observed that whereas a sentence . This behavior holds in (2) as well. Vendler divided eventualities into “states. 1993). and Vendler (1967). Tenny (1987. 1994). walk). 1991) proved wrong.” “achievements. together with Dowty’s (1979) set of diagnostics. can be considered the most influential work on classification system.” Sentence (3) does not behave like this. which also bounds the process. Although we lean toward thinking that the subject will not be walking forever. To make the characterization of each type clear. 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. The park can be conceived as the goal of the walking process undertaken by the subject of the sentence..

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

Identifying which type an eventuality behaves like may be helpful just to pinpoint the determining inner features of a predicate. 1991). (3)). like agency. Others make visible the features that two or more event types have in common. Finally. I will follow this second option throughout this work. Tenny (1987. Put in other words. for example.1 Events versus States. That is. as Verkuyl (1993) and Borer (2005). Juan está trazando un círculo Juan is drawing a circle . 1994). (2) vs.2. 1989.1. Vendler (1967). “event templates” do not have a grammatical reality. The test in (9). 1970) as a proof distinguishing statives versus nonstatives. put it. and Ritter and Rosen (1996). Juan está paseando Juan is walking d. proposes. Pustejovsky (1988. I start out by discussing some tests that distinguish statives versus nonstatives in (9)–(11). I will be adding some discussion about them.) 3. Some distinguish between states versus nonstates. 1993).Event Classes and Individual-Level Predicates 43 affected by other elements present in the sentence (cf. has been taken (since Lakoff 1966. as Pustejovsky (1988). rather than a lexical one. Each test establishes a distinction of a different kind. 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 then proceed with the tests that differentiate among different types of events. I present a summary of the tests to diagnose event-type membership from Kenny (1963). (The judgments I give are based on native intuitions from Spanish. ??Juan está dándose cuenta de que su madre tiene razón Juan is realizing that his mother is right c. The other (more radical) option is to think that there are no such event type primitives. Ryle (1949). Verkuyl (1972. the possible or impossible occurrence in the progressive form. Next. among others. among others. 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. made clear that internal arguments and certain types of locative phrases affect the category a predicate can be classified into before entering into syntax. (9) Occurrence in the progressive form a.. As I present them. others test out not aspectual properties strictly speaking. but properties that seem to depend on aspectual ones. and Dowty (1979). aspect is a crucially compositional issue. Only the latter can appear in such a form. *Juan está siendo alto Juan is being tall b.

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

which divide activities and states.4 Whereas (11a).” as their reaction against combination with habitual adverbs proves. on the other. The eventualities whose interpretation in the present is not habitual are states. . However.2).Event Classes and Individual-Level Predicates 45 habitual (Comrie 1976. whereas (SL) statives can be said to be ambiguous. it might seem that this test is a diagnostic for IL-hood rather than for stativity.1. and achievements and accomplishments. sentences like Juan is usually sick or Juan is usually at his best friend’s place. Juan traza un círculo Usually. I argue that the crucial condition for a predicate to appear in habitual form is its possibility of restarting. involving an IL stative predicate. and those that do not. (11) Interpretation as habitual in present tense a. are completely grammatical. Such compatibility cuts the pie of eventualities in two: those that have a habitual interpretation. on the one hand. Juan realizes that his mother is right c. In this subsection I go through a number of tests that separate activities and states from achievements and accomplishments. Juan draws a circle Nevertheless. according to the temporal adverbials that refer to the time an event lasts they admit. Normalmente. Juan pasea Usually. A durative adverbial. Normalmente. I show that this is a possibility for every type of predicate (including IL predicates) if a suitable context is created. 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. I start out with the tests in (13) and (14). nonstatives are interpreted as habitual. States are argued to be the only type that cannot be interpreted as “habitual. Normalmente.2. 3. whereas Juan walks does. with stative though SL predicates. is excluded in combination with a habitual adverb (which proves it cannot be interpreted as habitual). Juan knows mathematics b. Juan walks d. *Normalmente. Juan se da cuenta de que su madre tiene razón Usually. Bertinetto 1986. Cinque 1999). 4 I thank Tim Stowell for interesting discussion on this topic.3.2 Activities and States versus Achievements and Accomplishments. it is worthwhile to notice that when no (habitual) adverb is present. 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. Juan sabe matemáticas Usually.

see Piñón 1999. For a recent account about the different readings available in such sentences. Pablo estuvo enfermo durante tres semanas Pablo was ill for three weeks b. in-adverbials refer to the period of time it takes to complete a certain event.46 Individuals in Time ending point. which is the reason why they are compatible just with eventualities entailing a delimit point. *Pablo viajó en tres semanas Pablo traveled in three weeks c. . As the ungrammaticality of (14) shows. 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. as (13) shows. *Pablo estuvo enfermo en tres semanas Pablo was ill in three weeks b. Pablo viajó durante tres semanas Pablo traveled for three weeks c. *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. 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. Pablo construyó una casa en tres semanas Pablo built a house in three weeks d. *Pablo construyó una casa durante tres semanas 6 Pablo built a house for three weeks d. (13) In + x time a.5 (12) For + x time a. the entailment of a delimit 5 Note that this test does not distinguish between IL and SL predicates. 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.

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

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

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

To the extent that an agent is conceivable only if a proper event (a nonstate) is at hand. an achievement. *¡Estate enfermo! Be sick! b. As can be appreciated. and (21c). an accomplishment. the subject did not get to walk or realize. Pablo casi construyó la casa Pablo almost built the house b.2. can only have the second interpretation. I want to introduce another group detecting the presence of an agentive subject (Ryle 1949. the subject was involved in the process of building up a house but did not finish it. states and achievements. an activity. To conclude this survey of aspectual tests.1. get differentiated from activities and accomplishments. yielding ungrammatical results. Pablo casi se dio cuenta de que su madre tenía razón Pablo almost realized that his mother was right Sentence (21a). ¡Pasea! Take a walk! d. in (22)–(25). the tests in (22)–(25) can be used to diagnose stativity. 1970). (21) Interpretative entailments from the adverb almost a.2 gives a more precise notion of “agent. that is. ??¡Date cuenta de que estás confundido! Realize that you are wrong! . which are grammatical. However. (Section 3. too.5 Agentivity Tests. In the other. ¡Construye la casa! Build the house! c. can have two interpretations. (21b). Lakoff 1966.50 Individuals in Time The possible interpretations of almost put in contrast accomplishments with other event types.”) The roles that DPs play in the event and the analysis of the event itself usually overlap. Pablo casi caminaba con ocho meses Pablo almost walked when he was eight months c.5 in this regard). as van Voorst (1988) and others point out (see also section 3. (22) Occurrence in command imperative form a. 3. In one of them. the subject did not even start out the process of building.

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

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 3.2. 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. 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. and a “%” if the event type shows a positive behavior in the test but just under a certain interpretation. The symbol “/” means that the test does not apply to the event type for some reason.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). (20). a “–” when it cannot.

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

given that they behave differently in certain contexts: (28) a. 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. and John who causes it in (27). *El portazo rompió el cristal deliberadamente The bang broke the glass deliberately b. inanimate DPs are perfectly natural in do-pseudo-clefts. a scenario where past tense was possible might be created. in this respect. A: Wasn’t he being treated with insulin? 10 . However. which suggests that such adverbs prove the presence of animate causers.54 Individuals in Time In sentence (26) it is the bang that causes the breaking of the glass. Observe (i). this is not totally true. which mark volition. the feature that differentiates between bang and Juan seems to be animacy. they can be considered on a par. yielding a general statement interpretation. where the DP cucumber cannot be understood as a volitional agent. but just as a cause. and the acceptability of the do-pseudo cleft construction in (ii).10 Although this is the traditional view. In particular. Interestingly. can appear just with certain causers. 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. Both the bang and Juan are the arguments “responsible” of the event. Juan rompió el cristal deliberadamente Juan broke the glass deliberately (29) a. 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. it is easy to note that they are not completely the same. the pseudo-cleft with “do” is possible only when the cause is animate (29b). tense seems to play a role in this regard. Likewise. (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.

but. Collins 1997. . multiplying elements in the lexicon is an uneconomical move. where agreement markers switch according to the relative animacy of the arguments. 2000. What the insulin did was reestablish sugar levels. 12 It would be crucial to investigate whether there are languages that distinguish overtly between volition or animate causers and nonvolition ones. A possible alternative is the following. The first guess can be that there are actually two types of small v. lo que hizo la insulina fue restablecer los niveles de azúcar. languages exist that show agent/patient splits.Event Classes and Individual-Level Predicates 55 In the framework of the generative Minimalist Program (Chomsky 1995 and subsequent work). and references therein about this issue. 11 See Chomsky 1995. Assuming the generative premises that economy principles are observed by the system of language.11 any uneconomical proposal becomes automatically undesirable. such as Algonquian languages (Berardo 1999) and Southern Tiwa (Rosen 1984). Since volition seems to be tied to properties of the argument in the specifier. I am grateful to Olga Fernández Soriano for discussion on this point. It can very well be the case that it is the whole set “specifier + head” that licenses intentional adverbs (or not). There are languages with animacy splits. rather than the head (small v) by itself. the DP in its specifier gets interpreted as the ‘causer’. 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. One conveying something like “animate cause” and another one involving “inanimate cause.12 (30) vP 3 deliberately vP → volition causer 2 DP[+animate] v (vi) B: Sí. thus far. a possible question might be how volitional cause is distinguished from nonvolitional cause. given that not all causers are understood the same way. As a result. die. I have not found any clear example showing that an inanimate causer is treated morphosyntactically different from an animate one. Now. 2001a. but it could not control the acetone levels. it is usually considered that a functional projection (socalled small v) contributes the content of ‘cause’. pero no pudo controlar los de acetona. such as Lakhota (Mithun 1991). There are many other kinds of splits regarding subjects (Ritter & Rosen 2003). 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.” However. As is known. or be tired).

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

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

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

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

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

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

3.1 regarding the present-tense interpretation as habitual (see also chapter 5 for more important .1.” do not “take place. I will make use of the tests introduced above in the same order. Juan es cruel con Pedro Usually Juan is cruel with Pedro Despite all the observations introduced in section 3. *Juan estaba siendo esquimal Juan was being an Eskimo b. *Normalmente. simply. The eventualities yielding ungrammatical results are taken as stative.” They are considered to be eventualities that lack any bounding and are inherently durative. *Lo que sucedió fue que Juan era rubio What happened was that Juan was blond c. in general. To do that.1 Events versus States. *Juan estaba siendo rubio Juan was being blond c. Juan es rubio Usually Juan is blond c. states are eventualities that do not “happen.3. IL predicates are.62 Individuals in Time 3.2. to identify the aspectual properties of IL predicates. Juan estaba siendo muy cruel con Pedro Juan was being very cruel with Pedro (44) Pseudocleft with happen (‘take place’) a. Normalmente.1. taken to belong to the group of states. Juan es esquimal Usually Juan is Eskimo b. As described above. (43) Occurrence in the progressive form a. *Lo que sucedió fue que Juan fue esquimal What happened was that Juan was an Eskimo b. I will concentrate on copular clauses and test out whether all IL predicates are appropriately grouped in the aspectual set of states. as has been widely agreed on in the literature.1 Aspectual Tests on IL Predicates The aim of this subsection is. *Normalmente.3.” but just “hold. The predicates I will test are a classificatory AP (Eskimo) and two qualifying APs (blond and cruel). ?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. In (43)–(45) I reproduce the tests marking for stativity versus eventivity.

a clear contrast emerges among different types of APs in combination with the copula ser. although it gives ungrammaticality for one of the predicates that above tested out as a state (Eskimo). .2. Since it appears that the AP in (c) behaves like an event. Actually. 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. which is why they are compatible only with eventualities entailing a delimit point.2. 3.3.2 Activities and States versus Achievements and Accomplishments. as shown earlier. (45c) gets a habitual interpretation. when they hold of an individual. It does so as well with IL predicates. as the contrasts between (46c) and (47c) show. we see that one group comprises APs showing a state nature (a. as reasoned in the literature.1. the reasons for its ungrammaticality are independent from the test itself. Rather.Event Classes and Individual-Level Predicates 63 remarks). on the other. states are durative and for-adverbials are durative adverbials. it seems that these tests consistently separate the APs in two groups: Eskimo and blond. rather than a state. If we compare these results with the ones obtained in the previous section. por eso sabe hablar Inuit “What happens is that her cousin is Eskimo. once again.18 states and activities from accomplishments and achievements. hold for that person’s entire lifetime. Even in the absence of the habitual adverbial. the next step is to find out exactly what kind of event it is.b)17 and the other comprises one kind of AP manifesting eventive properties (c). in-adverbials refer to the period of time it takes to complete a certain event. If. Consider a scenario like the following: (i) Lo que sucede/pasa es que su primo es esquimal. predicates like Eskimo refer to properties that. and cruel. for-adverbials express how long someone has been engaged in a particular activity or a state has held of an individual. but the oddness in (46a) calls for further explanation. The test in (46) distinguishes. that the period that a stative IL adjectival predicate such as blond holds of the subject can be restricted with an adverbial (46b). pseudoclefts with happen are acceptable with states when present tense is involved. on the one hand. that is why she can speak Inuit” 18 As mentioned in section 3.1. 17 Much like do-pseudoclefts (see footnote 10). its acceptability in (46b) is expected. As I introduced in chapter 2 (and will discuss further in chapter 6). These three tests show that not all APs in combination with ser behave the same way. which is why the limitation of its duration with a for + x time complement sounds weird. It is interesting to note. In turn.

which argues in favor of considering the predicate (ser cruel) as an activity.1. *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). Juan ha sido cruel con Pedro Juan has been cruel to Pedro 3. tests like (49) divide activities and accomplishments from states and achievements.3.1. only activities.2. However. Juan está siendo cruel con Pedro Juan is being cruel to Pedro b. *Juan fue rubio en dos horas Juan was blond in two hours c. and not accomplishments.3.1. Once again. the combination of ser and an AP such as cruel ‘cruel’ looks good. a predicate like ser + AP would be unexpected as a complement of parar de ‘stop’.2.3 Activities and Accomplishments versus States and Achievements.64 Individuals in Time (46) For + x time a. 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. In principle. *Juan fue esquimal en dos horas Juan was an Eskimo in two hours b. where there is no endpoint. (48b) is a licit entailment from (48a). As mentioned in section 3. such a predicate shows a nonstate patterning. . yield a perfect entailment from the progressive form.2. *Juan fue esquimal durante tres semanas Juan was an Eskimo for three weeks b. since this is only possible with eventive predicates. Juan fue rubio durante tres años Juan was blond for three years c. (48) Perfect entailment from the progressive form a. As shown in section 3.

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

3. confirm be cruel as an activity. ?Juan casi fue cruel con su adversario Juan almost was cruel to his opponent b. *Juan casi fue esquimal Juan almost was an Eskimo 3. The next set of contrasts aim at testing whether the process that be cruel can be considered is agentive or not. which also test accomplishments versus all the other eventualities. (52) As a complement of finish a.2.4 Accomplishments versus All the Rest.4. the latter is with activities. like (53c).1. ?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. Tests (52) and (53) show some more contrasts consistent with the results from (49). since the unique licit entailment from (53a) is (53b). . as mentioned in section 3.20 Although (53a) does not sound extremely natural. Although be cruel predicates do not give perfect results after finish. Juan finished being an Eskimo b. Juan no empezó a ser cruel con su adversario Juan did not start being cruel to his opponent c. None of the AP predicates with ser. Only ser cruel is possible under a command imperative form (54) or in combination with adverbs like deliberately (55). (53) Interpretative entailments from the adverb almost a. except for cruel. *Juan terminó de ser rubio Juan finished being blond c. which suggests that dynamic properties are somehow involved in these predicates.66 Individuals in Time 3. All tests give the same results.5 Agentivity Tests. 20 Recall that. fit in canonical agentive contexts. *Juan terminó de ser esquimal. 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.3. they clearly contrast with the other two. The former reading is available with accomplishments. whose combination with almost is impossible.1.1. it contrasts with other APs with ser. and (b) that the subject did not even start out the process.

an agentive subject for their infinitival complements. *Juan fue esquimal deliberadamente Juan was an Eskimo deliberately b. *Sé esquimal Be Eskimo! b. Sé cruel con tu adversario Be cruel to your opponent! (55) Combination with volition adverbials (deliberately) a. force (59).Event Classes and Individual-Level Predicates 67 (54) Occurrence in the command imperative form a. As argued in section 3. 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. a sharp contrast with the other APs is observed. *Sé rubio Be blond! c. but. *Juan fue rubio deliberadamente Juan was blond deliberately c. . (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. all of which need. still.2. I take this to show that the subject of this predicate unequivocally involves “volition.” besides being the instigator (controller) of the process. because of their inherent semantic reasons.1. or regret (60). Be cruel may not be absolutely natural in do-cleft constructions (61). be cruel is a grammatical complement for verbs like persuade.

The leaf was falling down from the tree).1. 21 .22 I dedicate detailed attention to these predicates that do not behave as states in chapter 4. Among them.2. only processes (i.e. as shown by a number of tests. as shown in the set of aspectual tests. differing from other authors’ account.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). *Juan forzó a Pedro a ser rubio Juan forced Pedro to be blond c. agency is notable. activities and accomplishments) can be agentive. we have learned that. among state be-predicates. *Lo que Juan hizo fue ser rubio What Juan did was be blond c. where it is attributed to the copular verb be instead. there seems to be a group whose behavior patterns with canonical activities.3 In this section we have learned two things. I will attribute such an eventive-like behavior to the properties of the AP itself. *Juan lamentó haber sido esquimal Juan regretted to have been an Eskimo b.2 Summary of Section 3. First. There seems to be a group of APs that can be taken as Examples (60a.68 Individuals in Time (59) As a complement of persuade or force a. Rather. not all of them can be considered alike. 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. Second.. *Juan lamentó haber sido rubio Juan regretted to have been blond c. 22 Recall that although not all processes are necessary agentive (cf. Juan lamentó haber sido cruel con su adversario Juan regretted to have been cruel to his opponent (61) Pseudocleft with do a. This amounts to claiming that there is a group of IL cases that involve typical eventive properties. ?Lo que Juan hizo fue ser cruel con su adversario What Juan did was be cruel to his opponent 3. *Lo que Juan hizo fue ser esquimal What Juan did was be Eskimo b. *Juan forzó a Pedro a ser esquimal Juan forced Pedro to be Eskimo b. taking into account the properties of this group introduced in section 3.3. not all IL predicates belong to the category of states.

