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



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)


© 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

.......................................................................15 2.....................2............53 3..3 Summary of Section 2............................... When There Is More Than One Copular Verb: Spanish Ser and Estar...... Kratzer (1988.....................39 3..........53 3..5 Summary of Section 2.................................................................2..3 IL Predicates as Inherently Generic Predicates....1 The IL/SL Distinction as a Semantic Distinction........2 The IL/SL Dichotomy as a Syntactic Distinction...1...................14 2...................1 .............1 Chapter 2 Individual-Level Predicates .........2 Event Types and Event Structure...................... Raposo and Uriagereka (1995)................................................................................................... The Structure of Copular Constructions ...............2 ...................2.........................1..........3 Summary of Section 3................................1 Distribution of Copular Verbs in Spanish .....2........................................................8 2.........................3 Aspectual Differences among Individual-Level Predicates ...........1 .............................................1 Proposals about the Individual-Level/Stage-Level Dichotomy ......1.................2....................40 3................ Chierchia (1995) .............................................................1.......................3 Summary of Section 3.........................................Table of Contents Acknowledgments.....................2...2..........4 The IL/SL Contrast as a Categorical/Thetic Distinction...........................................33 2.... Milsark (1974) and Carlson (1977).........................................2 Ser and Estar as an Aspectual Differentiation ......... Introduction to the Tests Discerning the IL/SL Status....................................11 2.........5 2.............................xi Chapter 1 Presentation of the Study ...................1..32 2.......5 2...............2......38 Chapter 3 Event Classes and Individual-Level Predicates ..............................................1 A Cluster of Notions ....... Higginbotham and Ramchand (1996) ................................................16 2....................39 3...1........61 ..... Inner Aspect and Event Types ...ix Foreword.61 3..39 3........................................2 Agents in Event Structure ...........................59 3....4 Summary of the Chapter .........5 2.............................................................25 2..................1..........................1 Inner Aspect...12 2............................ A Brief Stop at “Agentivity”.................2 ......51 3.................1.....3.1.................... 1995) ... by Tim Stowell ..........................

........................... 69 3.......6..........4 Summary of Section 3..76 3..........................98 4.............................................. States and Activities: A Grammatically Relevant Difference?.....1 On the Interpretation of the Relational PP .................4...3 .62 3...............................6 An Account Based on the Relational PP Complement ..6.......1 Aspectual Tests on IL Predicate ....................1 Prepositions as Aspect Encoders .......................116 4. The Cruel-type as a Small Clause Integrant.............4.......................5..3..143 ...................................133 4..........3...............4 Summary of Section 4..................................................................................... Summary of the Chapter ........3.............4............7.....................7....69 3....6.3 The Relational PP with Other APs ..................................2....................................7........1.............96 Individuals in Time 3.. Summary of the Chapter ....84 4................................3 Some Confusing Arguments in Defense of a State/Activity Distinction............................137 4............................................90 4..6............117 4..........................2 Summary of Section 3.........1 Cruel-type Small Clauses Taken by Verbs Other Than the Copula .................3 Relational Mental Properties....72 3........2 On the Optionality of the (Affected) Goal PP .......5....................................................................................................................108 4.................105 4.................................143 4........5.......................3.........................2 Be as a Copula Shifting a State into an Activity ........3 Summary of Section 4..............................1.................5 .........4...................................138 4..................................3....96 4...................2 Summary of Section 4....1 Similarities between Activities and States ...1...............8.81 Chapter 4 Aspectual Alternations in Individual-Level Predicates ..............68 3....................................91 4........4 ...................4 The Issue: Aspectual Alternation in IL Copular Clauses ....................................106 4............2 The Preposition Introducing the “(Affected) Goal”: An Activity Inductor123 4.......................107 4.......5.........................................................2 Summary of Section 4....7 ......83 4.....................................................84 4....................................................1 Lexicalist and Logico-Semantic Approaches .......108 4..............1 The Two-Copulas Hypothesis ........................................1 Lexical Semantic Classes of Adjectives .....2 .....3 Adjectives Predicated of an Implicit Event Argument .2 Differences between Activities and States ............110 4.3 .......5 Justifying the Approach ................2 Syntactic Approaches ....4.......................................................................1 Previous Explanations of Nonstative Copular Clauses ...95 4....2 Adjectival Predicates Showing an Activity-like Behavior .........................3 The DP Subject of Relational MPs .2................85 4.................4 Summary of Section 4.............86 4.......80 3.135 4.................................. The Relational PP Complement....................3..117 4.................6 ...........

...2...197 6...1 Tense...................147 5..................173 5..1 When the Subject Is a QDP .....................4...........2..... 187 5..........................2 Context Associated to Individuals ............205 6.....................................................................4.......................... Kratzer (1988........................................218 6...1 Quantifying over Occasions ..................4.....................1 Habitual Heterogeneous Predications..... 1996) ..5 Adjectival Individual-Level Predicates and Viewpoint Aspect ......2..157 5.............. and Systematicity ....3 Brief Notes about the Complement of Mental Properties APs ......1 Temporal Interpretation as a Consequence of Argument Structure.....212 6..........................................................3........ 1997) ..158 5.................................................................. Proportion...... A Brief Summary of Aspect Notions .....6...4.......5.....4 ..............2 Nonpermanent IL Predicates........3 The Arising of Lifetime Effects ................1........5....... 179 5.................199 6...3 Inner and Outer Aspect ...5 The Lifetime Reading in Compound Sentences .............2 Perfective Homogeneous Predications ........................................3............................................................................ 177 5........................2.................................................................................................................205 6.........................................5....... 186 5....3..............2..2 The Habitual Interpretation: Iteration........................................................208 6.....1 The Imperfect and Adjectival IL Predicates ................................................... Stowell (1993..............................................210 6..148 5..........4 The Determination of the TT Content and Lifetime Effects ............3 On the Relation between the TT and the Habitual Q<occ> ......................1 Permanent IL Predicates .....4 Summary of Section 6.......148 5......................................................................................2 Aspect as a Quantifier over Occasions ......................193 6.......................................................................................................................3 The Content of the TT in the Arising of Lifetime Readings ........222 ..Table of Contents vii Chapter 5 Outer Aspect and Individual-Level Predicates ......2.................2 ...........................171 5....152 5.163 5........2 The Internal Argument of Tense: The External Argument of Aspect.............2 Introducing the Determining Role of Contextual Factors........ 176 5.............. 180 5................................202 6... Musan (1995...................................220 6....3 Articulating the Account............205 6.........215 6......2 The Perfective and Adjectival IL Predicates ..194 6............................................. Aspect as an Ordering Predicate .........3..............................213 6.................................. 1995) ................. 192 Chapter 6 Tense and Individual-Level Predicates .............199 6.3 The Progressive and Adjectival IL Predicates . Summary of the Chapter .................................................................1.............3.....................4..........1 Tense and Aspect as Ordering Predicates ..........1 Necessary Conditions for Lifetime Effects .......2 Differentiating Temporal Extensions of IL Predicates ............................2........................... 174 5...............................................4 Summary of Section 6..

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

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

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

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

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

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


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

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

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

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. I show that inner aspect properties are independent from outer aspect ones (i. In this respect. Chapter 6 centers on the temporal interpretation of IL predicates.4 Individuals in Time duced by complement these adjectives can have. Chapter 5 is devoted to the analysis of some of the outer aspect forms: specifically. focusing on Lifetime Effects. ordering and quantification over occasions). the influence of contextual factors on the arising of the lifetime reading is derived. the lifetime interpretation depends on the content of the Topic Time.. 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. the perfective. paying special attention to whether the perfective plays any role in making atelic eventualities telic. the imperfect. I will put forth two claims: first. and second. Finally.e. . the lifetime reading is a salient reading only with lifetime IL properties. and the progressive.

1 The IL/SL Distinction as a Semantic Distinction. (1) (2) a. Likewise. I concentrate on the most influential proposals about the contrast between individual-level (IL) and stagelevel (SL) predicates. a part of this chapter presents the differences between the two Spanish copulas (ser and estar). There are several policemen in the corner a. 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. In section 2. There are several policemen available b. I present other two aspects basic for my purposes in this work. *There are several policemen intelligent b. indicating some of the points on which my proposal differs. I will introduce the syntactic analysis that I will assume for copular sentences.1 Proposals about the Individual-Level/Stage-Level Dichotomy 2.4 summarizes the chapter. For example.1. 2. PP.Chapter 2 Individual-Level Predicates In this chapter I introduce the concept of the individual-level predicate. ser and estar. *There are several policemen insane . the one offered by Stowell (1978. In section 2. I present the syntactic analysis I assume for copular cases—namely.2 is devoted to the differences between the two Spanish copulas. according to most authors. I will critically revise those analyses. since I am chiefly concerned with copular cases.1. This chapter is organized as follows. and introduce some of the concrete aspects for which this work makes alternative proposals. according to which the copular verb selects for a small clause (SC) containing the nominal argument and the predicate (AP. Milsark (1974) and Carlson (1977). who discuss the constituency of the SCs that copular verbs take (in particular. 1981). and pragmatic distinctions. IL and SL. I also introduce proposals by Heycock (1994) and Becker (2000).3. the presence of functional projections) and its relationship with the IL/SL distinction. which. After that. Section 2. I will first present the main descriptions of this notion found in the literature. arguing for an aspectual distinction. specifically discussing arguments for semantic. lexically distinguish the two types of predication. the predicates of (1) are allowed. syntactic. but those in (2) are excluded. NP). Since the data in the work are from Spanish. Section 2.

the relation between in Los Angeles and John is distinct. as “states. as in (5). Carlson argues that subject bare plurals in English are interpreted differently depending on the type of predicate. and the latter stage-level (SL) predicates. j for John. (4) means that there is a stage in Los Angeles that realizes the individual John. However.” Types of properties differ. SL predicates need an extra semantic operation (the realization function R). M stands for mammal. For example. the type of those allowed in them. Compare these two sentences. the DP subject can only be understood as generic. Whereas. a mammal (M) is a property that applies to the individual John (j). j) & in (Los Angeles) (y)] For Carlson. 1 Section 2.6 Individuals in Time Milsark (1974:211) describes the contrast in the following terms. . IL predicates apply to their subject directly. However.1 contains a few remarks about this test. If the predicate is SL. Carlson (1977) proposed a distinction between those predicates that apply to individuals and those that apply to stages or happenings of individuals. stages and individuals. this distinction between IL and SL also draws the restriction between being able to appear as a complement of a perception verb1 or not. whereas states are conditions whose removal does not cause any change in the essential qualities of the entity.2. possessed by the entity. in turn. In support of the difference between IL and SL.” He defines properties as those facts about entities which are. in some sense. according to Carlson. if the predicate is IL.” An individual. then. (5) (6) Dogs are in the backyard Dogs are mammals (existential or generic) (generic only) According to Carlson. is defined as “that whatever-it-is that ties a series of stages together to make them stages of the same thing. a stage is defined as “a spatially and temporally bounded manifestation of something. (3) (4) John is a mammal M (j) John is in Los Angeles ∃y [R (y. Carlson (1977) brings up other scenarios where each kind of predicate behaves differently. Quoting directly from Carlson (1977:115). in what they are predicated of. In Los Angeles is not predicated of the individual John but of a slice of him. Elaborating on these insights. He argues that the predicates that are excluded from there-sentences can be characterized as “properties”. the DP subject is ambiguous between an existential and a generic reading. This distinction is linked to another previous distinction: the differentiation between two ontological categories. The former are called individual-level (IL) predicates.

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

denoting nonactions but performing actions and proposed that events could be treated as individuals. 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. Kratzer (1988. PP. in the bathroom. This gives rise to a difference in their argument structure which is the stem of the difference between the two types of predicates. denoting actions. at midnight. and nouns. (In the default case. 2.3 In Kratzer’s proposal. as in (i). such as Kratzer (1988. as Diesing 1992 proposes.1. 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. (∃e) (buttered (Jones. to mention just a few. the toast. from Davidson 1967. such as Jäger (1999) and Becker (2000). Kratzer’s proposal looks like (13). As will be shown in section 2.2 The IL/SL Dichotomy as a Syntactic Distinction. whereas SL predicates project this extra argument (which can be understood as a spatiotemporal argument).8 Individuals in Time This view has been adopted by some authors. e) & (at midnight. 1995) claims that IL and SL predicates differ with respect to what they are predicated of. Whereas IL predicates are predicated of the individual directly.2 The eventive variable is considered a primitive element in the logical semantics of events. e) & (in the bathroom. (i) Jones buttered the toast slowly. He proposed that action sentences include an event variable as a primitive element in their logical semantics. SL predicates are predicated of something else: the eventive variable proposed by Davidson (1967). like nominal variables. Schematically. 1995) and Diesing (1992) but disputed by others. it is the DP referring to the individual itself which functions as external argument. based on Williams’s (1981) Argument Linking description. which. Kratzer (1988. NP). 1995) In the spirit of Carlson (1977). e) & (slowly.3. the event variable is existentially quantified). can be modified and quantified over by adjuncts.4 2 Davidson (1967) took the classical differentiation from Plato between verbs. . When it is not projected. 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. IL predicates do not. with a knife. e)). 3 4 I will add a few more words about the eventive variable in chapter 4. The Davidsonian argument is the external argument of the predicate.

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

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

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

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

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

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

A large part of the population uses Bambara as its mother tongue. Spanish and Portuguese have two lexically different copular verbs. those clauses where the topic is an individual are instances of IL predications.9 In Russian. hinging on whether the predicate expresses an inherent or definitional property of the subject (38) or not (39) (Greenberg 1994). The Hebrew copula alternates between overt or null. 1996. In the past-tense cases.2 When There Is More than One Copular Verb: Spanish Ser and Estar A good number of languages show copular alternations. Other languages. the postcopular NP (the predicate) alternates in case (nominative or instrumental) depending on whether the property is permanent (40) or nonpermanent or noninherent (41).8 distinguish three types of copular verbs depending on the category the predicate belongs to (37). such as Bambara (Koopman 1992). In the next section. 2. I expand the discussion regarding what seems to be an adequate description for the IL/SL dichotomy. I have introduced the idea that equating permanency of a property and ILhood is inaccurate. When the predicate is an AP → copula “ka” b. 9 For further discussion on the overtness of Hebrew copula. and as secondary language it is employed to communicate nationwide. ser and estar. found across most of western Africa. When the predicate is a PP → copula “bè” (38) Ha-kli ha-ze *(i) patis he tool the this (COP)MASC-3SG hammer “This tool is a hammer” me’od ‘ayef (39) Dani (*hu) Dani (COP)MASC-3SG very tired “Dani is very tired today” byl durak (40) Oleg Oleg-NOM was fool-NOM “Oleg was a fool” ha-yom today 8 One of the closely related languages collectively called Manding. Bambara is one of the officially declared languages of Mali. When the predicate is an NP → copula “don” c. the copula is null in the present tense but overt in the past tense (Kondrashova 1995. where the semantics of the two Spanish copulas are introduced. I take up this issue also in chapter 7.Individual-Level Predicates 15 are instances of SL predications. Matushansky 2000). see also Rapoport 1987 and Rothstein 1995. (37) a. . In turn.

Pedro es/*está Pedro ser/estar-PRES-3SG teacher “Pedro is a teacher” When the predicate is a PP. How can the distribution of the two copular verbs be described? As traditionally observed. the copula used can be estar. there is no general agreement in the literature on how the copula alternation should be characterized. However. The classic counterexample in this respect is estar muerto ‘be dead’. insufficient to characterize the copular alternation.1.2. if the subject is a physical entity (43).” as cited previously (see section 2. In the case of Spanish. discussed in section 2. as . I would like to note Demonte’s (1979) observation that this distinction poses an asymmetry in the account of copula distribution. which led to the claim that ser corresponds to permanent properties and estar to temporary ones. 2. if the subject denotes an event (44). obviously designates a nontemporary property.” The copula ser is used to refer to properties that are considered essential of the individual. when the predicate is an NP.12 10 Note that these descriptions fall in line with the ones given by Milsark (1974) for “properties” and “states. 11 See also comments by Milsark (1974) and Carlson (1977) about the same predicate. un ordenador de 1960 (42) a.11 which. Ese artefacto es/*está that artifact ser/estar-PRES-3SG a computer from 1960 “That artifact is a computer from 1960” una cabra b.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.1. if not a mistake.1. other authors observed that the distinction between permanent and temporary was.1 Distribution of Copular Verbs in Spanish Some traditional grammarians (Bello 1847.10 Essential properties are understood as permanent. 12 Although I will not work out this difference here. if. despite being ungrammatical with ser. Este animal es/*está that animal ser/estar-PRES-3SG a goat “That animal is a goat” professor c. whereas estar is used to refer to those properties considered accidental. Roldán 1974) describe the distribution of the two copulas ser and estar on the basis of the Aristotelian dichotomy “essence” versus “accident.1). the copular verb must be ser (42). and accidental properties as temporary. or ser.