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

e. both activities and states share the so-called subinterval property. as pointed out by Vendler (1967). Mourelatos (1978). (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.M.” That is. it is entailed that John was sick at every subinterval between Monday and Friday. Likewise. As notably put by Bennet and Partee (1972). any part of an activity or a state has the same properties as the whole. to use a more precise term. Pablo construyó una casa en tres semanas Pablo built a house in three weeks d. “homoemerous.70 Individuals in Time (63) In + x time a. and Dowty (1986). and assert that it was true of John from Monday to Friday. including every moment of time I. an activity) from 2 P.. to 3 P. 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. Bennet and Partee (1972). *Pablo estuvo enfermo en tres semanas Pablo was ill in three weeks b. then the sentence is true at every subinterval of I. it is entailed that at every subinterval between 2 P..M. (66) Additivity Property The result of the sum of a number of portions of x is x .M. Contexts like (65) also prove the subinterval property. defined in (66). *Pablo viajó en tres semanas Pablo traveled in three weeks c. (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). John was pushing a cart. Carlson (1981). makes both eventualities homogeneous. toward which to tend. If we take a state. if we assert of John that he pushed a cart (i. or. to 3 P. At each moment of (65) it is correct to say both Pablo is talking and Pablo has talked. be sick.M.

and Bach (1986). There are other state predicates that also allow for such a frequency reading. . The legs of a table are not a table. or it can also mean that Pablo goes walking very often. it is ambiguous: in (68). “water” can be divided into parts. Mourelatos (1978). Observing such properties. (71)). each of which is water. (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. However. Pablo is able to walk 45 miles without stopping).2.Event Classes and Individual-Level Predicates 71 For example. noticed the close parallel between the mass-count distinction in the nominal domain and the aspectual classification of predicates. consider (72). of building a house. or subintervals. 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. Likewise. we always obtain walking or being sick as a result. However.g.. However. Carlson (1981). much is a quantifier proper for mass nouns. a table is not the result of minor portions of tables. 24 I thank Tim Stowell for bringing these cases to my attention. with activities. if we sum subintervals of walking or being sick. which yields bad results with count nouns (cf. it can mean that the amount of walking is big (e. For example. meaning that ‘Pablo drawing a circle takes place very often’. For more detailed discussion of frequency readings. see section 5. among others. the subdivision and the sum of the parts of a count noun and the result are not of the same nature. Quine (1960). building a house is not the result from summing portions.23 In this respect. (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). As is known. and the sum of portions of water is always water. In (67) the adverbial mucho refers to the intensity with which Pablo loves Mary.

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

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

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

however. (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. (xiv) *This table is missing a leg more and more (xv) *This table is getting to miss a leg 26 On my view.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. Martha. however. about examples from English such as those pointed out by van Voorst (1988). 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’ . with an activity such as walk around the park. the modal has two meanings— namely. the modal has just an epistemic reading. the speaker makes a guess about Martha’s state. is usually involved in. The examples below show where stative verbs have to appear in the progressive form in English. Piñón (1995) notes that modal verbs have different interpretations with dynamic and stative predicates. the simple present form being ungrammatical. (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. the speaker expresses a guess about a particular activity she thinks the subject. The same can be said for accomplishments (80) and achievements (81). On the former. However. epistemic and deontic. the different interpretation of predicates with modal verbs diagnoses agency versus lack of agency.26 I have stopped knowing her/getting to know her I have nothing to say. On the latter the speaker expresses a command that Martha has to effectuate. rather than stativity versus dynamicity. the oddness of (xiv)) or an inchoative one (xv). Roughly described.

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

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

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

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

properties like agency cannot be attributed to the lexical entry of the verb but to the whole configuration. based on Hale (1984). (4) (5) John broke the window on purpose The ball broke the window (*on purpose) . as I discussed in chapter 3. the two-copulas hypothesis is unappealing because it multiplies lexical entries with no independent evidence to support it. as shown in chapter 3.84 Individuals in Time presence of a constituent acting as a dynamicity inductor.1. and empirically.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. as has been maintained thus far in the literature. undesirable in itself). In section 4. rather than to the properties of the copular verb itself. semantically null. They proposed a theta-assigning active copula for cases such as (1). different from the “regular copular be”. I will defend the preposition heading the PP complement as an actual aspect head. 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. therefore. Theoretically. there are the following three. 4. 2004). among others. of cases such as (3). Stowell (1993). with a meaning close to act (2). Recall pairs like (4) and (5). First. where the verb is the same but it is observed that agency is a property just of the animate DP subject.1 The Two-Copulas Hypothesis Partee (1977) and Dowty (1979). 4. Specifically. and Demirdache and Uribe-Etxebarría (2000. 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. this is an uneconomical move (and. 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.1 Previous Explanations of Nonstative Copular Clauses In this section I summarize previous proposals concerning the active behavior observed in certain copular clauses.

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

of an individual and an event simultaneously. color. etc. whose behavior I described in chapter 3 as proper to activities in English constructions such as those in (9) and (10). (13) *To invest in the bonus market was wide/green/squared/old (14) John was wide/old to invest in the bonus market Second. Rothstein’s Gricean explanation does not account for the fact that (7) is ambiguous but just. (11) John was cruel (12) To punish the dog was cruel So. no adjective referring to physical properties (dimension. If the adjective cannot be predicated of an individual (15). and therefore creating confusion to the hearer. the adjective must be able to be predicated of individuals. when the adjectives are predicated of just one argument.1. Thus. Finally. or dyadic. this proposal leaves unexplained why it is a concrete type of adjectives which activates the set of dynamic properties. it gets excluded from the dyadic usage (16). at best. to be able to participate in the so-called dyadic constructions. as in (9) and (10). The speaker would be adding something unnecessary. intelligent) can be predicated of an individual. as the account of the two copular entries. as in (11).3 Adjectives Predicated of an Implicit Event Argument. shape. First. 4. kind.) can appear in dyadic sentences. MP adjectives can be monadic. I will mention four points. (15) *John was premeditated (16) *John was premeditated to invest in the bonus market . for the preference of one reading over the other. the adjective must be able to be predicated of events. mean. in cases such as (11) and (12). when the adjectives are predicated of two arguments at the same time (an individual and an event). age. Stowell studies the set of adjectival predicates. (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. and. optionally. of an event (12). Stowell (1991) suggests that the active behavior observed in certain copular clauses is due to an implicit event argument. With respect to the dyadic usage.86 Individuals in Time Quantity and Manner.

” (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.Aspectual Alternations in IL Predicates 87 Third. from a finer grained typology of eventualities. given the fact that only activities and accomplishments refer to “actions. (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. the event. refers to “actions” rather than “facts.” Compare (17) and (18). the infinitival arguments can be neither states nor achievements. 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. only activities and accomplishments fit. The tree in (23) represents the syntax of the adjective in its monadic usage. it is very imprudent of you to run every day/to swim in the ocean In sum. represented by the infinitive clause. This is expected. and the tree in (24) depicts the dyadic one. The structures Stowell (1991) proposes for these potentially dyadic APs are the following. (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. (23) John/To tell the truth is intelligent AP 5 Spec A′ | | | A John intelligent To tell the truth .

whose PRO subject is understood as an agent (recall that the only predicates possible here are. Compare (i) and (ii). (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 . formed by the preposition (a) (‘to’) plus the definite article. which allows predicates to project additional maximal projections to accommodate all of their arguments. Stowell proposes (24). based on Larson’s (1988) proposal of phrase structure. by the same token. in Spanish they are not entirely parallel to those in English. the dyadic interpretation is obtained this way: the adjective is predicated of the action denoting argument and. precisely. whereas the event denoting argument in English is conceived as an infinitive. those that can be agentive. which is the DP (John). Thus. The adjective takes two arguments via a double-shell structure. in Spanish it has to be introduced by the contraction al. the performer of such an action. the subject of the predication is in the specifier of the adjective. As Stowell notes. in Spanish there is no such a restriction. in the dyadic cases the adjective is predicated of the action denoting argument. since. For the cases where the adjective is understood as simultaneously predicated of the individual and the action. In the first place. whereas in English the event denoting argument cannot co-occur with the complement some MP adjectives can have.3 I will not undertake an analysis of these dyadic constructions in this work. The controller of this PRO is the DP subject (John). gets also qualified as intelligent for having performed the action. (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. states and achievements are excluded).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).

(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. I will make two observations.1) regarding the APs that can appear with either copula (cf. but he is not cruel to him usually” Later. 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. to admit there is an eventive argument that is implicit amounts to admitting that (25) is a SL predication. Ser is completely grammatical. the event argument represented by the infinitive is implicit in simpler copular cases such as the ones under investigation here. . 4 For details of Kratzer’s (1988.2. Note that the contrast between ser and estar in this case falls in line with the contrasts studied in chapter 2 (section 2.3). In Spanish. he argues that the action denoting argument is an overt version of the (Davidsonian) eventive argument.4 he takes it that the presence of such an argument is proper of SL predicates. (25) John was cruel to Peter Regarding Stowell’s analysis. I repeat one of the examples here in (29). and between the dyadic one and stage-level (SL) predication. and the contrast between the examples with ser and those with estar (27) can be captured by paraphrases such as the one in (28). see chapter 2 (section 2. in his account. that is not case. 1995).2. (28)). According to Stowell.1. the copular verb should be. as (26) illustrates. First. Specifically. However. 1995) account.Aspectual Alternations in IL Predicates 89 Stowell (1991) establishes a correspondence between the monadic version of these adjectives and individual-level (IL) predication. in chapter 6 (section 6.2). estar. I take up the discussion related to the interpretation of the “infinitival argument” in Spanish and its status (as argument or adjunct). necessarily. this proposal would entail that in all instances of APs followed by a relational PP. Following Kratzer (1988. which is the lexical marking for SL-hood.

4. since they are not the same. 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. I will then offer an explanation regarding the stative reading also available with such predicates (cf. as shown above. My purpose in the remainder of the chapter is twofold. (8) above). As shown in the previous chapter. Thus. to account for nonstative IL copular clauses. I will consider these sentences as instances of IL predication.2 Adjectival Predicates Showing an Activity-like Behavior In chapter 3 (section 3. which is proved by their acceptation of during-adverbials and the rejection of in-adverbials. it cannot be said it is the copula that induces the active features. Second. the event argument Stowell proposes corresponds to an accomplishment (21) or activity (22).90 Individuals in Time guapo. I will first elaborate an explanation as to what specific kind of adjectives are the ones that show an activity-agentive patterning and why. we observed that there were certain copular combinations that behave like states and others that behave like activities. I have argued in the previous section that the activity-like behavior should not be attributed to properties of the copular verb. These two facts lead me to discard this hypothesis based on the existence of implicit event arguments. More specific- .3). 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. my proposal will not be based on properties of the copular verb. as a consequence. 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. Since the activity-like behavior depends on the adjectival predicate combining with the copula. since it is not the copula that varies in the minimal pairs that can be construed (be Eskimo/be cruel) and. 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. but he looks very handsome in that suit” Given the grammaticality of ser.

mean. *Juan estaba siendo alto/bajo Juan was being tall/short b. I will attribute such an active behavior to a particular syntactic structure. 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. blue. d. new. 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. (31) a. g. Progressive Form (32) a. heavy. slow Apt. c. squared Young. cruel. I will not argue that activity-like properties are due to the intrinsic lexical properties of the adjectives. round.2. and as complements of force or regret). b. 4. Concretely. horrible Quick. *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 . cunning. intelligent. wide. stupid. That is. recent Beautiful. short. 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. f. he distinguishes the following classes of semantic concepts. brown. I begin by describing the type of qualifying adjectives giving rise more often to the nonstative patterning. where a set of adjectives is conceptually more natural than others. e. farsighted. capable. Dixon distinguishes different classes of adjectives according to the type of concept expressed. small Light. shrewd. Dimension Physical property Color and Shape Age Evaluative Velocity Human aptitudes and dispositions Tall. kind. dense White.Aspectual Alternations in IL Predicates 91 ally. old.

*Juan era alto/bajo a propósito Juan was tall/short on purpose b. *Juan estaba siendo joven/viejo Juan was being young/old b. *Juan era joven/viejo a propósito Juan was young/old on purpose b. *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.92 Individuals in Time (35) a. *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.

the group of MPs manifests a variety of behavior. Apt. Intelligent. However. this does not make them agentive. there are two groups that pattern with activities: namely.5. .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).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. Furthermore. Cruel. kind Inner aspect patterning States States States States States Agentive activities States Activities Agentive activities Most of them show a uniform patterning with states. cunning c. capable b. More specifically. A subset of them pattern with states and the two other subsets pattern with activities. those APs referring to velocity and those referring to aptitudes and human dispositions or mental properties (MPs).

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

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

In this section.3. and discuss all the consequences derivable from this. I investigate the nature of the PP complement. for example. offend someone else either by saying something unpleasant. set on fire and bother. to name just a few. Likewise. although it seems intuitively clear that the PP specifies the goal addressed by the subject’s action. 1988.96 Individuals in Time correlation with their aspectual characteristics. and such an action is qualified as cruel. focusing on two aspects: its interpretation and its optionality or obligatoriness. even if we finally concluded that the PP is an “affected argument”. Thus. I will show that its syntactic presence correlates with the aspectual patterning of the construction. 1994] and . offend or regale. Incidentally.3 Relational Mental Properties: The Relational PP Complement As just described. 4. whether or not the PP expresses an “affected argument” depends on the concrete type of action undertaken by the subject. it is understood that Juan was the agent of an action that had Pedro as its affected argument. 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. 4. 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). in italics in (58). abuse. can be considered as an “affected goal. harass. One can. or by acting in a certain way.1 On the Interpretation of the Relational PP Stowell (1991) suggests that the DP inside the relational PP. (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. 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. 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”).” This captures his intuition that such a DP refers to the individual that gets affected by the (underlying) action undertaken by the subject DP.

Jackendoff (1996) challenges Tenny’s correlations and points out that it is not always an internal affected argument which delimits the event and. (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. after this brief discussion. I will deal with all this in turns. in what sense can we say that the relational PP is an “affected” goal? Although.Aspectual Alternations in IL Predicates 97 Pustejovsky [1988] mentioned in the previous chapter). as the suitability of in x time adverbials show. however. as the ungrammaticality of (64) shows.” since they undergo a change. there is no delimitation of the event in (63). in effect. Both examples below are from Jackendoff (1996). 1994) describes “affected argument” as the direct internal argument which undergoes some change. and delimits the event. event delimitation and affectedness has been discussed by other authors. there are internal arguments that can be considered as “affected. 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. nor do they delimit the event. it has become clear that the correlation among internal argumentaffectedness-delimitation as conceived by Tenny is too strict. where. 1988. Humiliate has direct internal arguments and. like the city in (60). also. nevertheless. distinct PPs (into the house. over the bridge) delimit the event. it . The first point I will address is the meaning of the notion of “affected argument. 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. Getting back to the relational PP of our adjectival cases.” Tenny (1987. (60) The soldiers destroyed the city That the event in (60) is. but a PP. it is not the case that all affected internal arguments actually delimit the event. In turn.

In this case. for example. I have argued that its status as an “affected argument” depends on the nature of the action actually undertaken by the subject. in the second case it is not so obvious. 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. 4. By ‘optional complement’. very similarly to the destruction of a city examples (60). As to relational PP complements. that is the case in . harass. criticizing him in public. we would not say that Pedro has undergone a change. As seen in (65) humiliate someone is not a delimited event either. if it is not phonetically overt. depending on the action itself. In sum. by hitting him in such an aggressive way that Pedro undergoes a change.3. can humiliate someone else (Pedro). whereas in the first case the DP inside the PP can be considered an affected argument. it is considered implicit. maybe. 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. I am going to treat the PP as a “goal. by ‘obligatory complement’ I refer to a complement that has to be syntactically present. it seems possible for one person to be kind or cruel to another without this actually affecting the other person in any way. Similarly.” which can be considered as “affected” just potentially. whether they undergo any change. As Bosque (1999) points out. the internal argument of humiliate can be considered as a licit affected argument. as an optional one. However. I mean a complement that is not necessarily present. the DP inside the PP in the relational MP cases could be argued to be “affected” or not. As Stowell (1991) also notes. humiliate (like offend. say (Juan). either phonetically overt or not. I will explore whether it should be taken as an obligatory complement or. (65) *Juan humilló a Pedro en una hora Juan humiliated Pedro in an hour As mentioned earlier. on the contrary.9 8 9 I want to thank Tim Stowell for all the conversations around this section. Again. another important property of obligatory complements is that they are headed by a specific preposition. by uttering Juan has been cruel to Pedro we are just describing a scene where Juan has ridiculed Pedro. if one (Juan) humiliates someone else (Pedro) by. As will be specified.2 On the Optionality of the (Affected) Goal PP 8 Now we have established the interpretive status of the relational PP. that is. One.98 Individuals in Time would be debatable whether they can be considered as affected or not. Or.

Aspectual Alternations in IL Predicates 99 I will start by observing the behavior of other APs that also have prepositional complements. Now. . supporter of such and such proposal and compatible with such other computer features. Spanish. one can think that if a person is cruel or kind or mean he has to be cruel. distressed or prone become ungrammatical if the PP complement is not explicitly present. the PPs are massively headed by to. On the one hand. constructions with adjectives such as eager. In other words. kind and mean to someone else. on the other. maybe even like a contradiction. but. Examples (69)–(72) are trickier. although their presence can be phonetically null and its content is recovered from the context. they cannot be taken as ungrammatical as they appear. although with can appear as well. we may judge a sentence like (74) as odd. In English. The complement is interpreted as something specific which is not necessary to be repeated for whatever reason: guilty of such and such crime. 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. For example. it is quite clear that someone has to be guilty and a supporter of something. immune to something and compatible and consistent with something. (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. The following examples are from Bosque (1999). since all of them are introduced by the preposition con. the complements in these cases are obligatory too.

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

One context is after dejar de ‘≈give up’. so that there is no crucial difference between (78) and (79). as already shown. These contrasts suggest that dynamic predicates keep as such even in the contexts they share with states. habituals of non state verbs keep their dynamic and agency properties in the habitual forms. note that. (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. states are excluded from such contexts (v) and (vi). . To begin. In fact. habituals are stative. it is true that states and habituals behave the same way in certain contexts. the opposite assertion does not sound that strange: One could argue that. the interpretation as habitual in present tense is not the unique reading available. which invites us to take the parallelism between habituals and statives with a bit of caution. While we can have habitual predicates (go walking every day) combined with volition adverbials (iii) and as complements of verbs like force (iv). as a consequence. (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. as has been claimed in the literature. whereas (74) above was considered a contradiction. I will keep the idea that the habitual interpretation should not be considered stative in inner aspect terms. which suggest that habituals and statives should not be considered to be exactly the same thing.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. Examples (i) and (ii) illustrate this.