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

Further discussion is presented in chapter 7 (see section 7. and Demonte 1999.6). However. Bosque and Picallo 1996.13. For more details about classifying adjectives. 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.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. one of the properties that Schmidt (1972) and Bache (1978) provide to distinguish classifying adjectives from qualifying ones is. *The trip was presidential). 14 Note. When the adjective is referred to the place where one is born. it is not gradable. and those in (51) combine with both copular verbs. These tests are (a) impossible participation in comparatives (47). (b) impossible modification by a degree adverb (48). also. that classifying adjectives are excluded as copulative predicates (The presidential trip vs. precisely. and (c) impossible participation in binary correlations (49).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. 15 13 . see Bosque 1993. 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. If the subject is a resultative nominal. that their acceptability as predicates in copular constructions seems to be confined to those cases where the subject is a physical entity.

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

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. (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” . I am dealing with an IL predicate. Thus. the copula designing SL-hood.1. Interestingly. (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). all of the inflected versions in (55)–(58) that we can construe as paraphrases have to be done by using estar. With the copular distinction in Spanish in mind. see section 2. 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. not ser— that is. Correspondingly. pero está muy guapo (53) Pablo no es Pablo not ser-PRES-3SG handsome but estar-PRES-3SG very handsome “Pablo is not handsome. the copula estar will be considered a lexical sign of an SL predicate. where a perception verb (note) has a SC (NP + AP) as a complement.20 Individuals in Time nada gracioso.1.1). pero está moreno (54) Pablo es Pablo ser-PRES-3SG very light-skinned but estar-PRES-3SG dark-skinned “Pablo is very light-skinned. Consider the next group of examples. but he looks very handsome” muy pálido. throughout this work. Consider (59)–(62). where the pound symbol indicates when the paraphrases are not appropriate. whenever the copular verb is ser. I will consider that.

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

verbs such as percibir ‘perceive’ or advertir . correspondingly. then. Notar seems sensitive to other properties of the predicates appearing in the small clause selected by it. they argue. correlates with an “absolute” (i. context dependent) standard of comparison: (v) partially/half/completely {empty. non-context-dependent) standard of comparison. 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). inexpensive} We could.22 Individuals in Time Other perception verbs. open. gradable adjectives incompatible with such modifiers involve an open scale structure (without minimal/maximal elements).e. For example.. (vii) and (viii)).c. In turn. which. it is not the case that any perception verb allows for any type of adjectives as its complement.) notes. 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. 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. The following sentences where they are used in comparatives and are modified by proportional degree modifiers such as parcialmente ‘partially’. full. (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. closed} (vi) ??partially/half/completely {long. However. completamente ‘completely’. whereas it allows open-scale adjectives. interesting. adjectives compatible with proportional degree modifiers are those gradable predicates with a closed structure scale (with minimal/maximal elements).18 such as ver ‘see’ seem to discriminate IL versus SL at first sight.19 since the adjectives yielding grammatical sentences in contexts 17 The semantics of the adjectival predicates admitted with notar deserves a much deeper exploration than what I can offer here. correlates with a “relative” (i. it rejects closed-scale ones. although cansado ‘tired’ and preocupado ‘worried’ seem to behave as proper closed-scale adjectives (cf.. alterado ‘agitated’ sounds quite natural when combined with a proportional degree modifier. This and other issues concerning the semantics of the predicates allowed by perception verbs need further examination. this does not mean that predicates such as desnudo ‘naked’ or descalzo ‘barefoot’ are not gradable. short.e. which.

they appear to contrast (Tu hermano puede venir al cine. rather than ser (72).Individual-Level Predicates 23 such as (70) and (71) are those that combine with estar (cf. where a state such as know languages is excluded. the verb see distinguishes between states and events. Consider the following contrast. 19 With infinitive complements. as in (ii). pero tú te quedas ‘Your brother can come to the theater. (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. physical. and disambiguate between verbal forms (cantaba ‘sing-IMPF’ can refer to the first and the third person). emphasize (Quiero que vengas tú ‘I want you [and not other] to come’). it is interesting to note the role that overt subjects (pronouns) play in the meaning of this perception verb in Spanish. or gracioso. and so on that contribute to the acceptability of the sentence. 20 . when the adjective is combinable with the two copulas. 50 above) and. Other adjectives. not all the results are so clear cut. moreno. rather than IL/SL-hood. pálido. the corresponding paraphrases contain estar. the semantics of perception verbs deserves a deeper investigation than what I can offer here. however. 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. but you stay’). 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 16.20 When they are overt. Among the issues to be explored would be the specifics of the meaning of these verbs when combined with small clauses of the type under discussion (whether they refer to mental. subject pronouns need not be overt in Spanish. the sense of the verb is that of sub- ‘notice’ are odd with predicates such as guapo. Roughly speaking. bastante ‘quite’. improve the sentence. or both types of perception) and the role of modifiers like muy ‘very’. In this regard.

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

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

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

kind. ser + some APs (such as cruel.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. it is not surprising that just predicates of the kind like cruel are the ones that are grammatical in progressive contexts. by virtue of the aspectual underspecification of ser. (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). However. Since the progressive favors subjects that have some control over the predicate. since it would be the aspectual head that licensed the aspectual form. Schmitt (1992) argues that ser incorporates into an aspectual head (the progressive in (80)) and becomes a predicate with internal temporal structure. according to Schmitt. where the copula does not play any role in terms of selection and does not add any semantic aspectual content to the construction. Schmitt continues. In sentences like John built a house. the house has not been built yet. if the reason for the grammaticality of (80) is the incorporation of ser into the . 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. which looks like a circular explanation. (79) María está *estando/siendo simpática Maria is estar/ser-PROG nice “Maria is being nice” Schmitt argues that this contrast is explained by the fact that the progressive cannot range over result states. With the progressive. In this respect. estar predicates cannot. then. there is no result reading available. I will make two brief remarks. a house has been built but in John was building a house. First. it seems that the incorporation of ser into the aspectual head (progressive) is what licenses the progressive. mean. nice) can appear in the progressive.

like all verbs in the perfect. they may just show that estar predicates behave as states. the reason why estar and the perfect are excluded as complements of perception verbs may be because both behave as states.1. it is not clear that tests along the lines of (82) and (83) show that estar designates a result state. 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. 1995). (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.2. and (infinitive) states are not accepted as perception verb complements. (c) Inability of estar to appear as a perception verb complement In relation to its inability to appear in the progressive. However. such as when(ever)-clauses: .” (I will introduce these notions properly in chapter 3). as I already suggested. (See section 2.) (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. (81)) remains unaccounted for.28 Individuals in Time aspectual head. Chapter 4 develops a proposal about the strict correlation between the type of APs and this syntactic behavior. Schmitt also notes that estar. the fact (also noted by Schmitt) that not all adjectival predicates combine with the progressive (cf. is not allowed as a perception verb complement. 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. (82) *Juan vio a María estar descalza Juan saw Maria estar-INF barefoot (83) *Juan vio a María haber construido una casa Juan saw Maria have-INF built a house Although Schmitt does not discuss this point in detail. In other words.

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

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

Another author arguing in similar terms as Luján (1981) and Schmitt (1992) is Fernández Leborans (1999).” As chapter 7 discusses in greater detail. the opposition.4 Summary of Section 2. I argue that the availability of the progressive is related to the internal temporal properties of the construction. Luján conceives of ser as a predicate that applies to an individual during a stretch of time with no beginning or end assumed.” As I understand her proposal. Fernández Leborans considers that in contexts such as the progressive (91) or temporal forms suggesting that the property does not hold any longer (92). to say that ser is allowed because it does not involve internal temporal structure is. are also perfectly grammatical in such scenarios. In sum.” she means that serpredicates are not tied to temporal limits—the reason why she argues such predicates can be described as “durative. Luján (1981) and Schmitt (1992) defend the idea that the distinction between ser and estar should be understood in aspectual terms. in her view. I argue that the oddity (or not) of such aspectual forms is not related to the IL/SL distinction. alludes to the lexical content of the predicates. by “inert with respect to aspect.Individual-Level Predicates 31 parallel to when(ever)-cases. as she explicitly points out (Luján 1981:206). Likewise. which are conveyed by a type of adjectives in a specific way. Although Luján refers to this differentiation as an imperfective/perfective contrast. is founded on aspectual properties.2. at best. ser predicates work as SL predicates. However.2. Whereas estar possesses internal temporal structure. where no . 2.2. Ser predicates are considered parallel to write or admire. which do involve internal temporal structure (estar in her own proposal and many others).2. For this reason. ser lacks any inherent temporality and is described as “inert with respect to aspect. and estar as a predicate that is true of an individual for a delimited period of time. since other predicates.2. I consider any appearance of ser to denote an instance of an IL predicate. I show that the perfective preterit can be licensed under the appropriate contextual conditions. which. 2. just a partial answer. (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. I will not have to propose that when the copula ser combines with them it becomes an SL predicate.2. she argues that the ser/estar distinction can be described as an IL/SL dichotomy.3 Fernández Leborans (1999). In her study on copulative clauses of Spanish.

yielding minimal pairs. known as “inner aspect. I will assume that with estar (SL) predicates. Whereas NP predicates obligatorily combine with ser. 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.” “lexical aspect.” or “aktionsart. That is. The terms in which Luján treats the dichotomy IL/SL refer to the internal temporal properties of the predicates. Fernández Leborans (1999). when a predicate appears in combination with ser. whereas estar involves aspectual properties (it corresponds to a result state of an accomplishment verb). etc. I have concluded that the opposition ser/estar is adequately described in terms of the IL/SL split. lining myself up with Bosque (1990). What I can advance here is the conclusion I will draw. I have considered the native intuitions about the different interpretation of such pairs. as I advanced. ser is underspecified with respect to aspect. which. I undertake this point in chapter 7. rather than states. Schmitt (1992) considers that the difference between ser and estar is aspectual in the sense that. the property is associated to a concrete circumstance (also along the lines of Higginbotham & .2. it can be said. Since the descriptions of the readings fall in line with Carlson’s (1977) distinction IL/SL. It is premature to introduce and develop a discussion about whether the real differentiation between IL and SL concerns inner aspect.). mean. 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.2 In this section I introduced the distribution of copular verbs in Spanish. I have critically reviewed her arguments. I will describe ser-predicates as classificatory. whereas estar predicates are taken as parallel to write a letter. for the time being. 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. More concretely.3 Summary of Section 2.32 Individuals in Time endpoint is involved. others that combine just with ser. that there are some that combine just with estar. a process where a delimited point is involved.” Along similar lines. paying special attention to her discussion on the behavior of ser in combination with certain adjectives (cruel. depending on which verb is used as the leading hint to define the opposition ser/estar. after having introduced a formal description of inner aspect and other temporal domains. Therefore. 2. the subject is categorized as belonging to the class denoted by the predicate. and others that are able to combine with both copular verbs. APs present a more complicated paradigm. Fernández Leborans (1999) argues that the IL/SL distinction is a primary parameter of lexical aspect. which is that the IL/SL distinction is not a matter of inner aspect. based on Escandell-Vidal and Leonetti 2002. Demonte (1999). and Escandell-Vidal and Leonetti (2002). is the topic of chapter 4. Although I will add some remarks in chapter 7.

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

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

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

1995) or its lack thereof. 28 See section 2. can host an argument (a subject) in its specifier. The evidence adduced to by Felser for the presence of an eventive argument is that perception verbs complements have to be “eventive. as in (107). Felser argues for the existence of an eventive argument and for the existence of a functional projection that.1 for a discussion in which opposition (eventive/stative or IL/SL) is at stake. besides hosting the event argument. in the presence of a functional projection (Aspect) and an Event Phrase (in the spirit of Kratzer 1988. I saw John draw a circle b. *I saw John know the answer As for the existence of an extra projection.” rather than stative. Felser presents evidence of a filled specifier position.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. .28 (107) a. 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.

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

the difference between IL and SL copular clauses. Schmitt (1992). for a lexical item whose semantic import has a correlation in the aspectual realm. 1981). In the second place. as I will argue. As a result. that is. dynamicity. dynamic. etc. taking into account the analysis of the temporal domains presented. much in the line of Milsark (1974) and Carlson (1977). I continue with an analysis of their outer aspectual properties (their combination with imperfect. and Chierchia 1995. PP. I consider IL predicates to be classificatory predicates that apply directly to the individual. I propose that the aspectual import I will be dealing with (i. 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. based on the different aspectual properties they show.e. a preposition. section 2.. I will not argue for a plain functional head but. Following Stowell (1978. In particular. aspectual content does not amount to SL-hood. I will develop a hypothesis about Aspect in predicates combining with ser and. perfective.e. I consider that the copular verb takes an SC containing the nominal argument.3 introduced the syntactic analysis of copular constructions I will assume. temporal interpretation) in chapter 6. NP). I defend IL SCs with aspectual content.38 Individuals in Time and SCs with aspectual content are not incompatible.4 Summary of the Chapter This chapter introduced the dichotomy between IL and SL predicates. from Kratzer 1988. such as the distribution of copular verbs in Spanish. I develop this hypothesis in chapter 4. as the minimal pairs of “copula + AP” in Spanish suggest. I therefore discarded those proposals establishing a strict correlation between IL-hood and the permanency of the property in the subject (differing. In essence. subject of the predication. etc. I do not submit the idea that only estar-predicates involve aspectual content. I do not conceive of the IL/SL contrast as a permanent/temporary contrast. strictly speaking. and the predicate (AP. Chapter 7 takes up the debate about the characterization of the SL/IL distinction. 2. unlike Luján (1981). and Fernández Leborans (1999). then.) in the next two chapters.) in chapter 5 and finish by examining their tense properties (i. In this work I defend a finer grained distinction among IL constructions with ser. based on Hale 1984 and Demirdache and Uribe-Etxebarría 2000. I argued that such a distinction is operative in the grammar of natural languages since it captures important contrasts. which contributes activity properties) comes from a lexical item—namely. I will not consider that Aspect is. I start by dealing with their internal temporal characteristics (discussing whether they are stative. Centered on ser + AP cases. Therefore. whereas SL predicates refer to properties that are presented linked to a situation. 1995.. among others). Because my primary concern is copular IL predicates. .

. Such properties tell us whether the event inherently tends . refers to the internal temporal properties of events in sentences. I first introduce the notion of “inner aspect” and the different types of events that can be distinguished according to it. according to a set of tests.Chapter 3 Event Classes and Individual-Level Predicates In chapter 2. I will bring the proofs onto IL predicates. I also emphasized that this should not be taken to mean that IL predicates are stative and permanent predicates. the criteria to distinguish them.1 Inner Aspect and Event Types In this section I deal with two points. Other more specific issues. and the tests to diagnose them. such as where the stem of the contrast resides (in the copular verb or in the adjective itself). Because permanency and stativity are temporal concepts in nature.1. Such a contrasting behavior raises a number of questions and invites reflection on some general issues. their internal temporal properties (inner-aspect properties). Once we have become familiar with the different event types. I will demonstrate that copular adjectival constructions can be classified in two groups. or whether there is more than one (IL) copular verb. Inner aspect has to do with properties such as duration. 3.4—namely. I defend a finer grained distinction among IL predicates based on the different inner-aspect properties they show. Specifically. as activities). whereas another can be justified as dynamic predicates (i. also called “aktionsart” or “situation aspect” (Smith 1991). I then present a survey of the distinct types of predicates traditionally distinguished in the literature. I start by investigating. I will show that not all cases in which the copula ser participates share the same inner-aspect properties. 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. First.e. in this chapter. I will examine the temporal properties of these predicates. culmination. 1993). or delimitation. I described individual-level (IL) predicates as predicates that are said of their subject independently from a given situation. are relegated to the next chapter.1 Inner Aspect “Inner aspect” (Verkuyl 1989. 3. To accomplish this task. One group tests out as states. I present one of them in section 3. as has been widely assumed across the literature. According to classical and well-established aspectual tests. the extent up to which the dichotomy between states and activities is real and grammatically relevant.