I will conclude that cruel-type APs are not inherently relational. This could in principle lead us to think that. at the same time. (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. Stowell (1991) was right in considering the simple versions of relational MPs as IL and the ones accompanied by the PP as SL. 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.102 Individuals in Time (80) Juan no es cruel. SCs complement of consider can perfectly have 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. In the remainder of the chapter I will leave aside these (somewhat confusing) interpretive contrasts. and rejects SL predicates. The first scenario where cruel cannot appear followed by a complement PP is small clause (SC) complements of epistemic verbs such as consider. 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” . Sentences (83) and (84) are based on examples from Demonte and Masullo (1999). be claiming that the person is not cruel. without any apparent contradiction. I will concentrate on some syntactic scenarios where an overt relational PP cannot appear. However. (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. such as Demonte and Masullo (1999). have pointed out. Some authors have claimed that consider selects for SCs containing IL predicates. as other authors. in fact. From these cases. To show that the PP is optional.

then. Surely. even in the case where it appears alone. . such as activities or accomplishments. when the DP is inanimate the relational PP is excluded. this does not seem to give us the reason of the contrast in (81) and (82). for example. As observed in the examples below. I conclude. which leads to the conclusion that it is not always there. Consider now the following examples from English. where just some infinitival predicates are possible as predicates of consider SCs. 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. one can agree as a matter of opinion about whether someone is cruel to someone else or not. there should not be any difference between cruel by itself and “cruel + PP”. Consider the contrasts between (87) and (88). since. (87) *Considero a Pedro padre I consider Pedro a father (88) Considero a Pedro un buen padre I consider Pedro a good father However. (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. if cruel-type APs always involved a PP complement. In sum.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. the predicate should refer to something which can be a matter of subjective opinion. The contrast found in (81) and (82) suggests that the presence of the PP makes a difference. In the first place. a PP involving an arbitrary nominal could be syntactically noticed. 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.

such as the imperative form. En Canadá el frío es muy cruel In Canada the cold is really cruel *En Canadá. when the DP subject is inanimate. the construction cannot appear in agency scenarios either. el frío es muy cruel con los habitantes In Canada. (95) 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.104 Individuals in Time (93) a. and. (94) a. they show a correlation between the relational PP and the properties that the subject has to possess. When the subject is inanimate. b. ¡sé cruel! “Image. such as the progressive or as a complement of parar de ‘stop’. as a complement of verbs like force or in combination with volition adverbials: (98) Occurrence in command Imperative *Imagen. 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]. 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 construction cannot appear in contexts proper of dynamic eventualities. b. Examples (93)– (95) demonstrate that the possibility of the PP to appear depends on the nature (animate/inanimate) of the DP subject. more interestingly.

This is because they describe a property that can be understood in relation to another animate entity.3. dynamicity. not included in the group of relational MPs in the table above which. as mentioned before. Consider (101) and (102). which are not totally excluded. as Violeta Demonte makes me notice. (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. If the PP complement is not overt. (103) Juan es inteligente Juan is intelligent .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. while the subject of cruel is not an agent. but just a “theme. since. I will consider that the presence of the PP correlates with a subject involving particular properties and. otherwise. In the next sections. since it needs to bear the required properties for that (and animacy is the most basic one). conclude that the structure of cruel-type APs does not always involve a PP complement and aspectually behave as states in principle. I therefore. can also take a relational PP complement. as it is possible with the other adjectives (cruel. it is not inferred from the context or interpreted as generic. kind. also. I consider these cases with inanimate subjects as evidence that the PP complement cannot be an obligatory complement of cruel.” if we use traditional vocabulary. (104) is not the interpretation of (103). its impossibility to appear with certain subjects would remain unaccounted for. they do not have the same relationship with the PP. namely. which enables agency. In the first place. with particular characteristics of the construction. 4. I account for the relation between the two inner aspect forms.3 The Relational PP with Other APs There are other MPs. Summarizing. etc.). although these adjectives referring to human (or animate) aptitudes can appear in combination with a relational PP.

with other adjectives. they can be said to gain agentive properties. the subject is understood as an agent. I defended that the DP inside these PPs should be analyzed as a “goal. interestingly. either explicitly or covertly. which strengths the relationship proposed between the PP and agency. I studied whether the relational PP is always present.3 In this section I have been concerned with different properties of relational MPs. Jackendoff 1996). Finally. In this respect.” although I discussed the status as “affected” that Stowell also attributes to them. Agentive properties are in direct correlation with the presence of the PP: for the PP to be grammatical.3. Compare the following sentences. With some of them. although this complement does not maintain the same relationship with all adjectives. In this regard. note that.4 Summary of Section 4. whereas. Interestingly. I have debated the correlations between affectedness (understood as ‘change of state’) and event delimitation (Tenny 1987.106 Individuals in Time (104) #Juan es inteligente con la gente en general “Juan is intelligent to people in general’ As a second remark. In sum. 4. aspectual and thematic properties of the constructions are proved to depend on the presence of such a PP. when the PP is added to these adjectives. (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. Whereas without the PP. when the PP is added. 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. I do not believe we can talk of a real affected DP. with the PP present they become acceptable. The following sections elaborate on this point. the subject must be able to involve agentive properties. volitional adverbials are excluded. I considered two facts. it can be understood or inferred from the context if phonetically null. I first showed a syntactic scenario (relational MPs as . I first discussed the interpretation of the PP complements. 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. 1989. In copular constructions of the type of be cruel to someone. Examples (105) and (106) show that. it must be overt. in the be cruel to someone constructions. Second. Along the same line as Stowell (1991).

I showed that these nonagentive APs gain agentive properties as they combine with the relational PP. the matter is reducible to an aspectual alternation affecting the thematic properties of the arguments. activity properties are due to the presence of the PP. I will propose that we can have cruel. Otherwise. Whereas a [+animate] DP subject is compatible with the presence of a PP complement and it is interpreted as an agent. the cited active properties emerge. the relation between the AP and the PP is different. such cases would be unexpected. More precisely. I briefly discussed the combination of the relational PP with adjectives other than the ones considered “relational” by Stowell (1991). the thematic interpretation of the DP subject in each case (stative or dynamic) was different. I showed that.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. the issue is not very different from other contrasts noticed in the literature and men- . and “cruel + PP”. In other words. behaving as a state. which strongly suggests that their properties are different. 4. I then looked at cases where the subject DPs are [–animate]. although I also showed that if we create a suitable scenario and add a relational PP to other adjectives. a [–animate] DP subject is not compatible with the PP and it cannot be understood as an agent. on the one hand. 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”. I took this as additional proof that these adjectives are not inherently relational. whereas predicates describing properties that can be projected onto an animate entity can appear with a relational PP. which correlates with the syntactic presence of an argument (a PP). I argued that this aspectual alternation does not emerge with every type of copular predicate. It can be inferred if it is not present with the cruel-type. Finally. 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. whereas it must be overt to be understood with APs such as intelligent. Concretely.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 suggested that the APs that take a relational PP are those referring to MPs. In the copular cases in question. 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. I also pointed out that. and the other dynamic. In particular. behaving as an activity. and the type of DPs allowed was correspondingly distinct. but the activity behavior ensues with those APs admitting a PP relational complement. interestingly. all of which are odd with a relational PP. Put in these terms.

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

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

and van Voorst (1988). 1994) and van Voorst (1988) elaborate both on the observation that arguments grammaticize the points that give the temporal contour of an event. 1994). 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. Tenny (1987. As van Voorst puts it. as captured in the Event Correspondence Rule (116). each entity corresponds to an aspectual notion.2.1 Event Roles.110 Individuals in Time This approach leaves several questions unanswered. and the presence or absence of other elements in the sentence (the PP complement in our cases) that. among others. showed the relevance of arguments in the inner aspect properties of a sentence. both would contain the term “hold.” given the fact that both are atelic and do not “culminate. 4.2 Syntactic Approaches 4. there is no obvious way to distinguish between the two constructions in point (stative and dynamic copular clauses). (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 . this approach does not establish any relationship between the aspectual term (Hold). 1989.” Second. 1989. in principle. Dowty (1979). As discussed earlier (see section 3. In particular. its syntactic position and its role in the determination of the aspectual properties of the sentence. these authors emphasized the role of the internal object as an event delimiter. authors such as Verkuyl (1972). since.1). correlate with the aspectual outfits of the construction. 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.5. arguably.1. which opens the possibility of establishing a direct relationship between the thematic properties of the arguments present in a sentence. the dynamic or stative properties.5. First. Tenny (1987.

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

2003).5. but not so their syntactic positions. However. Ritter and Rosen (1996. Compare (120) and (121) with (122) and (123). therefore. 2005). the work of authors such as Borer (1994. 2000. However. As Rosen (1999) points out. In a sense. 4. In other words. as such. The old lady instigates the selling in (123). 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. different lexical items have different ways to project. the systematic correlation between object and delimitation and subject and origin). occupies the subject position. correspondingly. and Sanz (2000) elaborates on the idea that argument structure is licensed by functional syntactic structure. 1998. they cannot capture aspectual alternations such as the one we are interested here in a natural way. he does nothing that can be called selling in (123). and. the event-role perspective predicts the mapping of different arguments to particular syntactic positions more accurately than the UTAH.2. Benua and Borer (1996). but the position of event-roles seems to be so. regarding the stative/dynamic contrast in IL copular clauses. Furthermore.112 Individuals in Time In these two sentences the theta-roles are the same. Note further that the NPs play different event roles in each one of them. since. Although these event-role approaches provide accurate insights in important respects (the fact that arguments can affect the event type. 1998. The position of thematic roles is not restricted. Incorporating the insights from the lexical argument-oriented approaches just reviewed. and. Travis (1994. this assumption makes them similar to lexical approaches. For our present concerns. In sum. 2000). 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. . appears in the subject position. (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). a natural account of such compositionality cannot be derived from their lexicalbased approaches. these approaches would have to assume that each aspectual behavior corresponds to different adjectival projections. which I have argued are in strict correspondence. The car is the theme in both cases. she does nothing that can be called buying in (122). for them. although both Tenny and van Voorst explicitly acknowledge the compositionality of events.2 Syntactic Structure as Event Structure.

that gives the type of aspectual interpretation. Borer (2005) elaborates on the intimate relationship between quantification and telicity or lack thereof. In other words. 17 This was introduced in chapter 3 (section 3. events are not lexically specified for any type of aktionsart.4). that is. I will refer to her book (Borer 2005). since it presents the latest versions of her ideas and proposals. whereas a DP headed by a numeral. Borer (1994. Mostly along the lines of Verkuyl (1972. Compare: (124) John drank beer (125) John drank two beers A bare mass noun gives an atelic interpretation. 1989. In turn. aspectual properties can be recast into quantity properties. see Kratzer 2004 and Kiparsky 1998. which I take as a frame of reference for the remaining discussion. all the relevant grammatical processes get a unified account.Aspectual Alternations in IL Predicates 113 and functional structure is interpreted as event structure. the quantification properties of the DP direct object make a direct difference regarding telicity.17 Borer proposes that the core feature of aspectual distinctions is [±quantity]. When considering that functional projections are in charge of event-roles too. an object that can be marked as [–quantity] correlates with atelicity. This head has the property of “quantity” (AspQMAX or Quantity). with telicity. 2000) argue that functional projections responsible for Case processes (Agreement projections) assign the event-roles of delimitation and initiation. 2001. Borer proposes that the core functional projection of the clause is Aspect. events are not states or accomplishments from the lexicon. 15 For further discussion about the relationship between case and event structrure. 16 The work by Borer has appeared progressively in a number of articles.14. a telic one. That is. argument structure is computed on the basis of the syntactic structure. Thus. whose core is constituted by aspectual projections. 1998. 2001b). in turn. Likewise. Assuming the parallel distinctions between mass/count nouns and atelic/telic verb interpretation. Event-roles are not assigned by a particular head but are simply derived as an entailment from the aktionsart of the whole event. 1999. where functional projections are in charge of Case assignment and agreement processes. 2000. 2001a. Ritter and Rosen (1996. a verb projects an unordered number of arguments whose interpretation depends just on the functional projection they move to.16 For Borer. More accurately. 14 . A [+quantity] object. These works are framed in the Minimalist Program (Chomsky 1995. 2005) defends the idea that Case-checking processes are satisfied through aspectual projections. I introduce some points of the work by Borer (2005). 2000). she argues that aspectual properties of a determined event All the authors cited make proposals in a similar spirit in this regard. 1998. As mentioned in chapter 3.15 In what follows. It is the syntactic structure.

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

1996. who argues that adjectives occur in statives but not in eventive atelic predicates. Borer concludes that they are a derived notion and it is states that emerge from dedicated structure. activities are the event type by default. which denotes the presence of an originator (130). which denotes dynamicity but does not necessarily requires an originator (132)– (133). Borer (2005) proposes that states emerge in the presence of specific functional structure (a sort of “stativizer”). . (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. adverbials blocking a stative interpretation. out of the three event types possible (quantity. 2000. since they are compatible with both of them. or others such as quickly.’ ‘originator. a view commonly upheld in the literature (Kratzer 1994. this would capture the fact that adjectives are predicates of stative events. examples such as (128)–(131) show that adverbials denoting lack of telicity are compatible not only with activities but also with states.’ ‘state. activities are the option by default and states are the result of some specific structure. and activity). She further argues that there are no modifiers restricted to activities. Note that this entails the claim that adjectives can only be stative.’ or any other are grammatically real insofar as modifiers can make reference to them. among many others) and explicitly by Borer herself. That is.Aspectual Alternations in IL Predicates 115 types of homogeneous eventualities (activities and states). *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. According to Borer. Pedro molestó a María intencionadamente Pedro bothered Maria intentionally b. On the one hand. On the other. Such dedicated stative structure preempts the lexical stem entered into the derivation from verbalization. Borer argues that notions such as ‘activity. state. do not make any distinction between quantity and nonquantity (dynamic) predicates. such as intentionally. Bennis 2004.

and. Since. 4. which has been proved to be essential in explaining aspect properties. in particular. but it is the syntactic structure which gives the type of aspectual interpretation. by creating the structure based on functional nodes. accomplishments. and the aspectual properties of the construction. differing from Borer.5 In this section I introduced the approach I will take for the analysis of the aspectual properties of the IL copular constructions.3. Focusing on adjectival copular cases. the conclusion I draw. in the spirit of Borer (2005) and Ritter and Rosen (1996. provides the skeleton where lexical items merge and their interpretation is obtained as a by product. achievements or activities. 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. differing from Borer’s idea. The dynamic properties exhibited by the adjectival constructions studied here (those of the cruel-type) directly challenge this hypothesis. and its absence correlates with the stativity of the construction. Such a perspective is in itself an account of compositionality. I have suggested. by the intervention of the PP complement that some adjectives can have: John was cruel to Pedro.5. this complement is not obligatory. Likewise. 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. More in the spirit of the seminal work by Hale and Keyser (1993). as argued in section 4. .” and dynamic properties come induced from separate projections. 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. 2000). their interpretation. is that the stative status is. I assume that the interpretation of arguments emerges as an entailment from the aspectual structure. I will discuss the claim that adjectival predicates can only be stative. that states are the type by default. I also assume that heterogeneous and homogeneous eventualities are distinguished by a concrete syntactic projection (Quantity). I assume that eventualities are not specified from the lexicon as states. Regarding the main aspectual issue I am concerned with (the state/activity contrast observed in IL copular clauses). Specifically. My discussion of inner aspect properties of IL copular sentences in Spanish takes Borer’s (2005) work as a frame of reference. in some sense. I assume that it is syntax itself which. “default.3 Summary of Section 4.116 Individuals in Time Regarding this specialized stative structure.

the DP subject has to involve a particular set of properties. 4. I argue. 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. 2004). I will argue that the stative version of the cruel-type APs corresponds to their most “basic” structure. but it correlates with the existence of the PP relational complement. (b) When an animate DP subject and the PP are present. the construction has characteristics proper of processes. among others. the dynamic aspectual behavior) actually resides in the preposition itself. 2000. Stowell (1993). As I intimated before. the aspectual alternation (state/activity) is not unpredictable at all. based on Hale (1984). I will discuss the syntactic treatment of the opposition state/activity. From the description in the previous sections. As I develop the proposal. all of them very likely emerging from animacy itself: the DP has to be able to behave as a controller agent. I propose that the dynamic properties become a part of the construction through the preposition heading the PP. it has become clear that the account for the cruel-type constructions has to capture. that prepositions can be conceived of as heads encoding aspect properties. In support of this proposal. .6. and I will also be concerned with the aspectual treatment of the arguments expressed via a PP. As I advanced. as soon as they combine with a (relational) PP.Aspectual Alternations in IL Predicates 117 4.6 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. Let us begin by discussing point (b) concerning the aspectual contribution of the preposition. I propose that the source of dynamicity and its related properties (agency etc. To capture (b).3. I will take such a correlation seriously and I will propose that the source of the aspectual properties (specifically. Point (a) is discussed in section 4.) in the active cruel-type constructions is a projection (the PP) that combines with the “basic” structure of the AP. and Demirdache and UribeEtxebarría (1997.1 Prepositions as Aspect Encoders I have shown that cruel-type APs. behave as activities.6. at least. which gains dynamic properties from the preposition heading the PP denoting the DP goal. the PP makes them a dynamic predicate. since their intervention in the event structure is not obvious (given that they are oblique arguments). the following two properties: (a) When there is a relational PP. To capture (a)—that is. That is. I will address two theoretical points.

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

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

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

where “come from” can be used with the aspectual meaning of the perfect. similar cases are found in Romance languages. For example. in Spanish and English. indicating movement toward (centripetal movement). In turn. 2000. in Dutch. For example. Also. 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. Specifically.Aspectual Alternations in IL Predicates 121 ‘(with)in/on/at’. the locative preposition at directly combines with the verb in infinitival form to construe the progressive.23 Prepositions like the Italian per (‘for’. they notice that. in contrast to stative locative verbs participating in progressive forms. such as from. the prepositions a (142). in Nigerian Margi and Palaung (from the AustroAsiatic family of languages). are present in perfect aspectual forms in some languages. 2004) conclude that Aspect can be analyzed as a head establishing relations between two terms. (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). 23 Note also the presence of a verb denoting movement. appear in the form used to express close future (going to). prepositions of noncentral coincidence are present in the auxiliaries expressing perfect and prospective aspect across languages. . to (143). ‘to’) (144) or the English about to (145) are constituent parts of prospective forms. 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. prepositions denoting ‘movement from’ (centrifugal movement). Demirdache and UribeEtxebarría (1997.

1996) proposal that the semantics of tenses can be described in analogy to prepositions like before or after. Maria had cradled the baby [----------------]///////// c. Demirdache and Uribe-Etxebarría (1997. This time is the Topic Time (TT). Following Reichenbach (1947). In (149). which establish a temporal ordering between time-denoting arguments. Demirdache and Uribe-Etxebarría argue that Aspect can be conceived in parallel terms. (149) a. the TT is captured “before. the kind of the relations being either of central coincidence (progressive aspect) or noncentral coincidence (perfect and prospective aspect). 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. Maria was going to cradle the baby //////////[------------The trees below schematize Demirdache and Uribe-Etxebarría’s (1997. 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. namely. 2004) argue that Aspect can be conceived as a head establishing a temporal ordering relationship between two times. 2000. When I entered the room.122 Individuals in Time Demirdache and Uribe-Etxebarría take as a point of departure Stowell’s (1993. he characterizes tense distinctions as relations of priority. When I entered the room.” The discontinuous line represents the event and the slashes the TT. 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”). Perfect AspP 2 c. 2000) proposal inspired in Hale 198424 and Stowell 1993. 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. (150) a. when he entered the room. and in (149c). In (149b) the TT is located “after” the span of time the event of cradling takes place. Maria was cradling the baby [----------////////------------b. Some clarifying examples appear below. . Progressive AspP 2 b. When I entered the room.