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

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

be sick. be tall. knock Table 3.” In (8). collapse. be born. write novels. Whereas the former two clearly refer to processes that need not ever end. Events with duration but no endpoint die. walk around the park be green. However. that is why. Event types Sentences (1)–(8) illustrate the event classification as proposed by Vendler (1957.42 Individuals in Time like (4) works as an accomplishment.1 briefly characterizes each event type and provides some more examples. 1967) (in addition to the class proposed by Smith). with an endpoint but no duration Actions with a culminating point that take duration to be completed. in itself. Table 3. Instantaneous and nonculminating events swim. Achievements Instantaneous events. belong. Verbs involving such a property are called “achievements. As noted earlier. it is not an action in any sense. there is a slight difference between activity-typed events like walk or push a cart and others like build houses or read novels. They lack any kind of internal structure. awaken. explode walk to the beach. has neither duration nor endpoint. the same with a bare plural as an internal argument instead (build houses) behaves as an activity (see especially Tenny 1987. the subject is involved in an action that is instantaneous and lacks an intrinsic endpoint. and it is the habit that is not understood as delimited in time. build houses and read novels are conceived as different building-houses and reading-novels events distributed over time. arrive. Eventualities like this are named “states. we also see that such templates cannot be a pure lexical matter. but. recognize. In (6). each building-a-house or reading-a-novel event is conceivable to reach an endpoint. hate cough. I return to the meaning of habituality in chapter 5. since they can be . read a chapter Accomplishments Activities States Semelfactives No-actions that hold in time but do not take time. know. 1989. push a cart. That is. build a house. love. They can be considered as complex events with internal steps toward an end. Semelfactives are not of great relevance to this work. fall sleep. importantly. sneeze. This predicate does not take time or involve any kind of “process”. Both are conceived as habitual happenings. the latter two refer to a habit that does not have to end. it lacks duration. fall asleep denotes an endpoint. However. 1994 and Pustejovsky 1988). so I will not discuss them further. such events are called “semelfactives” (Smith 1991).” The predicate in (7).1.

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

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

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

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

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

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

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

an achievement. (22) Occurrence in command imperative form a. can only have the second interpretation. the tests in (22)–(25) can be used to diagnose stativity. In the other.2. To the extent that an agent is conceivable only if a proper event (a nonstate) is at hand.50 Individuals in Time The possible interpretations of almost put in contrast accomplishments with other event types. To conclude this survey of aspectual tests. (21) Interpretative entailments from the adverb almost a. 1970). *¡Estate enfermo! Be sick! b. 3. can have two interpretations. the subject did not get to walk or realize.2 gives a more precise notion of “agent. and (21c).5 Agentivity Tests. an accomplishment. as van Voorst (1988) and others point out (see also section 3. that is. Pablo casi caminaba con ocho meses Pablo almost walked when he was eight months c.1. in (22)–(25). ??¡Date cuenta de que estás confundido! Realize that you are wrong! . ¡Pasea! Take a walk! d.5 in this regard). Pablo casi se dio cuenta de que su madre tenía razón Pablo almost realized that his mother was right Sentence (21a). the subject was involved in the process of building up a house but did not finish it. the subject did not even start out the process of building. which are grammatical. too. states and achievements.”) The roles that DPs play in the event and the analysis of the event itself usually overlap. However. As can be appreciated. ¡Construye la casa! Build the house! c. an activity. (21b). yielding ungrammatical results. get differentiated from activities and accomplishments. (Section 3. In one of them. I want to introduce another group detecting the presence of an agentive subject (Ryle 1949. Pablo casi construyó la casa Pablo almost built the house b. Lakoff 1966.

A “+” . the good behavior of activities and accomplishments in agency contexts is due to the property of dynamicity the two of them involve. Pablo forzó a Juan a construir la casa Pablo forced John to build the house c. Lo que hizo fue pasear What he did was walking d. I presented different tests usually employed to diagnose which event type a concrete predicate belongs to. Table 3. *Pablo forzó a Juan a estar enfermo Pablo forced John to be sick b. 3. *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. which pattern together because of the stative feature they share. I also introduced the idea that inner aspect does not depend on the inner properties of the verb itself.2.2 summarizes the tests and the behavior of event types. *Pablo estuvo enfermo deliberadamente Pablo was sick deliberately b.Event Classes and Individual-Level Predicates 51 (23) Pseudocleft with hacer ‘do’ a.1. Pablo construyó la casa deliberadamente Pablo built the house deliberately c. Pablo forzó a Juan a pasear Pablo forced John to walk d. *Pablo se dio cuenta de que estaba confundido deliberadamente Pablo realized that he was wrong deliberately (25) As a complement of force a. This accounts for the ungrammaticality of both states and achievements.1 The first part of this section presented the notion of inner aspect. Pablo paseó deliberadamente Pablo walked deliberately d. where other elements like the object play a determining role. but it is a compositional matter. *Pablo forzó a Juan a darse cuenta de que estaba confundido Pablo forced Juan to realize that he was wrong As discussed further in section 3. In the second part.3 Summary of Section 3. Lo que hizo fue construir la casa What he did was build the house c. *Lo que hizo fue estar enfermo What he did was be sick b.

Tests for event types Among other things.2. and (22)–(25) reveal that activities and accomplishments share an important part of their . (17). a “–” when it cannot. (20). and a “%” if the event type shows a positive behavior in the test but just under a certain interpretation. the results of the tests (9)–(11). The bracketed numbers in the leftmost column refers to the number of the test as they appear in the text above.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. as pointed out above in the description of each test. The symbol “/” means that the test does not apply to the event type for some reason. 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.

agency tests detect eventualities that denote stuff that can be “brought about.” Then. In particular.” and therefore “commanded. In agency. (26) (27) El portazo rompió el cristal The bang broke the glass Juan rompió el cristal Juan broke the glass . (20). agency tests separate states and achievements from activities and accomplishments. when I discuss the aspectual behavior of copular IL predicates. In this section. In the following section. activities and accomplishments may be the most persistent co-patterning of the table.Event Classes and Individual-Level Predicates 53 properties. but just holds. the term “agent” clearly denotes ‘the argument that brings about an event’.2 A Brief Stop at “Agentivity” The set of tests in (22)–(25). Vendler (1967) himself considered states and achievements a natural class. 3.3. I will briefly introduce the relation between agents and event structure. I consider two aspects related to this. volitionality. and control.2. causation is the most basic notion involved in agency contexts.” The cleft with “do” and the test of command imperative detect eventualities that can be “done. 3. As will become clear.” Also. Actually. which gives bad results in contexts with adverbials such a deliberately. may be used to discern properties of eventualities as well. three notions involved: causation. which diagnose agency. stuff that is not performed. To begin this investigation. Further discussion about the tests and about some concrete event types will come in section 3. in principle. there are. I will deal with the different notions that get blended in the term “agent. The results of (17). As mentioned before. at least. Likewise. First. cannot hold with any willfulness on the subject’s behalf. consider the classical minimal pair in (26) and (27). not all controllers involve volition. and (22)–(25) also suggest that states and achievements have characteristics in common. In fact. although volition usually entails control. while volitionality and control are properties that just animate causers can embrace. To the extent that just some types of events can be undertaken and commanded. it is important to distinguish the semantic features it contains and to establish whether there is some feature more basic than the other. agency tests work as event-type tests.1 A Cluster of Notions Although. I present a brief discussion about the agency tests above and about the notion of “agent” that I will observe in this work.

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

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

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

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

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

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

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

which has been very often linked to conscious volition only. This supposes a remark on the notion of agency.2. I will ignore the differences between imperfect and perfective aspect that exist in Spanish. Objects grammaticize telicity. 3. I introduced the issue of the role of agents in the eventive structure itself as “initiators” of the event. In the last part of the section. Agentive subjects can grammaticize an aspectual role—namely. but only animates can involve volition and control. We saw that not all arguments that could be labeled “agents” embrace all such properties. They are relevant to some points. as mentioned earlier. In particular. 16 . control over the ending. through the battery of tests introduced above to diagnose event types. Now. as Ritter and Rosen (2003) put it. and control.2 This section includes remarks about the tests regarding the thematic property of agency. which overlaps with aspectual notions.Event Classes and Individual-Level Predicates 61 (42) La rueda/Juan empezó a girar The wheel/Juan started rotating In sum. 3. the end is inherent. is the lexical expression of IL-hood).3 Aspectual Differences among Individual-Level Predicates This section explores a group of IL predicates from an aspectual perspective. but I leave the discussion about their interaction with aktionsart to chapter 5. I looked briefly at some of the notions involved with the concept “agent”: causation. volition. (which. in the event structure. 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. it is natural to wonder whether the agentive argument plays any role in aspectual properties. This will constitute the first step in the account of such predicates offered here. I examine the behavior of copular cases with ser. In the case of accomplishments. Specifically. given the correspondence between certain types of events and certain types of subjects. not all controller agents involve volition. it seems that three properties can be connected: dynamicity. in the case of activities. 1989. although all volition agents involve control. Tenny (1987. the ending is arbitrary and subject to the subject’s willingness. as argued in chapter 2.3 Summary of Section 3. all must involve causation. 1994) and Pustejovsky (1988) observed the role that objects have in the structure of events. In turn. and (with the considerations just made) control over the beginning of the event. Throughout the tests. the initiator of the event.

As described above. to identify the aspectual properties of IL predicates. The predicates I will test are a classificatory AP (Eskimo) and two qualifying APs (blond and cruel). *Lo que sucedió fue que Juan fue esquimal What happened was that Juan was an Eskimo b.” do not “take place. The eventualities yielding ungrammatical results are taken as stative. Normalmente.1 Aspectual Tests on IL Predicates The aim of this subsection is. Juan es cruel con Pedro Usually Juan is cruel with Pedro Despite all the observations introduced in section 3. Juan es rubio Usually Juan is blond c. *Juan estaba siendo rubio Juan was being blond c.” but just “hold. simply. *Lo que sucedió fue que Juan era rubio What happened was that Juan was blond c. To do that. 3.3. Juan es esquimal Usually Juan is Eskimo b. *Normalmente. In (43)–(45) I reproduce the tests marking for stativity versus eventivity.62 Individuals in Time 3. (43) Occurrence in the progressive form a.” They are considered to be eventualities that lack any bounding and are inherently durative. Juan estaba siendo muy cruel con Pedro Juan was being very cruel with Pedro (44) Pseudocleft with happen (‘take place’) a. *Juan estaba siendo esquimal Juan was being an Eskimo b.1 regarding the present-tense interpretation as habitual (see also chapter 5 for more important .1. states are eventualities that do not “happen. ?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. I will concentrate on copular clauses and test out whether all IL predicates are appropriately grouped in the aspectual set of states.1. as has been widely agreed on in the literature. IL predicates are.2. *Normalmente.1 Events versus States.3. in general. taken to belong to the group of states. I will make use of the tests introduced above in the same order.

Even in the absence of the habitual adverbial.2 Activities and States versus Achievements and Accomplishments. a clear contrast emerges among different types of APs in combination with the copula ser. but the oddness in (46a) calls for further explanation. Consider a scenario like the following: (i) Lo que sucede/pasa es que su primo es esquimal. .2. although it gives ungrammaticality for one of the predicates that above tested out as a state (Eskimo). Actually. In turn. for-adverbials express how long someone has been engaged in a particular activity or a state has held of an individual. These three tests show that not all APs in combination with ser behave the same way. as shown earlier. pseudoclefts with happen are acceptable with states when present tense is involved. which is why they are compatible only with eventualities entailing a delimit point. 3. the reasons for its ungrammaticality are independent from the test itself. As I introduced in chapter 2 (and will discuss further in chapter 6).Event Classes and Individual-Level Predicates 63 remarks). on the other. its acceptability in (46b) is expected. It is interesting to note. Since it appears that the AP in (c) behaves like an event. (45c) gets a habitual interpretation.1. when they hold of an individual. we see that one group comprises APs showing a state nature (a. on the one hand. If. the next step is to find out exactly what kind of event it is. as the contrasts between (46c) and (47c) show. in-adverbials refer to the period of time it takes to complete a certain event. If we compare these results with the ones obtained in the previous section.2. rather than a state.b)17 and the other comprises one kind of AP manifesting eventive properties (c). The test in (46) distinguishes. it seems that these tests consistently separate the APs in two groups: Eskimo and blond. that the period that a stative IL adjectival predicate such as blond holds of the subject can be restricted with an adverbial (46b). as reasoned in the literature. hold for that person’s entire lifetime. that is why she can speak Inuit” 18 As mentioned in section 3.3. predicates like Eskimo refer to properties that. states are durative and for-adverbials are durative adverbials. which is why the limitation of its duration with a for + x time complement sounds weird. It does so as well with IL predicates. and cruel. once again.1. por eso sabe hablar Inuit “What happens is that her cousin is Eskimo. 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. Rather.18 states and activities from accomplishments and achievements. 17 Much like do-pseudoclefts (see footnote 10).

*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). and not accomplishments. since this is only possible with eventive predicates. As shown in section 3. the combination of ser and an AP such as cruel ‘cruel’ looks good.2. (48) Perfect entailment from the progressive form a. *Juan fue esquimal en dos horas Juan was an Eskimo in two hours b. such a predicate shows a nonstate patterning.1. *Juan fue esquimal durante tres semanas Juan was an Eskimo for three weeks b. Once again. 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. which argues in favor of considering the predicate (ser cruel) as an activity. In principle. However.3. yield a perfect entailment from the progressive form.64 Individuals in Time (46) For + x time a. where there is no endpoint. a predicate like ser + AP would be unexpected as a complement of parar de ‘stop’.2. Juan fue rubio durante tres años Juan was blond for three years c. . tests like (49) divide activities and accomplishments from states and achievements.1. *Juan fue rubio en dos horas Juan was blond in two hours c. only activities.3. Juan ha sido cruel con Pedro Juan has been cruel to Pedro 3. (48b) is a licit entailment from (48a).2.1.3 Activities and Accomplishments versus States and Achievements. As mentioned in section 3. Juan está siendo cruel con Pedro Juan is being cruel to Pedro b.

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

like (53c). *Juan terminó de ser rubio Juan finished being blond c. *Juan casi fue esquimal Juan almost was an Eskimo 3. since the unique licit entailment from (53a) is (53b). None of the AP predicates with ser.66 Individuals in Time 3. The next set of contrasts aim at testing whether the process that be cruel can be considered is agentive or not. (52) As a complement of finish a. . confirm be cruel as an activity. The former reading is available with accomplishments.3. ?Juan casi fue cruel con su adversario Juan almost was cruel to his opponent b. they clearly contrast with the other two. as mentioned in section 3. except for cruel.4 Accomplishments versus All the Rest. the latter is with activities. *Juan terminó de ser esquimal. fit in canonical agentive contexts. whose combination with almost is impossible. Juan no empezó a ser cruel con su adversario Juan did not start being cruel to his opponent c. 20 Recall that.4. ?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.20 Although (53a) does not sound extremely natural. 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. Although be cruel predicates do not give perfect results after finish. and (b) that the subject did not even start out the process. which suggests that dynamic properties are somehow involved in these predicates. it contrasts with other APs with ser. which also test accomplishments versus all the other eventualities.1.1. Tests (52) and (53) show some more contrasts consistent with the results from (49). Juan finished being an Eskimo b.5 Agentivity Tests. All tests give the same results. 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.2.1.3.

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

where it is attributed to the copular verb be instead.2. *Juan forzó a Pedro a ser rubio Juan forced Pedro to be blond c. as shown in the set of aspectual tests. I will attribute such an eventive-like behavior to the properties of the AP itself. 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. *Lo que Juan hizo fue ser rubio What Juan did was be blond c.2 Summary of Section 3.3 In this section we have learned two things.e. not all of them can be considered alike. agency is notable.. This amounts to claiming that there is a group of IL cases that involve typical eventive properties. there seems to be a group whose behavior patterns with canonical activities. Among them. taking into account the properties of this group introduced in section 3. *Juan forzó a Pedro a ser esquimal Juan forced Pedro to be Eskimo b. Rather. There seems to be a group of APs that can be taken as Examples (60a. not all IL predicates belong to the category of states. as shown by a number of tests. activities and accomplishments) can be agentive. The leaf was falling down from the tree). 22 Recall that although not all processes are necessary agentive (cf.22 I dedicate detailed attention to these predicates that do not behave as states in chapter 4.1. only processes (i. Juan lamentó haber sido cruel con su adversario Juan regretted to have been cruel to his opponent (61) Pseudocleft with do a. we have learned that.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 lamentó haber sido esquimal Juan regretted to have been an Eskimo b. *Lo que Juan hizo fue ser esquimal What Juan did was be Eskimo b. ?Lo que Juan hizo fue ser cruel con su adversario What Juan did was be cruel to his opponent 3. Second. *Juan lamentó haber sido rubio Juan regretted to have been blond c.68 Individuals in Time (59) As a complement of persuade or force a. differing from other authors’ account.3. First. 21 . among state be-predicates.