6. 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: . the preposition con.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. denoting the goal.Aspectual Alternations in IL Predicates 123 Assuming that aspectual meaning is reducible to prepositions (and.6. it is conceivable that prepositions can also convey inner aspect properties. and assuming with Demirdache and Uribe-Etxebarría that all temporal relations can be reduced to the same set of primitives. as seen in the examples above. conversely. 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. Preposition con establishes a central coincidence relation between the figure and the ground (in the terms introduced before). and. 4.2. that prepositions have aspectual meaning). I propose that this is the case. ¡corre con tu papa! Pablito. con-PPs can also appear with motion verbs. 4. 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. The preposition introducing the PP in Spanish is. con ‘with’.1 Con: A Locative and a Directional Preposition. like in examples such as the following: (153) Pablito. This preposition is associated with stative locational meanings. paying special attention to the APs I am taking care of in this chapter. I am going to argue that the preposition introducing the (affected) goal carries inner aspectual properties affecting the aspectual nature of the whole construction. 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.

the complement of adjectives of the type of cruel appears introduced by prepositions such as a (‘to’). XIII] 25 I would like to thank Isabel Pérez for bringing this source into my knowledge. in previous centuries. with a goal. Most were locative predicates that can also have a directional meaning. contra (‘against’). S. XIII] (159) (…)& guardandosse segunt su aluedrio fue muy cruel contra los que lo non merecían & keeping himself according to his will. S. Svenonius 2004). 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. the only difference between con ‘with’ and a ‘to’ is the nature of the DP acting as their complements. . 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. in previous periods of Spanish.25 As reported in the corpus. con (‘with’) and a (‘to’) can be said to share properties and can be both analyzed as prepositions denoting a trajectory with a destination. as (157) summarizes. the prepositions introducing the relational complement in cruel-type cases were of same characteristics as con. It is interesting to note that. as a path with a goal (a Goal Path. (157) Con + [+animate] DPs A + [–animate] DPs Therefore. 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. he was very cruel against whom did not deserve it [Alfonso X: General Estoria V. S.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. or hurt them… [Alfonso X : Siete partidas. Consider first the following examples taken from the corpus of Spanish from the Illinois State University (Davies 1999).

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.

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

6. it locates the figure in its trajectory toward the ground. therefore. 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 . associated with ‘future’ in the realm of Tense and with ‘prospective’ in the domain of outer aspect. They are. the discontinuous dotted line for the trajectory. we saw that Demirdache and Uribe-Etxebarría (1997. …● ⌂ In section 4. 2004) proposed that centripetal noncentral prepositions translate as “before” in the temporal domain. (182) summarizes the description of the preposition. the DP subject can be considered as the figure and the DP inside the PP as the ground.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.1. (The bullet stands for the figure and the little house for the ground. The two of them are related by a preposition meaning ‘toward’. that is. and (183) represents it graphically. it designates the movement of a figure toward a place or ground.) (182) (183) TOWARD: preposition that establishes a centripetal noncentral coincidence relation between the figure and the ground. 2000. The two terms are put under a specific relationship according to the meaning of the preposition. 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). As a directional preposition. (181) Juan went to Pedro In the terms of figure and ground presented before.

That is. which translates as ‘nonaccomplishment’. are interpreted in present tense (or future). indicating a reached destination (i.). an incomplete process. in other words. then. The examples (ii)–(v) from Davis (in prep. a concluded process) are interpreted as past tense sentences. no process has been fulfilled33.3. in inner aspect terms. shrewd) that. but not in past.3. cunning. Due to this reason. but merely as a ‘process’ or ‘activity’. there are other adjectives (stupid. we can say that such prepositions convey also the meaning of ‘before’ in the inner aspect domain. illustrate this point. the PP is not contextually inferred if it is not overtly present. Since the figure is captured “before” the ground.e. by their lexical meaning. admit a relational PP complement. (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. although.132 Individuals in Time reached (“the destination has not been reached”). consider the temporal contribution that motion verbs make in Lillooet Salish. (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’ . In principle. The following table (Davis. Because of the same reasoning. sentences with t’iq or tsicw. when these adjectival predicates combine with the PP. Interestingly.) shows motion verbs in the mentioned language. in prep. both refer to an ongoing process. they gain dynamic and agentive properties. (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. to some extent.. Finally. as mentioned in section 4. which argues in favor of the hypothesis that the preposition heading the complement is an actual dynamicity inductor. sentences containing this verb and lacking an explicit temporal marker. it cannot be said it has reached its destination.

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. I repeat the contrasts below. 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” . If that were the case. This fact suggests two things.3 The DP Subject of Relational MPs As I mentioned in the beginning of the section.Aspectual Alternations in IL Predicates 133 construction is decided. the construction cannot appear in contexts proper of dynamic eventualities. First. when adjectives such as cruel are predicated of an inanimate subject. such as the progressive or as a complement of parar de ‘stop’. In sum. and we could not explain why it cannot appear in sentences like (186). ¡sé cruel! “Juan/image.6. it suggests that the properties of the noun in the DP subject and the PP stand in a codependent relationship.3. 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. eventuality properties are obtained from the very syntactic structure. when the DP subject is inanimate. it suggests that the PP is not a plain complement of the adjective. the relational PP complement cannot appear. when the subject is inanimate. (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. the construction cannot appear in agency scenarios either. the PP complement would be possible independently from the properties of the DP subject.2. 4. As shown in section 4. as a complement of verbs like force or in combination with volition adverbials: (190) Occurrence in command imperative Juan/*imagen. And second. such as the imperative form.

Therefore. but just a “theme. which. where the PP stands for a process predicate. located in the preposition. I would like to argue that the interpretive properties of the DP subject are naturally derived from the aspectual characteristics of the construction. as I mentioned). 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). I consider that the element that triggers the aspectual peculiarities of these copular constructions is the preposition heading the PP itself. As I have argued. the structure I propose is the one already introduced above.” to use the traditional vocabulary. 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. together with its animacy properties. the DP is the subject of a movement predicate. The properties of the DP subject come as a simple entailment from their interpretation as an animate figure in a “trajectory”. In this vein. makes it be interpreted as an agent.134 Individuals in Time (192) Combination with volition adverbials Juan/*la imagen fue cruel intencionalmente “Juan/the image was cruel intentionally” Above. the syntactic realization of an agent and that of an experiencer or goal comes together. while the subject of cruel is not an agent. 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. Just by virtue of the fact of being located in the specifier of a movement predicate. 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. This way. which I have justified above as an aspect head. since it needs to bear the required properties (and animacy is the most basic one. In other words. as an ‘agent’ just by virtue of the fact of being an animate figure of a dynamic projection semantically conveying ‘movement’. I took these contrasts to suggest that the subject of “cruel to someone” is a real agent. but the stronger configurational perspective (Hale and .

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

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

That is. Etymologically. red) and refer to dynamic processes. Both cases are similar regarding the elements that participate in them: a preposition. the paraphrases of both being ‘to’ (The Latin preposition in could be followed by a nominal either in ablative case. the preposition had a locative (stative) meaning (English ‘in’). (v) en–negre–c–(er) ? ? ? ? P A v (infinitive-ending) ? ? ≈? con cruel ser . a `centripetal noncentral coincidence preposition. they come from the Latin “in” and “ad”. 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. 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. The prepositions participating in the cases above are en (en-negreció) and a (a-largó).Aspectual Alternations in IL Predicates 137 structure to emerge. 4. black.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. it had a directional meaning (‘to’). which renders parallel deadjectival verbs and cruel-type cases. whereas if followed by an accusative. If followed by a nominal in ablative case. an adjective and a verbal piece. The structure I have proposed for deadjectival verbs exploit the participation of a preposition to account for their dynamic properties. or in accusative case. I suggest that such dynamic properties come from the semantics conveyed by the prepositions that participate in their formation.

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

Aspectual Alternations in IL Predicates 139 both (198) and (199). Parecer takes SCs containing predicates which. (200) and (201). I repeat here the structure I give for the SC the copular verb takes. aspectually. For the same . And third. As a consequence. 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. parallel to the examples with hacer ‘make’. the SC can be considered to bear activity-like properties. 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. I will argue that only those verbs selecting (or allowing) for SCs containing dynamic predicates are compatible with an “AP + PP” SC. only the bare AP is good. the preposition introducing the (affected) goal is actually an aspect head conveying activity-like properties to the construction. I will begin with parecer ‘seem’. (204) ser Juani ser 2 cruel 2 cruel 2 PP (con) = activity inductor 2 (∅toward) con 2 con DP cruel As I largely argued. are states. a state such as be tired looks good. under a causative form. 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.

This hypothesis could explain. therefore behaving as a state. for example.35 Regarding inchoatives like become. even in the simple “parecer + AP” cases. I will bring up two remarks suggesting that seem and “seem + PP” are actually different. for example. precisely. “seem to me” in (iiib). is accepted. in English. When a QP rises from the lower subject position across.c. (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. Although I do not have an explanation to it now. I argue. quantifier lowering is blocked across “seem + PP”. it cannot “lower” at Logical Form. they are aspectually compatible. whereas the simpler one is more defective (looking very close to a modal). the aspectual property they involve is. 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. 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. the contrast between (i) and (ii) is a piece of evidence suggesting that parecer has different properties than “clitic-parecer”. whereas the AP with the PP (an activity SC) gets degraded. (190) with the plain adjective. the same one I propose for the PP in these cases (both involve process properties). b. (iii) a. Tim Stowell (p. Therefore. (The judgments are from Spanish). and does not sound natural in other tenses other than the present. why the clitic forms can be found in all tenses. and the scopal ambiguity disappears.) observes that. (iv) ??María pareció cruel Maria seemed cruel (v) ??Maria ha parecido cruel Maria has seemed cruel .140 Individuals in Time reason.

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

(213) and (214). the next question is what happens with ser. The command cannot refer to the event involved in the SC. does not take the active SC but just the stative. Given the similarities existing between habituals and generics and statives (see chapter 3). However. The different behavior of ser. and AP+PP). confirms ser as a very light verb. on the other. is not as odd with the PP as the causatives.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. Since parecer is a state and. Obviously. its oddity in the imperative form is not a surprise. make) the imperative seems grammatical. it appears that the more semantically vacuous and syntactically simpler (note causative make has quantity in its structure. for example) the preferred selection of these predicates. additionally. which so nicely accepts both forms (AP. Bearing in mind the discussion thus far. In sum. when the active SC is present. which can be considered semantically more vacuous than the causative hacer. the combination of ser with AP or AP+PP gives good results in the imperative. when the command appeals to the main verb (become. inchoative volver(se) ‘become’. and the command over the particular eventuality referred to by “cruel + PP” is the command of the whole sentence. it is degraded. In the same vein. Cruel to animals seems. the command seems to sound better when the PP is not present. parecer gives an odd sentence under such a form either way. and volverse and hacer. Inchoative volverse and causative counterparts get odd in the imperative when the PP is added. That is. on the one hand. A perspective where the (mis)-matching is conceived in aspectual terms seems to make sense of the body of contrasts in (196)–(203). Regarding the inchoative (212) and causative cases. then. 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. this is not surprising. However. to have been reanalyzed as a property at some level. . 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. Finally.

according to which such dynamic properties are in correlation with the status of the predicate as SL.7 I have analyzed the behavior of relational MP SCs as a complement of verbs other than be. and Rothstein (1999). rather.2 Summary of Section 4. I assume a null PP when it is not overt. which can make sense of the preferences of different predicates such as seem. I have shown that the dynamic behavior is not inherent to this type of APs. makes more predictions. I have shown that this aspectual behavior is not idiosyncratic or unpredictable but. kind to someone). unlike previous accounts (Partee 1977. Taking into consideration the aspectual properties of the SC. 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. the better it accepts the active SC. more suitable with those adjectives describing properties that can be understood in relation to another animate entity (cruel. 37 In imperative forms with ser. I have shown that this is not necessarily the case. as the full grammaticality of Spanish copular clauses with ser suggests.1 previous accounts by Partee (1977).7. I argued that to attribute the dynamic properties to pieces inside the SC. . I have shown cases where the PP cannot appear overtly. the better it bears a command imperative. become. This fact led me to reject in section 4.8 Summary of the Chapter In this chapter I have accounted for those copular constructions that behave as activities from an inner aspect perspective. 4. which proves its empirical superiority. corresponds to the presence of a (relational-directional) PP. no null PP would be proposed. Dowty (1979).Aspectual Alternations in IL Predicates 143 verb is. but it is only present if the PP complement is also present. 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. Correspondingly. where they were attributed to the copula. mean. Dowty 1979. accordingly. In order to support the hypothesis that cruel-type APs can also appear without the PP complement.37 4. Rothstein 1999). 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). in the cases where an explicit PP is odd. and make. Although null. I showed that the properties attributed to the SC headed by the adjective are “active” independently of the taking verb. I explained as aspectual (in)-compatibilities the appropriateness of the relational MP SCs with each verb. I also discussed Stowell’s (1991) work. the P brings the dynamic properties enabling forms such as imperatives.

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

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

.

Centered on individual-level predicates. the external argument of Aspect is the internal argument of Tense.e. Throughout the discussion I will focus on three particular viewpoints: the imperfect. 1 The particulars of temporal interpretation are the topic of the next chapter.. I show that the perfective form does not contribute quantity (i. I assumed that different event types are decided through functional structure. for example.2 that aspect viewpoints also involve quantification over occasions. Since. through the functional projection of Quantity. In section 5. which I will capture through a quantifier inside the Aspect composition. Following Borer (2005). so that the complete articulation of Tense and Aspect and their parallel behavior can be easily contemplated. specifically.e. .Chapter 5 Outer Aspect and Individual-Level Predicates In the two previous chapters I dealt with inner-aspect properties. the perfective. I conclude that. I am going to deal now with another aspectual realm. and its absence amounts to leaving the predicate homogeneous. according to Demirdache and Uribe-Etxebarría’s proposal. namely.. I will also present how Tense works here. I also introduced some discussion about the difference between the two types of homogeneous predicates—states and activities. 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. 1995) and Demirdache and Uribe-Etxebarría (2000). Following Klein (1994. I discuss the relationship between inner and outer aspect. 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. I propose in section 5.1 Besides the ordering component just mentioned.1). in Spanish. inner aspect) properties. outer aspect. I will first describe the meaning of different aspectual viewpoints and the way outer aspect works (section 5. outer-aspect properties do not fulfill quantity (i. telicity) to the construction and the imperfect is compatible with quantity predicates (telic predicates). I argue that Aspect is a predicate (located structurally higher than Quantity) that establishes an ordering relationship between two interval arguments. Focusing on the imperfect and perfective. 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.3. and the progressive. Its presence makes the predicate heterogeneous.

1 Tense. as something that took place in the past). inner-aspect characterization may work as an aspect indicator (an accomplishment can be taken as a perfect action.. inner aspect can be as well. posteriority. as Tense and Aspect are considered ordering predicates. like to. therefore. These predicates translate into the temporal domain as ‘before’. in some languages (e. I turn my attention to adjectival IL predicates and show the different interpretations available with each type of viewpoint mentioned. I show that. Therefore. and the possible relationships that can be established among them (anteriority. I tried to show that all aspectual properties could be reduced to the same set of primitives. and Eventuality Time). the interested reader is referred to Partee. posteriority. for example) and.3 Authors such as Zagona (1990) and Stowell (1993.g.e. It turns out that this cannot be described in terms of presence or absence of quantity.. At that point.5. in Spanish at least. 5.148 Individuals in Time In section 5.6. Lillooet Salish). For an introduction of such notions. and inclusion are semantically defined as partial order relations fulfilling a number of properties. I argued that inner aspectual properties could also be conceived in ordering terms. Reichenbach gave a formalization of tenses in natural languages based on three intervals (Reference Time. specifically.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. The idea is depicted in (1). ter Meulen. this being why the eventuality is computed as “dynamic and unbounded” (i. I show that states are compatible even with the perfective. based on Zagona 1990. and simultaneity). which I mentioned briefly in chapter 4 (section 4. Since the line of this work is not formal semantics I do not discuss them here. I argued that dynamicity properties (that activities involve) can be described through prepositions. Consistent with the conclusion that viewpoints do not interfere with quantity properties. the progressive seems to impose restrictions on the type of eventuality. However. 2 . and Wall 1993.1). by the same token. aspect properties can serve as indicators of temporal properties (a perfect action is computed as finished and. atelic).1. In particular. 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. that establish centripetal noncentral coincidence relations between a figure and a ground. 3 Anteriority. 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. but in dynamicity versus lack of dynamicity. Stowell (1993. 5.2 One of the facts I alluded to was that. Speech Time.

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

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

(i) (ii) Maddi had left school at 5 P. ZP(ET) 2 ET PP (at 5 P.6 I will adopt this conception of Tense 6 As for ZPs. In turn. In (ii). In sum. 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. I examine this situation in chapter 6. However. with no additional stipulation.M.e.M. the interpretation is the so-called reference-time reading: the time of Maddi’s leaving is presented as culminated prior to the TT (i. The different readings thus emerge depending on which temporal argument they modify. for the moment.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. .. These authors argue that time adverbials can be considered as modifiers of temporal arguments. Demirdache and Uribe-Etxebarría (2004) provide empirical evidence supporting their actual representation in the syntax. prior to 5). temporal relations are accounted for by arguing that Tense is a dyadic predicate that takes two temporal arguments (ZPs) plus general syntactic rules. the adverb restricts the reference of the ET. yielding the reading according to which Maddi’s leaving takes place at 5. the grammatical existence of such ZPs is demonstrated. the value of the external ZP is the UT.) ZP(TT) 2 TT PP (at 5 P. if the adverbial modifies the TT (iii). Modifiers become the evidence for modified arguments. For example. nothing else hinges on this.M. where I explore the temporal interpretation of IL predicates in simple and compound sentences. the two interpretations of (i) are derived from the different ZP modified in each case.) (iii) To the extent that time adverbials can be considered modifiers of time-denoting arguments.

The lens of a camera.152 Individuals in Time as an ordering predicate. but the interval the speaker is referring to. he intends (10) rather than (9). The analogy between DPs and ZPs enables Stowell to distinguish between “definite” and “indefinite” (with just pure existential force) ZPs. 1996) work. Depending on the focus . the total interval during which María has been washing the car is left in the dark. in principle. Smith (1991) makes an insightful comparison between aspect and the lens of a camera. That time is in the past. as Klein (1994) pointed out: Tense does not order the ET with respect to RT. when someone utters (8). I call the interval the sentence is about (or “the interval the speaker talks about”) “Topic Time” (TT). In (8). This leads to a big conclusion. (10) wants to capture a different assertion: the speaker has a concrete time in his mind and he is asserting that. (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. The former would correspond with readings of the sort in (10). I assume that the element that does this job is Aspect. It orders the time the speaker talks about with respect to a time of reference. for example. Aspect as an Ordering Predicate The temporal interpretation of a sentence like (8) can. 5. the speaker is asserting that María has been washing the car some time in her life. whereas the latter with those along the lines of (9). the obvious question is: what tool picks out the interval the sentence makes an assertion about? Following Smith (1991) and Klein (1994). The observation that Tense can refer to a concrete interval traces back to Partee (1973. makes them visible for the receiver to be able to catch them in a photograph. 1984b) and was also noted in Stowell’s (1993. María was involved in the task of washing the car. However.2 The Internal Argument of Tense: The External Argument of Aspect. (10) There is a specific time at which Maria was involved in the task of washing the car. at that time. by focusing objects. be paraphrased as (9) or (10). 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.1. (10) seems the most accurate paraphrase of (8). That is. 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. Following Klein (1994). According to (9). Between the two. Now.