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

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

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

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

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

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

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

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

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

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

Event Classes and Individual-Level Predicates


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


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


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

First. This strengthens the hypothesis that the properties exhibited in the dynamic copular clauses are due to the properties of the structure where the adjectives taken by the copula are inserted. attributed the agentive properties displayed by adjectives such as cruel to a different lexical entry of the copula. Specifically. it finds the obstacle that there is no evidence proving the actual existence of such two homophonous copulas. (4) (5) John broke the window on purpose The ball broke the window (*on purpose) . 4. Stowell (1993).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. where the verb is the same but it is observed that agency is a property just of the animate DP subject. Theoretically.1.1 The Two-Copulas Hypothesis Partee (1977) and Dowty (1979). as I discussed in chapter 3. semantically null. of cases such as (3). Second. this is an uneconomical move (and. as has been maintained thus far in the literature. In section 4.84 Individuals in Time presence of a constituent acting as a dynamicity inductor. 4. properties like agency cannot be attributed to the lexical entry of the verb but to the whole configuration. undesirable in itself). the two-copulas hypothesis is unappealing because it multiplies lexical entries with no independent evidence to support it.1 Previous Explanations of Nonstative Copular Clauses In this section I summarize previous proposals concerning the active behavior observed in certain copular clauses. 2004). and empirically. as shown in chapter 3. (1) (2) (3) John was cruel on purpose John acted cruelly John is tall (*on purpose) Among the limitations of this analysis. among others. and Demirdache and Uribe-Etxebarría (2000. therefore. They proposed a theta-assigning active copula for cases such as (1). The last section summarizes the chapter. based on Hale (1984). there are the following three. Recall pairs like (4) and (5). different from the “regular copular be”. rather than to the properties of the copular verb itself. I will defend the preposition heading the PP complement as an actual aspect head. with a meaning close to act (2).

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

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

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

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

the copular verb should be. Ser is completely grammatical.2. to admit there is an eventive argument that is implicit amounts to admitting that (25) is a SL predication.1) regarding the APs that can appear with either copula (cf. necessarily. I take up the discussion related to the interpretation of the “infinitival argument” in Spanish and its status (as argument or adjunct).2. in his account. he argues that the action denoting argument is an overt version of the (Davidsonian) eventive argument. 1995) account. In Spanish. . However. as (26) illustrates. (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. Specifically. in chapter 6 (section 6. According to Stowell.4 he takes it that the presence of such an argument is proper of SL predicates. Note that the contrast between ser and estar in this case falls in line with the contrasts studied in chapter 2 (section 2. 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. 1995). (25) John was cruel to Peter Regarding Stowell’s analysis.1.3). and the contrast between the examples with ser and those with estar (27) can be captured by paraphrases such as the one in (28). that is not case. 4 For details of Kratzer’s (1988.2). but he is not cruel to him usually” Later. (28)). Following Kratzer (1988. see chapter 2 (section 2. which is the lexical marking for SL-hood. and between the dyadic one and stage-level (SL) predication. the event argument represented by the infinitive is implicit in simpler copular cases such as the ones under investigation here. First. estar. I will make two observations. this proposal would entail that in all instances of APs followed by a relational PP.Aspectual Alternations in IL Predicates 89 Stowell (1991) establishes a correspondence between the monadic version of these adjectives and individual-level (IL) predication. I repeat one of the examples here in (29).

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

given their lexical meaning. I will attribute such an active behavior to a particular syntactic structure. dense White. b. kind. and as complements of force or regret). cruel. d. 4. f. Dixon distinguishes different classes of adjectives according to the type of concept expressed. blue.2. where a set of adjectives is conceptually more natural than others. (31) a. brown. wide. g.Aspectual Alternations in IL Predicates 91 ally. horrible Quick. I begin by describing the type of qualifying adjectives giving rise more often to the nonstative patterning. e. c. intelligent.1 Lexical Semantic Classes of Adjectives I will take the basics of the classification of adjectives proposed by Dixon (1977) as a reference. mean. round. heavy. *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 . slow Apt. he distinguishes the following classes of semantic concepts. 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. I will not argue that activity-like properties are due to the intrinsic lexical properties of the adjectives. recent Beautiful. small Light. *Juan estaba siendo alto/bajo Juan was being tall/short b. Dimension Physical property Color and Shape Age Evaluative Velocity Human aptitudes and dispositions Tall. Progressive Form (32) a. capable. That is. new. Concretely. farsighted. old. short. shrewd. cunning. squared Young. 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. stupid.

*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. *Juan era alto/bajo a propósito Juan was tall/short on purpose b. *La mesa era ancha/pequeña a propósito The table was wide/small on purpose (40) *El mueble era ligero/pesado a propósito The piece of furniture was light/heavy on purpose (41) *La mesa era blanca/azul/marrón/redonda/cuadrada a propósito The table was white/blue/brown/round/squared on purpose (42) a.92 Individuals in Time (35) a. *Juan era joven/viejo a propósito Juan was young/old on purpose b. *Juan estaba siendo joven/viejo Juan was being young/old 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 .

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

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

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

(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. or by acting in a certain way. Incidentally. 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. focusing on two aspects: its interpretation and its optionality or obligatoriness. 4. I will show that its syntactic presence correlates with the aspectual patterning of the construction.3. set on fire and bother. One can. offend or regale. I investigate the nature of the PP complement.96 Individuals in Time correlation with their aspectual characteristics.” This captures his intuition that such a DP refers to the individual that gets affected by the (underlying) action undertaken by the subject DP. note that the DP inside the PP has to have a [+animate] noun: (59) Juan fue cruel con el gato/Pedro/*el armario Juan ser-PRET-3SG cruel to the cat/Pedro/the closet Although the concrete action itself is left unspecified (we do not know exactly what Juan did to Pedro). to name just a few. abuse. it is understood that Juan was the agent of an action that had Pedro as its affected argument. for example. whether or not the PP expresses an “affected argument” depends on the concrete type of action undertaken by the subject. Likewise. harass. 4. although it seems intuitively clear that the PP specifies the goal addressed by the subject’s action. and such an action is qualified as cruel. 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”).1 On the Interpretation of the Relational PP Stowell (1991) suggests that the DP inside the relational PP. and discuss all the consequences derivable from this. In this section. can be considered as an “affected goal. 1988. offend someone else either by saying something unpleasant. 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. 1994] and . Thus. in italics in (58). even if we finally concluded that the PP is an “affected argument”.3 Relational Mental Properties: The Relational PP Complement As just described.

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

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

The following examples are from Bosque (1999). one can think that if a person is cruel or kind or mean he has to be cruel. it is quite clear that someone has to be guilty and a supporter of something. In other words. but.Aspectual Alternations in IL Predicates 99 I will start by observing the behavior of other APs that also have prepositional complements. The complement is interpreted as something specific which is not necessary to be repeated for whatever reason: guilty of such and such crime. they cannot be taken as ungrammatical as they appear. the PPs are massively headed by to. Examples (69)–(72) are trickier. we may judge a sentence like (74) as odd. kind and mean to someone else. since all of them are introduced by the preposition con. (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. maybe even like a contradiction. which is the status of the PP complements in the relational MP cases? (73) Juan es cruel/amable/mezquino Juan is cruel/kind/ mean In principle. the complements in these cases are obligatory too. on the other. immune to something and compatible and consistent with something. supporter of such and such proposal and compatible with such other computer features. . distressed or prone become ungrammatical if the PP complement is not explicitly present. although with can appear as well. For example. although their presence can be phonetically null and its content is recovered from the context. Spanish. Now. In English. constructions with adjectives such as eager. On the one hand.

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

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

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

I conclude. (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. for example. if cruel-type APs always involved a PP complement. even in the case where it appears alone. . which leads to the conclusion that it is not always there. such as activities or accomplishments. the predicate should refer to something which can be a matter of subjective opinion. Consider now the following examples from English. there should not be any difference between cruel by itself and “cruel + PP”. Consider the contrasts between (87) and (88). then. one can agree as a matter of opinion about whether someone is cruel to someone else or not. As observed in the examples below. The contrast found in (81) and (82) suggests that the presence of the PP makes a difference.Aspectual Alternations in IL Predicates 103 acabado (86) El debate está/*es the debate estar/ser-PRES-3SG finished “The debate is finished” It seems that in the licensing of the predicates of consider two factors play a role. 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. since. In sum. 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. this does not seem to give us the reason of the contrast in (81) and (82). (87) *Considero a Pedro padre I consider Pedro a father (88) Considero a Pedro un buen padre I consider Pedro a good father However. when the DP is inanimate the relational PP is excluded. where just some infinitival predicates are possible as predicates of consider SCs. In the first place. a PP involving an arbitrary nominal could be syntactically noticed. Surely.