In sum. …but he did not finish it.) points out. as a whole. However. Before proceeding further. 8 García 1999 is part of the Gramática descriptiva de la lengua española (edited by Bosque and Demonte 1999). the portion of the situation that Aspect “privileges” is the portion of the situation asserted by the speaker in the utterance. we see neither the initial nor the final bound. …and he is still working on it. Focus would make things blurry as opposed to excluded. For example. nothing can be deduced about its ending bound. the event is presented as completed—that is. which I consider a significant work of reference for Spanish. b.” as opposed to focus. some things or others will appear in the picture.c. as they appear in García 1999.7 Only what is made visible is asserted in the sentence. Distinct “aspectual viewpoints” (in Smith’s words) correspond with distinct lenses. As a consequence.Outer Aspect and Individual-Level Predicates 153 setting. The perfective lens allows us to see the entire event: (11) Pablo made a cake [/////////////////////////] Compare (11) with (12). the appropriate analogy here seems to be “wide angle” versus “telephoto. Similarly. there is an asymmetry between these two. for instance. let me introduce a few notes of clarification about the aspect nomenclature I assume and its correspondences with the verbal forms in Spanish. in (11).8 As Tim Stowell (p. the progressive makes visible just some portion inside the event. (13) a. depending on what Aspect focuses. 7 . That is why (12) can be continued either of the ways in (13). some portions or others of the situations will get “registered” and become relevant for the interpretation of the sentence. Smith argues. Whereas the existence of the initial bound (onset) gets presupposed because the asserted part belongs to the inside part of the process. (12) Pablo was making a cake [------//////////-------In (12).

That is. Aspect nomenclature (García 1999) As indicated earlier. The verb estar acts as an auxiliary and can appear in any tense and aspect form. that of ordering temporal arguments. I restrict my attention to the forms called “imperfect” and “perfective” (and concretely to the perfective simple form) as well as with the progressive. The progressive is construed by the verb estar (locative be) plus the ing-form of the verb (Spanish -ndo). is an ordering predicate. Returning to the internal working of Aspect. whereas with the perfect. the adverbial refers to the time at which the event takes place.1.3SG 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. the difference between the composed forms of “have” corresponding to the perfective and the perfect is that. I mention them here only as clarifications. (i) Perfective → the leaving occurs at 3 (ii) Perfect → the leaving occurs prior to 3 . 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. Tense and Aspect consist in the same mechanism—namely. the event has already taken place at the time denoted by the adverbial. which is the element in charge of establishing a relationship between these two intervals.3SG Trabaja-ba he-work-IMPF-PRET. I am not going to analyze all these aspect viewpoints. with the perfective. the next question is how the singled-out portion of the situation (TT) relates to the rest of the event.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. More strictly speaking. I will thus use “perfective preterit” and “imperfect preterit” to refer to the two simple past forms. In the glosses for my Spanish examples. Thus.3SG Trabajó he-work-PERF-PRET. 9 As many authors have pointed out. Aspect. 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.

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

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

there are “two floors”—namely. among many others) are those of closed versus open interval. Closedness and openness are descriptions that can be derived from the ordering properties of aspect. just different arguments to order. “delimitation. Tense and Aspect are explained via the same set of primitives: same mechanism.” because if completed. among others). If the TT is located ‘within’ the event. Perfective AspP 2 TT Asp′ 2 Asp ET (after) c. this is distinct from the projection 11 Two other popular notions to account for Aspect (Comrie (1976) and Bertinetto (1986. If the TT is located ‘within’ the event. Following Verkuyl (1972. the widespread description of the progressive as an “ongoing viewpoint” is derived from the ordering property. Progressive AspP 2 TT Asp′ 2 Asp ET (within) b. Traditionally (Comrie 1976. and a quantificational one. the ordering one just presented. just about its developing. which is explained given the fact that the TT is located just ‘before’ the event. perfective aspect has been associated with “completion” and “delimitation. 1999). it is hard to demonstrate that factors such as “open” or “closed” have any real grammatical import where aspect is concerned. As I will argue. 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.2 Aspect as a Quantifier over Occasions In the articulation of Aspect I propose here. Although these metaphors may be illustrative and clarifying. . 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. where perfective events correspond to closed intervals and imperfect ones to open intervals. 1993.11 Similarly. prospective aspect is described as an indicator of futurity. In sum. delimited. If the TT is located ‘after’ the ET. Aspect orders the TT with respect to the whole span of time the eventuality extends over.Outer Aspect and Individual-Level Predicates 157 (21) Aspect structures a. it does not say anything about its beginning or ending. Finally. the event can be conceived as open. as a consequence. 5. Whereas Tense orders the TT with respect to an RT whose content depends on the syntactic environment. the event can be seen as closed. same contents. This posteriority ordering explains by itself “completion” and. 2000).” These characteristics can be derived from the fact that the perfective locates the TT ‘after’ the ET.

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

Outer Aspect and Individual-Level Predicates 159 with predicates that “hold. García (1999) mentions (i) as an example of continuous imperfect. as in (iiia). it can be defended a cardinality of |1|. (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. the quantifier can be argued to be ∃ (iiib). I would like to point out that its articulation regarding the quantification over occasions is similar to the imperfect. 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.” and can. Prospective |1| AspP 2 TT Asp′ b. it can be also present with eventive verbs. Whereas for sentences like (i). Perfective AspP 2 TT Asp′ 2 Asp Q<occ> (after) 2 1 walk Although I will not work on the prospective aspect. therefore. for other sentences (ii) with a different predicate. be conceived independently from a particular number of instances. where we can distinguish the habitual and continuous forms.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.13 In (27) the differing ordering and quantifying components of each aspect form are represented: (27) a. Progressive AspP 2 TT Asp′ 2 Asp Q<occ> (within) 2 1 walk b.” rather than “take place. con insistencia (i) Durante la reunión me miraba during the meeting me he-look-IMPF-PRET-3SG with insistence .

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

that of “containing. for example). That is. to go walking in the park. the period of his teens is included in the period of time during which Pablo used.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). the progressive and the habitual share the same ordering component. Pablo was walking around the park” por el parque (35) En su adolescencia. (36) -----------xxxxx[/]xxxxxx---------------↑ when I saw him (37) ------x------[x--------x-------x]---------x-ie his teens In (36). and maybe still use in the present. the TT in (35) is specified in the phrase in his teens. . Although the habitual viewpoint will be more extensively discussed shortly. the TT is contained in the period of time when Pablo was walking. Likewise. 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. Pablo used to walk around the park” The relationship established between the TT and the ET is the same in both cases—namely. Pablo paseaba in his teens Pablo walk-PRET-IMPF-3SG around the park “In his teens. the TT interval in his teens of (37) is included in the period of time during which Pablo used to walk in the park.” The TT in (34) is identified by the temporal clause when I saw him. Pablo estaba paseando por el parque when him saw Pablo was walk-PROG around the park “When I saw him. consider the following contrast to illustrate this shared characteristic: (34) Cuando lo vi.

it is interesting to consider accomplishment predicates in perfective followed by a “for + time” adverbial. Although I will discuss neither the process of coercion itself. as she conceives habituals to be. Compare (i) to (ii). whereas the imperfect habitual.) points out. Consider (38). which turns the event (make the cake) into a state. 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. simply. it is interpreted as a progressive (ii). in and of itself. the iterative interpretation seems possible when the DP object can be understood as different instances of the same object description. (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. the sentence is. this does not mean that the number of occasions that an eventuality in the perfective takes place cannot be modified by an adverbial.15 the perfective and the progressive can never denote a plural number of times. as Tim Stowell (p. as a consequence. One can be reading just sections of the paper and playing just parts of the sonata. where this interpretation is much more difficult to get. In (i). However. several factors seem to play a role. nor an account for these cases. refers to a plural number of occasions.162 Individuals in Time Regarding the number of occasions that the different viewpoints can allude to. (iii) Leyó el periódico durante dos horas he-read-PERF-PRET-3SG the paper for two hours . their compatibility with for-adverbials gets explained.16 Likewise. there is no need to interpret that the paper has been read completely or the sonata has been played entirely. that cake is not physically the same cake year after year. the point I want to make with (27) is that.c. and. 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). I would like to note that in the iterative interpretation with the perfective and for-adverbials. Once accomplishments are states. 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). where. 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. For example.

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. . Pablo estaba llamando a la puerta tres veces In the two occasions I heard him. and Systematicity Verkuyl (1999) and Stowell (2000). whereas the progressive and the perfective (in the absence of adverbial modification) refer to a singular occasion.2 The Habitual Interpretation: Iteration. (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. the habitual viewpoint refers to a plural number of occasions by itself.2. 5. Now suppose we interpret (40) and (41) against the following scenarios in (42) and (43). However. Specifically.1 Iteration. I assume that the meaning of habituality is lexically expressed in quantifier adverbials such as habitually or usually. Pablo was knocking at the door three times. as described in (27c). define habituality as an iteration of instances of a given eventuality distributed in time. (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. 5.2.Outer Aspect and Individual-Level Predicates 163 progressive. The speaker’s assumptions about the actual number of instances of a habitually quantified event depends on heterogeneous (pragmatic) factors. Proportion. Following the quantificational-based perspective defended by Verkuyl. In this subsection I deal with the properties of one such quantifier that gives rise to habitual interpretation. Consider the contrast between (40) and (41) as a starting point.17 among others. As just mentioned. (42) Juan smokes four times a year. I consider that habituality corresponds to a quantifier over occasions denoting a plurality of occasions of a particular eventuality.2. the exact number of event instances is not specified. and propose that it is expressed through functional quantificational structure. I will consider that habituality is encoded in the “quantificational floor” of Aspect.

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

This kind of observation is of the same spirit as that by Zemach (1975) and Carlson (1977). such as Kearns (1991).19 Obviously. Just as the interpretation of quantifiers like many is affected by contextual information. However.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. 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. 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. of such an eventuality. In fact. traveling to a foreign city (41). for example. in general. they are not basing it on a strict statistical background but on their perception of what the average can be. In the terms I am proposing here. consider that sentences like (47). but people call them “writers. if the mere assertion of the event means to remain indeterminate about the number of occasions it takes 18 . I represent the interpretation of the habitual quantifier (Hab) as in (48). The habitual quantifier refers to a number of occasions considered “average. In parallel with many in (46). habitual) can be understood as a property characterizing an individual. are not understood as proportional statements but as simple assertions of the existence of the event. with no overt habitual quantifiers. I am defending the opposite view: it is possible to capture habituality in quantificational terms. that an action in imperfect (I would not say.18 That is. 19 Other authors. who argue that not all events are quantified under the same parameters. A ∩ B) is greater than the cardinality of students by a contextual (“norm”) parameter (c). where the contextual parameter captures the average ratio relevant in this case.” 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)). note that ‘John is a writer’ is a possible paraphrase for John writes but not for John usually writes. often observed in the literature.). that is.” where the margins to consider a certain amount of times significantly close to the ratio may be flexible and subject to other pragmatic considerations. so is the interpretation of the habitual quantifier. This makes these authors conclude that it is virtually impossible to talk about quantification over events in an effective way. This would explain the fact. etc. the aforementioned cases would correspond to the continuous imperfect (covering present tense and imperfective preterit). (Think of those who have written only one novel in their life. we say John is a teacher. One can say of John that he is a writer even though he does not usually write. then. Both authors argue that if.”) Other illustrative examples would be sentences like Pablo anda ‘Pablo walks’. 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 the paraphrases for John writes as ‘John is a writer’. 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. but it has to be characterized by the semantics of the right quantifier. when speakers use a habitual form. unlike many. assuming that this perception is shared by most of the society.

contextual information participates in establishing one of the members of the proportion the quantifier refers to. the contextual parameter captures the average ratio relevant in each case. 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. in a simple case like (47). which give us quantities of individuals. rather than discussing the semantics of the present or imperfect preterit. but. Juan fumaba.2.1. then. despite the fact that. (49) En las fiestas de la empresa había marihuana. my main point in this section is to describe habituality. However. As I mentioned in section 5. this is not always the case.” (50) En verano los niños jugaban al tenis en el jardín. at the parties of the company was marijuana Juan smoke-IMPF-PRET-3SG “At the company parties there used to be marijuana around. The cited contextual parameter would capture. the number of occasions that people smoke (in general). 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. in this case. we can paraphrase John smokes by using ‘John is a smoker’. the semantic intuition about examples like (47). As before with many. 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). In principle. Consider (49) and (50) as samples of other possible cases. . to my understanding.2. Juan used to smoke. 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. rather. 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. to the number of occasions that Juan smoked marijuana at the parties with place. it would not be appropriate to use (47) to describe a situation where John has smoked only four times in a year. As shown. aspectual quantifiers give us the number of instances of a particular event. this does not fully capture.

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

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

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

and (b) the co-occurrence of perfective viewpoint and homogeneous (nonquantity. Their telic nature is tested by the appropriateness of the modifier in x time.174 Individuals in Time relation between inner and outer aspect. (84) and (85). either in the perfective or imperfect form. are odd in the presence of such a modifier.3. prepare the meal and write the report.1 Habitual Heterogeneous Predications The Spanish data show that the imperfect-habitual viewpoint is compatible with telic predicates. 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” . Bearing in mind this telicity proof. 5. I will discuss the following points: (a) the compatibility of telic (quantity) predicates and habituality. In contrast. 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. which points to the conclusion that the viewpoint does not affect in any sense the nature of the predicate. atelic) predicates. To begin. which tests the predicate swim as atelic. and offer a general overview of the relationships in the temporal system.

Or. habituality implies that the action at stake has been undertaken again. 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). although the fact that the action takes place several times leads to the inference that each occasion meets an endpoint. Nevertheless. which. put the other way around. . habitual interpretation is fine with atelic predicates. Likewise. an imperfect habitual leaves the predicate telic. In (89) and (90). Each occasion in which Juan wrote a report is heterogeneous. Quantity properties. the aspectual component of quantification over occasions is structurally higher than the projection of inner Quantity. Each occasion will prove to have heterogeneous additivity and subinterval properties. as discussed here. as well as the habitual suffix. these predicates are not turned into homogeneous predicates.3). 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. Each occasion in which Pablo prepared the meal is heterogeneous.21 That is. as seen in (88). activity (homogeneous) predicates such as walk retain their nonquantity properties. the habitual quantifier is predicted to have scope over the adverbial in x time. Likewise. the habitual quantifier in and of itself does not say anything about the actual culmination of the event. Accordingly. correspond to the opposition homogeneous/ heterogeneous.22 21 As I mentioned before. In fact. section 2. prepare the meal and write the report are heterogeneous predicates. can co-occur.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. a telic predicate is compatible with habitual interpretation.1. such as Chierchia’s (1995) (see chapter 2. 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. too. Inner-aspect properties. the telic modifier in x time and the habitual adverbial.

unlike states. a perfective form does not interfere with the quantity nature of the predicate.” as mentioned in the previous section. the TT is located ‘after’ the ET. and others.24 as subinterval and additivity proofs show.2 Perfective Homogeneous Predications I turn my attention now to homogeneous predicates appearing in perfective form. If we say John walked from 2 to 3. As discussed back in chapter 3 (section 3. he has had that car at every single moment between 1974 and 1985.4).3. Contrary to Mourelatos (1978). The meaning of these sentences is that Pablo was once involved in the task of swimming (91) and walking (92). who conceive the perfective as a modifier that makes an eventuality telic.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. In contrast. at least in Spanish. In other words. Piñón (1995). 24 23 . With the perfective. (95) *Pablo nadó en una hora Pablo swim-PRET-PERF-3SG in an hour “Pablo swam in an hour” Other authors. 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. (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. I will argue that. such as Bertinetto (2000).176 Individuals in Time 5. a perfective form does not make a homogeneous predicate heterogeneous. However. 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. In the same vein as before. if we say John had that car from 1974 to 1985. de Swart (1998). it does not mean that he has been walking at every single moment of that interval. draw a conclusion in the same direction. which leads to the inference of “completeness” and “boundedness. activities are not 100 percent homogeneous at a micro-level. Bach (1986).

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

∃) 2 e VP Thus. the ordering component takes the quantified eventuality and orders it with respect to the TT interval. This unifies the working of Aspect. Elaborating on Stowell’s (1993) proposal. (98) ZP 2 Zi VP 2 ei VP (99) AspP 2 TT Asp′ 2 Asp0 ZP (after/before/in) 2 VP Z0 (|1|. where there are no occasions to enumerate (think of stative homogeneous predicates such as Peter is tall). if “one.” perfective or progressive. See section 5. and makes the prediction that the quantifiers (|1|. I argued that. which does not make the sentence habitual. and not only with SL ones. I assume to be present with every type of predicate (SL and IL).4. independently from the kind of predicate. ∃) should be possible with any type of predicate. |>1|. following Stowell (1993. presented in (3) above (repeated here as (98)) about the internal constituency of ZPs. but these predicates are also compatible with the habitual quantifier (She was sick usually). |>1|.” habitual imperfect is at stake.178 Individuals in Time cation) is depicted in (99).26 The quantifier over occasions binds the eventive variable (e) which. Specifically. 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). this is borne out. as Kratzer (1988) originally proposed.1 for further discussion.25 The existential quantifier corresponds to the continuous imperfect. 26 This aspect form is involved with stative SL predicates (She was sick).5. I used Spanish habitual and perfective data to show that aspectual viewpoints do not have any impact on the inner-aspect properties of predicates. in Spanish. In the last section. When the quantifier denotes a number roughly describable as “greater than one. 25 . As will be shown in section 5. 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). 1996).

specifically. on the other. and cruel. In (100). Finally. The copular sentences in Spanish showed that not all predicates in combination with the IL copula ser behave as states. I established an aspectual division inside adjectival IL predicates. Adjectives like Eskimo and other qualifying adjectives like blond behave. ?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” . differently from the cruel-type. The next section continues the discussion of aspectual viewpoints by referring. which confirms it as an activity and not as an accomplishment or achievement among event types. consider just the following three sets of contrasts. (100) As a complement of happen (‘take place’) a. the suitability as a complement of happen shows that the cruel-type in copular constructions can bear typical eventive properties. to the behavior of adjectival IL predicates and states. those that can have a complement introduced by a directional preposition. 5.5 Adjectival Individual-Level Predicates and Viewpoint Aspect In chapter 3.” which suggests that quantification over occasions and quantity (inner aspect) properties belong to different levels. *Sucedió que Juan fue esquimal happened that Juan ser-PERF-PRET-3SG Eskimo “It happened that Juan was Eskimo” b. As a brief reminder. Classical tests show that a well-circumscribed set behave as activities: specifically.Outer Aspect and Individual-Level Predicates 179 mentary distribution with “quantity. on the one hand. (101) shows the different behavior in the progressive form of Eskimo and blond. as has been widely assumed. *Sucedió que Juan era rubio happened that Juan ser-IMPF-PRET-3SG blond “It happened that Juan was blond” c. together. (102) shows the (perfect) entailment derivable from the progressive of the cruel-type.

5. imperfect. Juan está siendo cruel con Pedro Juan is ser-ing cruel to Pedro “Juan is being cruel to Pedro” b. This will lead me to discuss the restriction existing between the progressive and the property of stativity mentioned in chapter 4.3 (about the relation between quantity and perfective and imperfect viewpoint). (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. *(En aquella ocasión) Juan estaba siendo esquimal that time Juan was ser-ing Eskimo “(That time) Juan was being Eskimo” b. 5. *(En aquella ocasión) Juan estaba siendo rubio that time Juan was ser-ing blond “(That time) Juan was being 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” .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. I will explore the interpretation that adjectival IL predicates have under the aspect forms mentioned thus far (perfective.180 Individuals in Time (101) Progressive form a. and progressive). 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. This restriction differs from the one discussed in section 5.

I want to show that. Thus. 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.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. That is to say. that the structures corresponding to (103) and (104) involve an existential quantifier. it is not the case that they are incompatible with a quantifier over occasions by . although the habitual paraphrase is not the most salient one when (IL) stative verbs (such as Eskimo or blond) are at stake. then. where we are not counting the number of times an eventuality holds. at least on a first approximation. that a Q<occ> with a value of greater than one is not present in (103) and (104). successive initial points and successive (inferred) ending points. It seems. (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. it seems. habituality entails that a given eventuality can hold or take place more than once. Since I argued that be Eskimo and be blond pattern with states. I propose. the absence of habitual interpretation with IL predicates such as Eskimo or blond can be explained by their inability to restart. I argued. therefore. but the property is attributed to the person as a whole. However. that states are incompatible with such quantification.

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

The two types of paraphrases (habitual and continuous) are possible with be cruel and each corresponds to a different viewpoint structure. (115) Pablo paseaba Pablo walk-IMPF-PRET-3SG (116) “Pablo used to walk” . The imperfect form can involve different quantifiers over occasions (Q<occ>): (112) a. Habitual AspP 2 TT Asp′ 2 Asp Q<occ> (within) 2 >1 b. 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). the habitual reading is the most salient. in a similar way as we say John was blond or John was Eskimo. That is.Outer Aspect and Individual-Level Predicates 183 (una persona) muy cruel. Nonhabitual (continuous) AspP 2 TT Asp′ 2 Asp Q<occ> (within) 2 ∃ Consider now the viewpoint interpretations when the PP complement is present. that is why his sisters hated him” Contexts (110) and (111) represent scenarios where the property is attributed to the person as a whole. When the PP complement is present. as in (113). 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. the only reading is the habitual reading (114).

they acquire activity-like properties. see section 5. Likewise.30 Finally. although the progressive form is correct with cruel. when a complement that can act as a “distributee” is added (125). whereas activities in imperfect (117) can also have a progressive paraphrase (118).3. the preferred interpretation of adjectives denoting intellectual aptitudes (121) is the continuous reading. a habitual interpretation emerges. as noted above. the progressive is not a paraphrase saliently available from the imperfect form. the progressive paraphrase of the imperfect is more natural in these contexts. However. make the habitual reading available.184 Individuals in Time It is worth noticing that. rather than the habitual (cf.1. (122)).5. accordingly. (119) and (120)). in Spanish. 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. as noted in section 5. this is just marginally available with be cruel (cf.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. 29 .2. 30 For the reasons why the progressive is excluded with Eskimo and blond. when a relational complement is added (123). which.

. and any eventive predicate. or achievements (132). nonstative IL (128).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. chino habitualmente (127) En sus reencarnaciones. activities (130). the objects (the house. stative SL (129). 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. the habitual viewpoint can be said to be compatible with any kind of predicate—stative IL (127). the typo) are subject to referential variability for habitual interpretation to be possible. accomplishments (131).

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

31 As mentioned in chapter 3. the progressive is possible with other class of homogeneous predicates. whereas processes take time and progress in time (although this progression in time does not mean they advance toward a culminating point). (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. As can be seen. the progressive in English works differently in some contexts. 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). consistent with the description of the perfective above. cruel/inteligente toda su vida (140) Juan fue Juan ser-PRET-PERF-3SG cruel/ intelligent all his life “Juan was cruel his entire life” 5.5. according to which it is compatible with basically any type of predicate.3 The Progressive and Adjectival IL Predicates I now turn my attention to the progressive viewpoint in Spanish. and therefore reject the expression of progression in time. in the sense that the progressive is not compatible with every kind of predicate.31 The progressive viewpoint does not seem to work like the imperfect and the perfective.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. . 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. SL (141) and IL (142) statives are excluded with 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.32.188 Individuals in Time The property usually invoked to explain the differences regarding progression (advancement) in time is “dynamism. it is difficult to find an explanation in more technical terms. as the ungrammaticality of cases like *Juan estaba estando enfermo ‘Juan was being sick’ shows. which. The input of energy correlates with the onset of the predicate. In fact. as a result. An eventuality can appear in the progressive if it is divisible in stages. such as Landman (1991). Some authors. does not allow us to use the progressive. 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. nondynamic eventualities do not. eventualities with such a temporal structure that between every two temporal points. we can distinguish different stages in. . progression in time is not possible for those eventualities that are “dense”—that is. based on Carlson’s (1977) notion of “stage. that is. Kearns (1991) explains the incompatibility between the progressive and states by alluding to their property of essentially lacking onsets and culminations.” related to the concept of “movement. However.”34 alludes to “temporal stages” to describe the distribution of the progressive. where the verb empezar ‘start’ refers to the onset of the predicate remain. examples such as the following. for example. 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. an event is a stage of another event if the second can be regarded as a more developed version of the first. See also chapter 6. strictly speaking. However. followed by Bertinetto (2000). Aside from the intuitive understanding of the correlation between dynamism and progression in time. unexplained under this view.33 Landman (1992). According to these authors. In a similar vein. a third point can be established. they are excluded in the progressive form.1) for the introduction of this concept. According to Landman. and. In some sense. since estar enfermo ‘be sick’ and estar preocupado ‘be worried’ are states and. nevertheless. if we can point at it and say ‘it’s the same event in a further stage of development. footnote 4. nevertheless.” Dynamic eventualities progress in time. a sickness. can be argued to lack any input of energy. accordingly. argue that progression in time is not possible for those eventualities lacking internal development.

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

Whereas the object houses does not refer to anything involving temporal extension. 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. because they refer to things that develop through time. these cases. a trip. 36 . This property draws the line between (145a) and (145b–d). 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. The restrictions affecting the acceptability of the progressive are not of mereological nature. Actually. we could not predict that the progressive combines just with one kind of them. some stative cases allowing for the progressive should be mentioned. Pretty much along the lines of the previous examples.190 Individuals in Time cates. then. It seems. (145) a. and only states (SL and IL) are excluded.36 However. that it is the nature of the DP object that matters. only in (145a) can we paraphrase have with ‘possess’. That is. Juan estaba teniendo un ataque al corazón Juan was having a heart attack c. since states and activities pattern alike in that sense. it is such temporal extension that can be conceived in progression. a heart attack. or a day do. the objects in these examples can be considered some sort of “event objects” (Dowty 1979). If there were no grammatical difference (of significance) between states and processes. where the nature of the object matters. either. Juan estaba teniendo un día horrible Juan was having a horrible day Besides the noun directly related to a verb (attack). mereological properties play no role in licensing them. By the same token. As Tim Stowell (p.) points out. It is those referred things that “take place” that can be conceived as involving temporal extension.c. Their defining characteristic is that they allow the progressive depending on the type of object they have. *Juan estaba teniendo casas Juan was having houses b. Juan estaba teniendo un viaje terrible Juan was having a terrible trip d.

the progressive is incompatible with adverbials denoting telicity. 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. the progressive viewpoint (in Spanish) is not a possible option with states. before getting the endpoint) are incompatible because both . Naumann & Piñón 1997.. (150) *Juan estaba arreglando dos sillas en una hora Juan was fixing two chairs in an hour As a dynamic predicate.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. Landman 1992. therefore. Parsons 1990. Summarizing thus far. This is known as the “imperfective paradox” (Dowty 1979) or the “progressive paradox” (see. makes possible their conception in progress. among others. do not come from the properties encoded in the aspectual functional projection of Quantity. Specifically. which. such as in + time. (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.e. Vlach 1981. it has to be noticed that the progressive interferes in a particular way with the mereological properties of the predicates. as a telic predicate. as a result. Asher 1991. However. it is compatible with an in + time adverbial (152). Bertinetto 2000. Delfitto & Bertinetto 1995. the expression of progression and the expression of telicity cannot cooccur. Kazanina & Phillips 2003). I have argued that the properties enabling the use of the progressive are not of mereological nature and. to fix two chairs is compatible with the progressive form (151) and.

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

by exploring the interpretation of IL predicates in compound sentences. I assume tenses are syntactically represented in the node of Tense. the domain of Tense. as a consequence. As mentioned in chapter 2. in particular. is located higher than the Aspect node. This chapter is organized as follows. lifetime effects refer to the interpretation whereby the referent of the DP subject can be interpreted as ‘no longer alive’. First. Second. Likewise. since this tense form gives rise to one of the most important interpretive effects with IL predicates—namely. past. Throughout the discussion I will be chiefly concerned with the interpretation of IL sentences in the past tense. The last section summarizes the conclusions. As it appears. In section 6. 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. which. formerly discussed by Kratzer (1988.Chapter 6 Tense and Individual-Level Predicates In this chapter I study another temporal realm—namely. 1995). I will make special reference to the work by Musan (1995. although it is not only the type of predicate that counts. it does not influence the ultimate temporal interpretation of the sentence.3 presents the discussion of lifetime effects and describes the properties the predicate has to involve to trigger them. In the first section I introduce one previous account for the particularities of the temporal interpretation of IL predicates. Section 6. 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. it is the content of the Topic Time (TT) that ultimately decides the temporal interpretation of the sentence. I will be concerned with the interpretive contribution that tenses (present. In this regard. As I have shown. IL predicates offer different inner-aspect properties. These two features suppose a difference with respect to previous accounts of IL predicates in general and.2. 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. in section 6. and future) have on copular sentences with adjectival predicates. to previous accounts of the temporal interpretation of IL predicates. those dubbed “lifetime effects” (Musan 1995). which has consequences for their temporal interpretation. I propose a division among (adjectival) IL predicates that correlates with different temporal interpretations. . Section 6. The two main points I would like to make here are the following.5 discusses the nature of the TT in greater detail. structurally.

SL predicates IP 1 I′ 1 I VP 1 <e> b.1 Temporal Interpretation as a Consequence of Argument Structure: Kratzer (1988. This author assumes that the factor that draws the line between the two types of predicates is “being temporary” versus “being permanent. a spatiotemporal or eventive (<e>) argument (Davidson 1967). see Anderson 1973. she argues that the eventive argument is the external argument of (SL) predicates. whereas SL predicates needed something “extra.1).1 I will briefly summarize her line of reasoning and show that there are cases it cannot account for. Kratzer proposes that the differentiation between SL and IL predicates resides in their argument structure. The configurational difference between IL and SL predicates can be depicted roughly as in (1). I will mention two points of this proposal.194 Individuals in Time 6. First. the representation in (1).1. in the same vein as Carlson (1977). Kratzer argues that SL predicates involve an extra argument in their argument structure—namely. In particular. SL predicates involve an extra semantic operation.3 (1) a. she proposed that IL predicates applied directly to the subject. temporariness and permanency is encoded in the argument structure of predicates: temporary predicates involve a spatiotemporal argument.” See chapter 2 (section 2. . Furthermore. as Diesing 1992 proposes. 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.2). IL predicates IP 1 I′ 1 I VP 1 θ-subj Regarding temporal interpretation. Kratzer proposes that the spatiotemporal argument is syntactically represented.”2 Specifically.1. the “realization function. 3 As I also said in chapter 2. whereas permanent ones lack such an argument. just aims to capture Kratzer’s idea on the location of the external argument with respect to the maximal projection of the predicate. That is. putting in direct correlation the presence of a spatiotemporal 1 2 For an earlier brief discussion on this phenomenon.” Likewise. 1995) Kratzer (1988. For Carlson (1977). 1995) is the first in establishing an explicit correlation between the temporal interpretation later dubbed “lifetime effects” and IL predicates. As said before in chapter 2 (section 2.

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

The reading disappears. I give an account for the way contextual factors can contribute to block the lifetime reading. so he did not have to go through the immigration process. Consider the following examples as an illustration. In chapters 3 and 4 I focused on the inner aspectual differences among them. On the other hand. Consider (7) as an example.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.4. such as hers. the second part—that is. predicts that the temporal reading of lifetime effects always arises. perfective viewpoint does not trigger lifetime effects. a purely syntactic approach. 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. In chapter 2. That is. In the next sections I will argue that the reasons for this differ. the past tense does not seem to locate in the past either the situation of his being from California or his lifetime. Harry was from California. 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. 1997) noticed. (7) That day. The second point I want to mention is that the lifetime reading does not arise with any type of IL predicate. the past tense of Harry was from California refers back to the moment at which the speaker and Harry arrived in the USA. Harry and I arrived in the USA. In section 6. based on the argument structure. lifetime effects arise more easily when the predicate is just temporally limited . I advanced that equating permanency to IL-hood is not accurate. On the one hand. which is why they do not appear in (8). In examples like (7). as Musan (1995.3 and section 6. Intuitively. but in none of them is a lifetime reading available. despite being parallel to the one mentioned by Kratzer Henry was French. Harry was from California— does not trigger a reading like ‘he is no longer alive’. First. these interpretive effects can be easily neutralized. Throughout this work I have suggested that not all IL predicates can be considered alike. However.

more accurately.” In other words. Regarding the temporal interpretation that gets the existence of the individual implicated (the lifetime reading). if the predication with ser is all right. I am not forced to say that the nature of the predicate has mutated from IL into SL. after having introduced the relevant temporal distinctions. “lifetime permanency” is a property of just a subset of IL predicates. the effects are not borne out. since I argue that permanency is not a necessary characteristic of IL predicates. Since be blond or be cruel does not have to overlap with the whole span of time that an individual’s existence extends over. As before. 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. among other facts. Pablo está Pablo estar-PRES-3SG very funny “Pablo is being very funny” . This perspective leads me to two outcomes. when an IL predicate comes temporally or spatially restricted (some way or other). As mentioned in chapter 2. I take the strong native intuitions about the different interpretation of adjectives that can combine with ser and estar to suggest that. despite the fact that the predicate is not understood as permanent. I consider ser as the copula marking IL-hood. I propose that being “nonpermanent” does not amount to being “stage level. I have consistently considered that every (copulative) predication with ser is IL. Thus. I dedicate a part of chapter 7 to such discussion. each of them involves a different semantics that can properly be described as corresponding to what Carlson (1977) denominated “individual level” and “stage level. I will construct an analysis of the temporal properties of IL predicates. The following paragraphs show examples of copular sentences with ser and different adjectival predicates. permanency or.” Consider. the next contrasts: (11) a. 6. as a reminder. in fact. First. why the Spanish copular verb does not change. as has classically been upheld.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. without being able to explain. Pablo es muy gracioso Pablo ser-PRES-3SG very funny “Pablo is very funny” muy gracioso b. In the remainder of the chapter. a result that proposals equating IL-hood and permanency are pushed to. Throughout this work. Second. I take it to mean that the predicate is IL.Tense and Individual-Level Predicates 197 by the existential limitation of the referent. I am expected to offer an alternative definition for IL predicates.

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

it is not so in the adjectival cases. . However. that the infinitival clause (in Spanish) spells out the interval of time the property holds. at least under this examination.” If the al + infinitive clause constitutes a satisfactory answer. Pedro fue muy amable In walking me home. one way to test whether an al + infinitive clause has a temporal meaning (concretely. this is not a possibility for al + infinitive clauses in the copular sentences at stake. #Al acompañarme a casa. similar to the during + DP or while-clauses above. as (35b) shows. (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). if it does not. Juan se rompió el pie In rolling down the stairs Juan broke his foot (35) a. (34) a. as the # symbol in (31) aims to capture. Pedro fue muy amable al acompañarme a casa Pedro ser-PERF-PRET-3SG very kind in walking me home b. According to García (1999). 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. the status of the clause is not temporal.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. Juan se rompió el pie al rodar por las escaleras Juan broke his foot in rolling down the stairs b. it cannot be concluded. it is proved it acts as a temporal modifier. that of ‘simultaneity’) is by asking “when the eventuality designed by the predicate took place. Al rodar por las escaleras. Hernanz (1999) points out that temporal infinitives can freely appear in either initial or final position (see (34)). 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. does not work as a temporal adjunct.

(i) Al ser tan tarde. Pedro estuvo muy amable In walking-me home. al + infinitive clauses are interpreted as temporal as discussed above or causal clauses. 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. According to Hernanz (1999). As I intimated in chapter 2. the linking to the particular situation is emphasized. Pedro estuvo muy amable al acompañarme a casa Pedro estar-PERF-PRET-3SG very kind in walking-me home b. when the copular verb is the SL estar. 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. which would be along the lines of a causal adjunct. it is necessary to point out that this is not the only meaning the al + infinitive clause has with estar. whose paraphrase appears in (ii).204 Individuals in Time To the contrary. 6 . like that in (i). it is typical of 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. no quise llamarte In being so late. Al acompañarme a casa. With estar. this discussion at least shows the nontemporal nature of these clauses. (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.7 However. 7 The meaning of the al + infinitive clause in these cases is not easy to capture.6 Although I cannot investigate further the status and meaning of al + infinitive clauses with ser here. It also has an interpretation very close to the one exhibited with ser. the al + infinitive clauses can behave as true temporal adjuncts.

nevertheless.2 I have made two main points in this section.1 Necessary Conditions for Lifetime Effects First. literally. 1997). that lifetime effects are subject to certain contextual conditions. according to which the temporal interpretation known as lifetime effects derives from the argument structure of IL predicates. 6. As already mentioned. 6. since there is a large number of predicates that. Differing from Kratzer (1988. I want to briefly enumerate the conditions that are necessary (not sufficient) for the lifetime reading to appear. according to independent grounds (such as their combination with a particular lexical choice of the copula in languages like Spanish). I do not consider that those predicates encode. First. For the interpretation of the DP subject as ‘no longer alive’ to be an option (in languages like Spanish). do not have to be permanent properties.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. for lexical reasons. compare the following sentences: . Second. I argued that those accounts. 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.g. I focus on another restriction for this temporal interpretation. can be argued to be IL and. as. these three conditions have to be met: (a) the predicate has to be an IL lifetime predicate.4 Summary of Section 6.2.3. More accurately. I am going to argue. I argue that lifetime effects are a salient option in those cases where the predicate.. That is. Ph.D. (b) the predicate has to appear in the past form. whether they denote a lifetime property. 1997) observed that contextual factors. In support of the first claim. in the line of Musan (1995. equating IL-hood to property stability is inadequate. In the following section. 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. necessarily produce an inaccurate overgeneralization. in their lexical entry. 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. and (c) the past form should be in the imperfect form. Musan (1995) does. e. can neutralize the lifetime effects.Tense and Individual-Level Predicates 205 6. for example. Musan (1995. 1995). such as the presence of another past tense around.).

does not activate the reading in (43). (43) Pedro is no longer alive The second necessary condition is that the predicate must be tensed in the past. This property will hold of its subject all through his life since it refers to his origin itself and cannot be modified or lost. in the present. an interpretation like (43) is available when the predicate is a lifetime IL predicate. However. A sentence like (44). but can give what can be called a “forward-lifetime effect” (46).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). either. the predicate is understood as referring to school time. 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 . such as be Eskimo. In (40). with an IL predicate that does not denote a necessarily lifetime property. 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.

2004). Now. lifetime predicates are not usually found in the perfective form. The arising (or not) of the lifetime reading hinges on the value that the TT of the sentence has. The interpretation in (49) is available from (47) but not from (48). Whereas imperfect locates the TT ‘within’ the span of time the predicate can extend over. Based on Demirdache and Uribe-Etxebarría’s work. Although I have used. the individual need not be understood as “dead.Tense and Individual-Level Predicates 207 Finally. However. only examples with the copular verb. If he has over-passed it. why does the lifetime reading arise with the imperfect form? In chapters 4 and 5 I claimed. but this form is not excluded with them per se. the sentences that prototypically trigger the reading come in imperfect form. and will keep on using. 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. I argued that the difference between imperfect and perfective was in the ordering predicate they denote. there are three necessary conditions for the lifetime reading to be available. perfective locates the TT ‘after’ the span of time the predicate can extend over. As mentioned in the previous chapter. 1995) and Demirdache and Uribe-Etxebarría (2000. the span of the individual’s life and the span of the predicate coincide.) (50) ------{//////--------------------(51) -----{------}//////----------------Imperfect: within Perfective: after With the imperfect. that viewpoint aspect is an ordering predicate. The first condition is that the predicate must denote an IL lifetime property. I will argue that the lifetime reading arises when the TT at play is the interval that an individual’s lifetime extends over.” In sum. following Klein (1994. I need to first introduce as an assumption something I will explain in more detail in the next section. 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 . bearing this in mind. (The slashes represent the TT. in consonance with the general agenda of the work.

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

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

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

Tense and Individual-Level Predicates 211 sentence—namely. Tense is locating in the past not the whole interval the eventuality (be from California) holds. Likewise. In this case. As I introduced in chapter 5 (section 5. lifetime effects can be reasonably considered a phenomenon of temporal interpretation.e. it seems more accurate to talk about the arising or not of the interpretation.2). no lifetime reading is available. 1997) talks about “the neutralization of lifetime effects. I propose that the lifetime reading arises when the TT is defined as the interval of the individual’s existence: 9 Musan (1995.1). this interval is called “Topic Time” (TT) (Klein 1994). as Stowell (1993) proposes for the other two (see chapter 5. 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. it can be observed that the value of the TT is decisive in the arising of such readings. The occurrence of the lifetime effect is the opposite side of the coin of their absence. I thank Tim Stowell for this remark. the interval corresponding to the arrival in the USA. is located within the span of time at which being from California holds and after the RT. In the spirit of Musan (1995). section 5. since we have proved that it is the content of the TT that matters.1. I assume that the nature of the TT is not different from other intervals like Reference Time (RT) and Eventuality Time (ET). here assumed to be the Utterance Time (UT) (i. In other words. . I assume that the TT is a time-denoting phrase (ZP). Thus.1. the interval at which Maria and Harry arrived in the USA. but the interval the speaker is referring to. From this..” However. since from the perspective I propose here it is not the case that the reading arises and then it becomes neutralized.9 What needs to be defined is which content gives rise to the lifetime interpretation. in the past).

we have seen two types of cases: “out of the blue” cases and sentences inserted in a context. In what follows. it is the whole span of time of Harry’s existence that gets located in the past. the lifetime issue is reduced to the working of temporal interpretation in general. I argue that the temporal TP 2 . 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). I argue that these two cases can be explained uniformly. it gets its content from salient elements in the conversation.3). accounts for them by appealing to different resources (Gricean Maxima of Quantity for “out of the blue” cases. I claimed. In sum. that the TT takes its content from the lifetime interval of the individual referred to by the DP subject. For “out of the blue” examples. in this way differing from Musan.4 The Determination of the TT Content and Lifetime Effects Thus far. a lifetime reading arises (or not) depending on the content of the TT. focusing on sentences with lifetime predicates. I will discuss the role of the contextual information conveyed by the DP subject itself in particular. based on Musan (1995). the reading that arises is a lifetime reading. that referring to the lifetime span of the individual referred to by the DP subject. The next sections discuss the sources for the TT. In this sense. I conceive the lifetime reading as an interpretive outcome derived from the location in the past of a TT with a specific content—namely. Thus. Put in these terms. who.4. For sentences inserted in a context with a previous past interval mentioned. as I said before. also in the line of Musan. and Tense restriction for contextual ones). 6. in (61). that the TT takes its content from the previous interval mentioned (I will refine this account in section 6.212 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. The leading idea in this regard will be that the TT is topic sensitive—that is. I proposed.

following von Fintel (1994). and von Fintel (1994). they are anaphoric to distinct “discourse topics. Among the elements constituting the discursive common ground. Regarding the specific relationship between the discourse topic and the elements influenced by it. the content of their TTs is different) because their TTs get their values from distinct “discourse topics”. lifetime effects do not arise. 1979). In section 6. I will show that the TT is influenced by the discourse topic. As defined by Stalnaker (1972. Consider the following sentences. among many others. a set of propositions that form the shared background of the participants in the conversation.Tense and Individual-Level Predicates 213 interpretation of these two types (with and without context) varies (i. Based on Keenan-Ochs and Schieffelin 1976. Grice (1975).4. 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. a “context” contains a “common ground”—that is. (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 .3 I will show how my proposal (to account for all the cases uniformly) works in a systematic way. Kratzer (1977. The common ground determines the options that are live at any point in the conversation.e.4. In the next two sections. I define “discourse topic” as the subject of immediate concern in the conversation.. 1981). where. I assume that every natural language discourse takes place in a context.1 When the Subject Is a QDP I am going to examine a set of examples without any previous explicit context. that it can be described as an anaphoric relationship. I assume. 6. I assume that sentences are uttered against this background.” Following widespread proposals in semantics and pragmatics. nevertheless.

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

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

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

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

I have argued that this is brought about through the context variable restricting quantifiers (free variable C). That is. so that all the surveyed cases can be accounted for in a uniform way. The relationship between the contextual background and the TT or the free variable C has been taken as anaphoric. The TT refers to a specific interval. A specific interval is an interval that is linked to a previous context. Conversely. 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. The relevant context can be brought to the scenario by the DP subject itself. and. 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. the utterance situation is the most salient context. 3. the context acting as background constitutes the set where anaphoric elements and free variables find their antecedents. how this is articulated. it becomes the source for antecedents. The leading ideas introduced thus far are the following: 1. which contains contextual information associated to the individual referred to by the DP. 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. In other words. following von Fintel (1994). I want to show now. Elaborating on Musan 1995. The occurrence of lifetime effects depends on the content of TT of the sentence. Enç 1991a. where the individual referred to by João is present. 2. 4. more systematically. (77) Salient context " TT . as a consequence. Beghelli & Stowell 1996). 5. I also argued that proper name DPs involve a similar context variable.4. The context relevant to find the “antecedent” for the TT does not have to be another past interval necessarily.218 Individuals in Time situation. 6. as Musan (1995) suggests. This seems to suffice to convert the utterance situation as the background that counts as relevant for the antecedent of the TT. along the same lines that specific DPs are generally understood (Pesetsky 1987.3 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.

(59) Maria and Harry arrived in the USA. 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. Then. and because the context involved refers to an interval that includes the UT. In other words. Result: a past form is allowed and. since it refers to the time of the trip. the temporal interpretations available vary because the contexts the TT interval is linked to vary. they are contain the information relative to previous contextual information. 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. three years ago. what definitely matters for the arising or not of lifetime effects is found in the contextual information of the DP subject. 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. 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. no lifetime effect arises. As topical elements. However. This is because. no lifetime effect can arise. Harry was from California. the contextual information associated to the DP subject becomes of crucial importance in determining the context where the TT gets its antecedent. the DPs are surface subjects. so he did not have to go through the immigration process. while others are simply explained by alluding to a previous TT. in the copular sentences I am analyzing. Since there is no past form. Result: a past form is not allowed. In the proposal I have sketched. which makes them sentence topics. so that a past form becomes excluded. where the information conveyed by the DP proves to be of . this creates an asymmetry in the explanation of (59) and (72) and (73).Tense and Individual-Level Predicates 219 What varies in each case is the salient context and its properties. (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. which resembles Musan’s (1995) examples. It could be the case that some cases have to be accounted for by appealing to the contextual information of the DP subject.

the value it takes by default is the individual himself and the interval associated to him.4 In determining the context relevant for the specification of the TT content (which is what finally decides the temporal interpretation of a sentence). it finds its antecedents in the most salient context. the determiners) have. there is no reason to consider (59) different from (72) and (73). it is the information associated to the DP subject that counts. (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. a lifetime effect does not arise. Finally. then. since it refers to the time of the arrival. As Kratzer (1988. these typically are out of the blue sentences. no lifetime effect arises. 6. I assume that. when there is no specific context the free variable is to be linked to. I propose. Result: lifetime arises With Musan (1995. (…) Topic of the sentence that may establish the connections to a previous background = Harry. Bearing in mind that the TT is a ZP context sensitive. let us consider a sentence where a lifetime reading can arise.4 Summary of Section 6.4. Gregory himself TT → the whole interval overlapping Gregory’s lifetime span. by default. the ZP . since they are the same in all relevant respects—all of them are copular sentences in context. The interpretive procedure is always the same and works the following way. 1995) and Musan (1995) assume.220 Individuals in Time crucial importance in the licensing and interpretation of a past form. I have argued that the establishment of the relevant context is brought about through the free variable DPs (more specifically. Result: a past form is allowed in the copular clause and. that in (59) it is the contextual information associated to the DP Harry that establishes the context that counts. (80) Harry was from California. Since the DP subjects of copular sentences can be considered “topical elements. Actually. (82) Gregory was from America (83) Topic of the sentence that may establish the connections to a previous background = Gregory (84) Gregory c= undefined. 1997). Since the contextual variable directs us to a previous context containing a past interval.” they establish the relevant context. Once a context is fixed.

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

among others. the ordering between the TTs should depend on the specification of the embedded RT. in compound sentences. yielding different interpretations. 6. yielding the interpretation in (iii).) I will deal with complement clauses and relatives in turns.M. the subordinate RT can be affected by the ET of the main clause.14 but rather the TT. yielding the reading in (ii) or the ET. whose nature I argue to be also a ZP. have noted. namely. . as the subindexes of the ZPs in (i) suggest. 1996) work. subordinate clauses (complement and relative). In simple sentences. I have just dealt with IL sentences which establish a link with a previous one paratactically.222 Individuals in Time 6. starting by the former ones. temporal interpretation is derived from the ordering relationship Tense establishes between the RT and the ET. The time adverb at 3 can modify the TT. However. (Thus far. As Hornstein (1990) and Demirdache and Uribe-Etxebarría (2004). this could lead us to a situation where the TTs are not ordered between each other strictly speaking. Peter’s leaving occurs prior to 3 P. and its eventuality will be located with respect to the eventuality of the main clause. by default.5 The Lifetime Reading in Compound Sentences For the sake of a more complete survey of cases. TT and ET can be modified by time adverbs. Thus.M. instead of with respect to the TT. I focus now on the interpretation of lifetime IL predicates in another type of syntactic environment. if the RT of the subordinate clause is in the c-command domain of the ET.1) I described the way temporal interpretation is accounted for following Stowell’s (1993. giving rise to different temporal readings. Following the logic of Stowell’s proposal discussed above. the embedded TT would become ordered with respect to the main ET.1. According to him.1 Complement Clauses In chapter 5 (section 5. Specifically. 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. the UT. the content of the RT is. to understand the temporal interpretation of a compound sentence is to understand the relationship holding between the two (or more) TTs.15 However. In this work. this does not mean that ETs always remain in the dark for interpretation. as in (i). As mentioned in chapter 5. This way. Peter’s leaving occurs at 3 P. I argued that what Tense locates is not the ET as such. (i) (ii) (iii) 15 14 Peter had left at 3 P. the subordinate clause will be temporally dependent on the main one.5. so that it can be controlled by the ET of the main clause. since the RT would be controlled by the main ET rather than by the main TT. the prototypical case of this is the modification of the past perfect. given that the ET is the closest ZP.M.

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

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

Abusch 1988. Ogihara 1996. These two TTs are further ordered between themselves in the way specified in (89). the situation is more complicated.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. As widely noted in the literature (Ladusaw 1977. such an ordering is derived from the control of the subordinate RT by the main clause TT. sentences like (88) have two . When a stative predicate is at stake. Enç 1987. respectively. Stowell 1993. among many others). 1996) terms. In Stowell’s (1993.

Simultaneous interpretation is then explained by arguing that the Tense head itself is null. 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. the interval of John’s being sick overlaps with the interval of Maria’s saying. while the (visible) morphological past is explained due to the presence of past in the ET ZP. Some authors assume a Sequence of Tense rule. One of them is a past-shifted reading (see the adverbial complements appearing in parentheses in (91)). Tense can locate the eventuality in the past.226 Individuals in Time possible temporal readings. which sort of deletes the content of the subordinate tense (Ogihara 1996 would be along these lines). though. Just when a past in the ET ZP is under a past in T. share a similar underlying idea: that the tense in the subordinate clause is not a regular “past”. However.17 17 For a fuller description of this account. but it originates in the ET ZP. That is. From a different perspective. (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. as it appears from the outside. it has to be c-commanded by a past in T. . a reading where the subordinate predicate is understood as temporally overlapping the main predicate. Basically all. sick sick ? ? (92) ----------------------//////X/////////-----------------UT---------↑ said The explanations to account for this phenomenon. stative predicates have another reading: the so-called simultaneous reading. as roughly represented in (92). Stowell (1993) claims that the past and present morphology is not a realization of the category Tense. A rough representation of the idea is in (93). please see Stowell 1993. For a past form to mean ‘past’ (which would give rise to a past-shifted reading). In (88).

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

(TTi). with respect to the (subordinate) RT. what we do is to order such an interval.228 Individuals in Time (94) a. there is something else different in the tree I have drawn: the content of the subordinate TT. María dijo que Juan estaba enfermo ‘Maria said that Juan was sick’ b. the subordinate TT can be either an interval different from the main clause one. First. either present (corresponding to the ordering predicate ‘within’) or null. Then. corresponds to be sick. 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. As the subindexes gloss. . different from the TT of saying. TTj. we can think that a concrete interval (TTj). or the same one. Let me spell out the two options. controlled by the upper TTi.

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

Once I have sketched the mechanics of temporal interpretation in complement clauses with stative subordinate verbs. (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. The past shifted reading is absent. (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. ‘Eva serfrom Albania in a previous life’. as can be tested by the unavailability of paraphrases such as (99). since it emphasizes the relevance of the aspectual head content.230 Individuals in Time to account for the simultaneous reading. I will follow what I said above. I will examine the temporal interpretation of (98) and. . 18 PERF-PRET-3SG At most. I will focus my attention on sentences containing a lifetime predicate. namely. the simultaneous one. Consider in contrast (100). specifically. this could be a paraphrase for Eva fue de Albania en una vida anterior. with a stative SL predicate. will examine whether a lifetime reading is available.

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. 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). First. Now. That is. it cannot shift any TT into the past. simply. Given that a lifetime reading arises when . does this state of affairs trigger a lifetime reading? According to natives’ intuitions. because the content of the subordinate T is assumed to be null. The reasons are two. which is “included” in the span of time over which Eva is from Albania. in (98) it does not arise. 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. The other reason is.

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

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

234 Individuals in Time relative. where the TT of the IL predicate refers to the arrival time. following Stowell’s suggestions. where a simultaneous reading would be an option (making the lifetime reading unavailable as a result). Furthermore. we can suppose that the RC moves along the DP. this case does not look different from (59) above and repeated below. This would seem to pave the way for a lifetime reading to arise. . (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). However. since the content of the subordinate TT is not in the domain of the higher one. and (107a). equivalent to a forward-shifted reading. In fact. can truthfully be captured by (105). the lifetime reading does not arise either. the indefinite DP a guy is interpreted as [+specific] and can be supposed to take wide scope (becoming an adjunct of TP). Bearing all this in mind. (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. consider now the following example with a lifetime predicate in the RC.

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

I examined the arising of the lifetime reading in syntactic composition—specifically. In this respect. are not necessarily lifetime properties. therefore. 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 argued that the contextual restriction of DPs can intervene in building up a previous background. 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. I concluded. I differ from Musan (1995. In both cases the lifetime reading does not arise. I also offered a number of examples of lifetime predicates in the past tense that. Finally. although the syntactic dependence between the matrix and the subordinate tense complexes was different. that interval cannot be “spelled out” by an adverbial. that the content of the TT is not sensitive to syntactic distance but to discourse prominence. Actually. I showed that a lifetime reading is not available in all of the cases where a lifetime predicate appears in the past tense. 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. Thus. as. which prevents a lifetime reading from arising. A large number of questions remain to which I have not even suggested a possible answer. etc. when it refers to the whole interval of an individual’s lifetime. do not have a lifetime interpretation. based on independent grounds. he was from California .5. Others (kind. I have shown that a good number of predicates that can be considered IL. (111) Harry and Maria arrived in the USA. In sections 6.4. if the TT refers to the interval corresponding to the span of time an individual’s lifetime extends over. 1997). and I argued that the arising of the lifetime reading depends on the content of the TT. a lifetime reading will be available. Consider (113) in relation to (111). blond. their good combination with restrictive temporal adjuncts proves. in section 6. who defends the position that the lifetime characteristic comes specified in the lexical entry of the predicate. For example. most of the properties are like the latter ones. Harry was from California. even without the overt presence of another past tense. I assimilated her observation to the theoretical terms assumed here.) can hold of an individual just for a determined period of time shorter than the individual’s lifetime. Among the issues I have not discussed in this work is the relationship between the TT content and time adverbials.3 and 6. I defended the lifetime reading as an interpretive outcome due to a specific content of the TT—namely. complement and relative clauses. for example. 1997) insight that lifetime effects can be neutralized by the presence of another past interval in the context. After introducing Musan’s (1995.

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

.

dark-skinned. the property is predicated of the individual as such: the speaker claims that the subject is a handsome. whose semantic characterization I have argued corresponds to the IL/SL distinction. in this book I sought to explore what lies behind the dichotomy traditionally known as individual-level (IL)/stage-level (SL).Chapter 7 Conclusions and Final Remarks As set out in the presentation. IL predicates are not necessarily permanent properties.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. When ser is involved (1). and. linked to external reasons (maybe because he is wearing a nice suit. (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. the speaker predicates the properties of the subject on a particular occasion. outer aspect. correspondingly. I examined whether this distinction is rooted in temporal terms. The length of the interval an IL property extends over can be restricted in time. I have consistently taken it that any instance of ser yields an IL predication. contrary to widespread belief. ser and estar. In the following pages. Spanish distinguishes between two copulas. by studying the temporal properties of IL predicates in copular sentences at three levels—inner aspect. In particular. and tense. 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. any instance of estar yields an SL one. Only a subset of IL predicates are permanent predicates. First. In the cases with estar (2). 7. or funny person. got tanned. or is in a good mood). most of them denoting properties related to the origin (ethnic or genetic) of the .

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

more in general. Previous claims regarding IL predicates Throughout this work I have analyzed IL predicates in their relation with the temporal realms cited.1 summarizes the claims previously made. By answering this question. and I have given an account for the different temporal interpretations that IL sentences can have without yielding inaccurate generalizations. classically alluded to in the description of IL predicates.2 The IL/SL Dichotomy and the Ser/Estar Contrast With major or minor differences. are temporal concepts in nature. As mentioned in chapter 2.1.” leaving the property without any association to a determined situation. most . Ser simply classifies the subject into the category denoted by the predicate that appears in combination with it. In examining this question we address the key difference between IL/SL predicates. between IL/SL predicates) can be cast in temporal terms. One of the crucial questions is how the association to a situation is brought about. while ser is more “innocuous. One of the important questions here is whether the difference between ser/estar clauses (or. a number of authors have proposed that the IL/SL distinction should be cast in “aspectual” terms. and the semantic domain they belong to. Table 7. where the definition of IL predicates has been mainly associated to the notions of permanent property and stativity. As noted in the beginning of the book.Conclusions and Final Remarks 241 which the Topic Time can take its value. we will be proposing something about the terms in which the IL/SL dichotomy is represented in the grammar. These conclusions differ from most previous literature. Although I will not be able to offer a definitive answer to this. Claim (in previous literature) IL predicates express permanent properties When an IL predicate is at stake. most authors dealing with the ser/estar contrast concur with the idea that estar-predicates put the property in relation to a situation. 7. I will introduce some discussion in the following sections. the notion involved. As was surveyed. 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. these notions.

if we say that SL and IL predicates differ in certain temporal terms. Thus. if we .2.242 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. once “acquired. One of the difficulties in these descriptions is the vagueness as to what exactly the authors mean by such terms.). and gave concrete definitions to each. aspect. To understand a property as permanent means either that the property holds for the entire lifetime span of the individual (intelligent) or that. their definitions. Different number of occasions eventualities occur and different orderings between the Eventuality Time (ET) and the Topic Time (TT) intervals. among others).” it lasts for the remainder of the individual’s lifetime (Ph. and the distinctions that can be made according to each one.3 The IL/SL Dichotomy Is Not a Permanent/Episodic Distinction IL predicates have been defined in the literature as permanent predicates (Kratzer 1988.D.” or “temporally anchored” (in contrast to IL predicates. nonstable predicates. I have considered that each of them is syntactically represented (quantity.” “temporally bound. However. Table 7. outer aspect. tense). In this work. I distinguished three temporal levels (inner aspect. 7. SL predicates are conceived as episodic. I understand that the difference consists of something describable along the lines of the rightmost column. Chierchia 1995. 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). 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. 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. 1995.2 summarizes the temporal units. I have argued that to attribute permanency to a property amounts to assigning it a determined temporal interval that the property extends over. which lack all such characteristics).

the distinction between IL/SL cannot be described in terms of permanent versus episodic. and in (5) the copular ser-clause appears with another number of adverbial complements restricting the interval the property holds. permanency cannot be a defining characteristic of IL predicates. all the following cases (discussed in chapters 2 and 6) become unexplained. the IL/SL dichotomy is not rooted in temporal terms. In (3). in correlation with the wa/ga marking of subjects in Japanese. an adverbial clause restricts the period the property holds. (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. 1 For a recent proposal on the IL/SL distinction in temporal terms. 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}. Since the length of the span of time properties hold does not define IL predicates. but then he changed his attitude (4) (5) In conclusion. at least in temporal terms regarding the length of the interval. The interval a certain property lasts seems independent from its nature as SL or IL. In other words. In (4). 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}.Conclusions and Final Remarks 243 endorse permanency of a property as a definitional characteristic of IL predicates. the copular clause appears as a complement of dejar de ‘stop’. the argument that Tense takes.1 Another important respect in which the proposal defended here differs from earlier accounts (particularly Kratzer 1988. see Torii 2000. which is unexpected if IL predicates are forced to apply for the entire lifetime span of an individual. in direct relation to this. which can be tested by the nonvariation of the copular verb in Spanish. 1995) concerns the structure of IL and SL predicates and.

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

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

I will briefly discuss four points. predicates can be distinguished between heterogeneous and homogeneous. “Estar enfermo” + “estar enfermo” = “estar enfermo” (12) a. ser-clauses do not tolerate in-adverbials under any circumstance. subject to empirical verification. I have described inner aspect properties as mereological properties. In this respect. However. Once we have established a concrete definition of inner aspect.246 Individuals in Time concrete way.). “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. that is. we can see that the differentiation between SL and IL predicates cannot be put in such terms so neatly. I showed that estar (SL) and ser (IL) predicates are both homogeneous predicates. Nevertheless. which we can test by the acceptability of adverbials such as in + x time. 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. both types of predicates give same results in the proofs testing mereological properties (subinterval and additivity tests). A subinterval of “ser de California” is “ser de California” b. as I will show in short. this conclusion deserves some remarks. relying on the subinterval and additivity tests (chapter 3). A subinterval of “estar enfermo” is “estar enfermo” b. as (9) and (10) exemplify here. (11) a. First. it is true that estar predicates can constitute heterogeneous predications. According to their mereological properties. which indicates their impossibility to construe 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’.

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’. from which participles derive. The projection of AspQMAX gets justified by the properties of the participles themselves (or the adjectives derived or related to them). see Bosque 1990. rather than llenar > llenado > lleno (‘fill > filled > full’).3 2 For a detailed discussion about participles. 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. the predicate gains a telic interpretation: morena *(en dos semanas) (16) Con esta crema. Following the theoretical framework I have worked with here.e. estar-clauses admit in-adverbials with certain predicates (participles and perfective adjectives that have been described as “cut-short”2). you get tanned in two weeks” Finally. only if the adverbial in +x time is present. with predicates other than participles and cut-short adjectives (i. which come from heterogeneous verbs.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. AspQMAX is projected. That is: lleno > llenar > llenado (‘full > fill > filled’). Bosque proposes that they are not adjectives derived from participles. ‘sick-INF’. I would like to suggest the idea that in the cases of participles (14) and cut-short adjectives (15). and cut-short adjectives. There are other adjectives (enfermo ‘sick’) related to a verb (enfermar. estás with this cream estar-PRES-3SG dark-skinned in two weeks “With this cream. 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. . 3 For participial and cut-short adjectives to yield a heterogeneous construction. Regarding cut-short adjectives. the ones of our minimal pairs in (1) and (2)). the verb they are related to has to be of heterogeneous nature too. due to their atelic nature. behaving. therefore.. ‘get sick’) which. but rather adjectives yielding verbs.

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

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

whereas the semantics of IL-hood would consist in the absence of such an association. and the other is due to the intervention of one of the copular verbs. That is. the external situational variable (s) exists in addition to the eventive variable 6 I leave aside here those adjectives that combine only with estar. I argue. precisely. describable in terms of categorical versus thetic judgment (Kuroda 1972). both options are quite close to each other). temporal heads are not capable of contributing this external situation. copular verb estar. This way. It is estar that introduces the properties typically associated to SL-hood. Following Demonte (1999). in the association to a particular situation. but it is the entire predicate together with the copula that is IL or SL. the semantics of SL-hood would consist.6 Demonte shows that adjectives have an IL interpretation when used as modifiers inside a DP. According to Higginbotham and Ramchand. 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). or tense. makes the predicate SL. Crucially. The next natural question is what such properties are. there are two possibilities: to propose that adjectives are not IL or SL. the interpretive status of those adjectives without the presence of any copula is the same as when they appear with ser. outer aspect. . there are no differences between ser and estar predicates regarding inner aspect. (In a sense. From this perspective. Higginbotham and Ramchand (1996) conceive the IL/SL contrast as a distinction pertaining to the Logical Form. I want to consider that adjectives are IL by default. Categorical judgments are predications about an individual (x). As I showed. among many others. As introduced in chapter 2. however. or to propose that adjectives are of one kind by default. 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. and restrict my attention to those that can appear with either of the copular verbs.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. Therefore. whereas thetic ones are predications about a situation (s). Along the lines of Higginbotham and Ramchand (1996). I think that the properties of SL-hood consist of the ability to link a property to an external situation.

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

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

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

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

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 . the metaphorical reading proves that the hearer tries to give an interpretation in consonance with the structure. Sometimes this alternation appears in correlation with the (in-)animate properties of the subject (i). Rather. in cases such as (33)–(36).) (i) a. note that there are adjectives of the type of (31) that can appear with ser. have an active or stative reading. Also. However. which only combine with estar (39). either.Conclusions and Final Remarks 255 accounted for.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. the interpretation of the adjective with ser gets. in principle. As many authors have pointed out. combinable with ser. as described in chapter 2. as Fernández Leborans (1995) notes. the necessary combination of (perfect) participles with estar looks consistent. I will not investigate this issue here. cansado Juan está/*es Juan estar/ser. then. depending on their combination with ser or estar. this does not affect the discussion about the combination with the copulas. Nominal predicates do not combine with estar (40).PRES-3SG tired cansado b. respectively. Likewise. a metaphorical reading. 15 There are other types of adjectives derived from participles that. (Note that they correspond to different participial forms in languages such as English. the impossibility of participial adjectives to combine with ser is complex and seems subject to exceptions. leaves unexplained other copular combinations. the idea that estar contributes the meaning of linking to a circumstance with any type of predicate is not accurate.15 The idea suggested previously that any predicate is. such as copula + locative PP.

which would leave (41) without explanation since in Spain is not a temporary property of Madrid. I proposed that the peculiar properties to the constructions these APs participate in emanate from the particular properties the APs themselves involve. the ungrammaticality of Juan es en Brasil is explained. those peculiar properties. The definition of ser given explains why nouns combine with ser (40) (since nouns denote classes). 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. but does not account for why they can appear just with ser and cannot be conceived as linked to a particular situation. This perspective also allows us to avoid. Consistently with this hypothesis. unlike adjectives. Ser classifies the subject by categorizing it into the class denoted by the predicate following the copula. although episodicity can be one of the consequences of the linking to a particular situation. it seems that both notions can be conceived apart from each other. among other things. In chapter 4. the cruel-type. Since a location is not a class.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. such as agency. 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 . en España (41) Madrid está Madrid estar-PRES-3SG in Spain Since a location is not a property of any individual. the obligatory predication with estar is explained. Thus. but something external to it. As we already know. 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. the description of estar as ‘episodic’.

The closeness in interpretation cannot be attributed to an effect of the perfective. poses some issues still unexplained. contrary to ser. the difference between ser and estar seems slightly neutralized. which. Whereas there is a neat contrast between (45) and (46). One of them directly refers to the copula alternation in Spanish. the adjectival property is interpreted restricted in time and tied to a particular occasion. where the contrast between the two copulas looks mitigated. This is left unanswered here. 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). 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)’ . since it appears that estar behaves in this respect as if it were a stative predicate. (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. waiting to be studied with the progressive form itself. which makes the meanings of ser and estar converge. as I mentioned in chapter 5. With cruel-type APs.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. the interpretive difference in (47) and (48) decreases significantly. There are other issues regarding these APs that deserve to be explored in deeper detail in future research. the fact that the progressive is incompatible with estar is not accounted for.

would locate the individual subject at the circumstance expressed by the predicate following the copula. The copula estar. either. Second. I have contributed the idea that the association to a particular situation cannot be recast in terms of tense (the length of the interval). One difference that exists between ser and estar regarding inner-aspect properties is that estar cannot appear in the progressive under any circumstance. links the property to a situation. whereas estar is the only copula capable to constitute heterogeneous predicates with certain adjectives (participles and cut-short adjectives). 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.16 Table 7. that the length of the interval a property holds does not differentiate estar from ser clauses. (1) and (2). regarding the minimal pairs where the IL/SL opposition is detected. outer aspect (by alluding to the completion or perfection of the property). I have argued that the lexical entry of estar includes as a part of its meaning ‘circumstance at which an individual is found’. both predicates ser and estar are homogeneous predicates.258 Individuals in Time 7. or inner aspect (mereological properties). Therefore. I have shown that both estar and ser clauses can appear with the same outer aspect forms. I have shown that. in the first place. Finally. then. Restricting myself to the contrasts argued to correspond to the IL/SL dichotomy of (1) and (2).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.3 summarizes all of these points. by virtue of its lexical characteristics. 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. I have argued so by showing. 16 . since we can construct both types of sentences with adverbials that restrict the duration of the property. regarding inner aspect.

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

.

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

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

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

Uribe-Etxebarría. The ingressive as a new category of verbal Aktionsart. In The Syntax of Time. Dowty. 1979. Demirdache. J. Donnellan. and M. Uriagereka (eds. Depraetere. Ms. Linguistic Inquiry 6:579–588. D. 1986. Philosophical Review 75:281–304. MA: MIT Press. MA: MIT Press. Dowty. 1999. Demonte. 1976. V. 157–186. Dordrecht: Reidel. 1991. . The stative in the progressive and other essence/accident contrasts. Egg. Language 67:547–619. I. Lecarme (eds. I. M.).). 1991b. R. 1995. Cambridge. M. Bosque and V.. Structure-Preserving. On the absence of the present tense morpheme in English. H. and Masullo. D. 2000. Journal of Semantics 12:311–356. 1995. Linguistic Inquiry 22:1–25. Thematic proto-roles and argument selection. The effects of aspectual class on the temporal structure of discourse: Semantics or pragmatics? Linguistics and Philosophy 9:37–61. 143–179. The primitives of temporal relations. V. Martin. 2004. Demonte (eds. and J. Uribe-Etxebarría. 1987. In Gramática Descriptiva de la Lengua Española. Where have all the adjectives gone? Studies in Language 1:19–80.264 Individuals in Time Demirdache. J. 1999. Indefinites. Linguistics and Philosophy 18:1–19. 1975. Demonte.). K. 1966. In Step by Step: Essays on Minimalist Syntax in Honor of Howard Lasnik. Madison. Clases de adjetivos: La posición del adjetivo en el SN. On the necessity of distinguishing between (un)boundedness and (a)telicity. Demonte. Enç. The semantics of specificity. H. 1979. New York: Academic Press. M. Emonds. V. Bosque and V. A Transformational Approach to English Syntax: Root. and M. Uribe-Etxebarría. Ms. 1991a. In Gramática Descriptiva de la Lengua Española. Dixon. Linguistic Inquiry 18:633–657. Dowty. Reference and definite descriptions. MA: MIT Press. Cambridge. M. Enç. Anchoring conditions for tense. Diesing. R. 129–215. J. D. Madrid: Espasa Calpe. Towards a theory of the diversity of temporal systems. Madrid: Espasa Calpe. Demonte (eds. 2461–2524. and Local Transformation. Demirdache. University of Wisconsin.. 1999. University of Nantes and University of the Basque Country/Basque Center for Language Research.). H and M. The syntax of time adverbs. Enç. Michaels. Dowty. M. 1977. D. Sintaxis y semántica de las construcciones con ser y estar. Guéron and J. M. I. Revista Española de Lingüística 9:133–171. Word Meaning and Montague Grammar. 1992. La predicación: Los complementos predicativos. D. M. Cambridge.

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

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

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

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

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

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

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

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

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

.

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

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

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

.

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

106. 189. 142. 14. 249 Existential coda 12 Existential quantification 8. 246. 117. 112 Juba Arabic 120 Kinesis 40 Krio 120 Lakhota 55. 100. 244. 122. 258 Hindi 120 Homogeneous (predicates) 70. 35 Figure 119. 192 Prospective 120–122. 184. 242. 147. 116. 144. 128. 129 Progressive 153–163. 257 Perform(er) 53. 123. 258 Imperfect Continuous 76. 132. 157. 131. 83. 116. 80. 165. 189. 79. 227 Simultaneous reading 76–79. 258. 224. 156 Focus 13. 122. 212 Mereological properties 3. 147. 224. 176. 230 of Narration 77–80 of Relative clauses 223. 134–136. 31 Imperfective paradox see Progressive paradox Initiation. 150. 189. 183. 211. 226. 156 Hawaii-Creole 120 Hebrew 15 Heterogeneous (predicates) 72. 148. 194. 105. 37 Thetic (judgments) 12–14. 94 Margi 121 Maxim of Quantity 85. 11. 168 French 121 Generic operator 11. 230. 144. 227. 58. 257. 11. 192. 227. 148. 141. 148. 201. 175–177. 221–223. 154–157. 212. see also Event(ive) argument Speech Time 122. 233–235. 159. initiator 61. 242. 148. 161.280 Individuals in Time Event-object 110. 158. 172. 72. 115. 180. 178. 183. 26. 129. 79. 148 Sequence of Tense 226. 184. 134 There-sentences 5. 136 Place/situational prepositions 120. 158. 158. 136. 207. 76. 187. 230. see also Utterance Time Stativizer 115. 35. 248. 162. 181. 162. 80. 187–192. 94 Pomo 94 Prepositions Directional prepositions 127. 246–248. 55. 245. 175. 180 Perfective (preterit) 31. 64. 112. 250 . see Temporal interpretation Russian 15 Salish 118. 57. 229. 113. 141. 236 Theme 105. 177–180. 252 Realization function 6. 131. 136. 132. 242. 137 Subinterval property 70. 195. 67. 155–157. 123. 207. 131. 130. 171– 175. 173. 215 Goal(Path) 96–98. 248 Swahili 120 Temporal interpretation of Complement clauses 76. 189. 134–136. 148. 88. 246. 176–178. 128–131. 186. 129. 194 Reference Time (RT) 3. 86. 160. 189. 123– 126. 212. 249. 192 (Im)perfective contrast 25. 134 Instigator 55. 244. 178. 184. 144. 173–180. 229. 158. 187. 118. 148–152. 186. 159 Quantification restriction 11 Quantity 113–116. 172–174. 153–160. 211. 186. 158–160. 176. 240. 248. 226. 148–152. 72. 249 Relative clauses temporal interpretation of. 125. 35. 174–178. see also Subinterval property and Additivity property Narration temporal interpretation of. 165. 152. 94. 155. 186. 209. 72. 221 Habitual 77. 135. 172. 246. 14. 222–225. 227–229. 178. 12. 191. 258 Progressive paradox 191. 158–160. 167. 234 Situation aspect 39 Small v 55 Spatiotemporal argument/variable 8–10. 231. 179. 147. 132. 147. 196. see Temporal interpretation Past-shifted reading 76–78. 233. 12. 186. 192. 233 Perfect entailment 47. 181 Existential reading 6. 191 Event(uality) Time (ET) 3. 189– 192. 154. 14 Generic reading 6. 202 Ground 119. 241. 226. 208. 111. 240. 222. 6. 232.

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

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

Sign up to vote on this title
UsefulNot useful