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

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

it must be overt. I have debated the correlations between affectedness (understood as ‘change of state’) and event delimitation (Tenny 1987. Second. I first discussed the interpretation of the PP complements. whereas. it can be understood or inferred from the context if phonetically null. In this respect.3 In this section I have been concerned with different properties of relational MPs. with the PP present they become acceptable. 4. 1989. With some of them. in the be cruel to someone constructions. 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. (105) *Juan fue inteligente a propósito Juan was intelligent on purpose (106) ?Juan fue inteligente con su jefe a propósito Juan was intelligent to his boss on purpose This fact argues in the same direction as the examples analyzed above with [–animate] DPs as subjects of the cruel-type. I first showed a syntactic scenario (relational MPs as . Along the same line as Stowell (1991). which strengths the relationship proposed between the PP and agency. I defended that the DP inside these PPs should be analyzed as a “goal.106 Individuals in Time (104) #Juan es inteligente con la gente en general “Juan is intelligent to people in general’ As a second remark. with other adjectives. Jackendoff 1996). Whereas without the PP. either explicitly or covertly. volitional adverbials are excluded. although this complement does not maintain the same relationship with all adjectives. interestingly. when the PP is added. In this regard. Agentive properties are in direct correlation with the presence of the PP: for the PP to be grammatical. In copular constructions of the type of be cruel to someone. The following sections elaborate on this point. Examples (105) and (106) show that.3. Interestingly. when the PP is added to these adjectives. the subject must be able to involve agentive properties. they can be said to gain agentive properties. Compare the following sentences. the subject is understood as an agent. I do not believe we can talk of a real affected DP. In sum.4 Summary of Section 4. I studied whether the relational PP is always present.” although I discussed the status as “affected” that Stowell also attributes to them. note that. I considered two facts. Finally. 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. aspectual and thematic properties of the constructions are proved to depend on the presence of such a PP.

the relation between the AP and the PP is different. It can be inferred if it is not present with the cruel-type. I showed that these nonagentive APs gain agentive properties as they combine with the relational PP. a [–animate] DP subject is not compatible with the PP and it cannot be understood as an agent. activity properties are due to the presence of the PP. although I also showed that if we create a suitable scenario and add a relational PP to other adjectives. I showed that. behaving as a state. In the copular cases in question. I suggested that the APs that take a relational PP are those referring to MPs. 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. such cases would be unexpected. which correlates with the syntactic presence of an argument (a PP). the issue is not very different from other contrasts noticed in the literature and men- . the matter is reducible to an aspectual alternation affecting the thematic properties of the arguments. which strongly suggests that their properties are different.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. Finally. but the activity behavior ensues with those APs admitting a PP relational complement. 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”. More precisely. and the type of DPs allowed was correspondingly distinct. whereas it must be overt to be understood with APs such as intelligent. I also pointed out that. I briefly discussed the combination of the relational PP with adjectives other than the ones considered “relational” by Stowell (1991). Otherwise. I argued that this aspectual alternation does not emerge with every type of copular predicate. In particular. I took this as additional proof that these adjectives are not inherently relational. I then looked at cases where the subject DPs are [–animate]. interestingly. the cited active properties emerge. the thematic interpretation of the DP subject in each case (stative or dynamic) was different.Aspectual Alternations in IL Predicates 107 predicates in SCs complement of “consider”) where the plain form (the adjective by itself) and the one with the PP complement (adjective + PP) are not fully interchangeable. I will propose that we can have cruel. Whereas a [+animate] DP subject is compatible with the presence of a PP complement and it is interpreted as an agent. on the one hand. all of which are odd with a relational PP. Put in these terms. 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. and the other dynamic. whereas predicates describing properties that can be projected onto an animate entity can appear with a relational PP. and “cruel + PP”. Concretely. In other words. 4. behaving as an activity.

according to which aspect properties are fully encoded in the lexical items. mean.5. In (107) it is a PP which is exerting a crucial aspectual role in making the event an accomplishment. two different entries (stative and dynamic) for each adjective (cruel.4).5 Justifying the Approach 4. by definition.2) the contrast between animate and inanimate DPs in pairs like (109) and (110). From a lexical point of view. it does not allude to the other elements present in the syntactic configuration. This perspective cannot capture any of the other properties (interpretation of the subject.). (107) El delantero avanzó hasta la portería The foremost advanced until the goal (108) a. That is.” Recall in this regard from chapter 3 (section 3.108 Individuals in Time tioned thus far. The pair in (108) from Borer (2005) shows an alternation stative/dynamic which correlates with the [±animate] properties of the DP subject. . kind. we would be forced to assume that each aspectual behavior of adjectival copular cases (stative and nonstative) corresponds to different (but homophonous) lexical items. I begin by discussing which approach to event representation can make sense of the observed correlations among the syntactic presence of a constituent.1 Lexicalist and Logico-Semantic Approaches The observed correlations among the presence of the PP relational complement. syntactic presence of a PP) as correlations since. In the remainder of the chapter I will be concerned with such a contrast in the realm of adjectival copular clauses. However. etc. I argued that such a contrast is grammatically relevant. whose contrast has been debated in the literature (see section 3. 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. 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. both atelic eventualities. María tocó la verja (dos veces) María touched the fence (twice) The alternation here is of states versus activities. Along the lines of Borer (2005) (among others). La pared tocaba la verja The wall touched the fence b. 4. since there are syntactic and semantic facts suggesting its actual substantiation. the thematic interpretation of the subject and particular aspectual properties.

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

5. since. In particular. 1994).” given the fact that both are atelic and do not “culminate. among others. 1989.1 Event Roles.1).1.110 Individuals in Time This approach leaves several questions unanswered. 4.5. correlate with the aspectual outfits of the construction. 1989.2 Syntactic Approaches 4. and van Voorst (1988). this approach does not establish any relationship between the aspectual term (Hold). each entity corresponds to an aspectual notion. these authors emphasized the role of the internal object as an event delimiter. authors such as Verkuyl (1972). Dowty (1979). as captured in the Event Correspondence Rule (116). the notions of origin and termination give the characterization of events (113)–(115) and the temporal points identifying the beginning and the end correspond to physical entities (the arguments): that is. showed the relevance of arguments in the inner aspect properties of a sentence. there is no obvious way to distinguish between the two constructions in point (stative and dynamic copular clauses). and the presence or absence of other elements in the sentence (the PP complement in our cases) that. which opens the possibility of establishing a direct relationship between the thematic properties of the arguments present in a sentence. This perspective does not make any direct prediction about the correlation between the aspectual properties of a VP and the thematic properties of the DPs. Tenny (1987. arguably. (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 . Tenny (1987. both would contain the term “hold. its syntactic position and its role in the determination of the aspectual properties of the sentence.” Second. the dynamic or stative properties. in principle. 1994) and van Voorst (1988) elaborate both on the observation that arguments grammaticize the points that give the temporal contour of an event.2. First. As van Voorst puts it. As discussed earlier (see section 3.

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

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

a telic one. Compare: (124) John drank beer (125) John drank two beers A bare mass noun gives an atelic interpretation. which I take as a frame of reference for the remaining discussion. whose core is constituted by aspectual projections. A [+quantity] object. a verb projects an unordered number of arguments whose interpretation depends just on the functional projection they move to. 2000. that gives the type of aspectual interpretation.15 In what follows. Mostly along the lines of Verkuyl (1972. 1999. 2001a. events are not lexically specified for any type of aktionsart. 16 The work by Borer has appeared progressively in a number of articles. aspectual properties can be recast into quantity properties. all the relevant grammatical processes get a unified account. Ritter and Rosen (1996. 1998. When considering that functional projections are in charge of event-roles too. since it presents the latest versions of her ideas and proposals. 2000) argue that functional projections responsible for Case processes (Agreement projections) assign the event-roles of delimitation and initiation. where functional projections are in charge of Case assignment and agreement processes. 17 This was introduced in chapter 3 (section 3. In other words. In turn. events are not states or accomplishments from the lexicon. 1989. whereas a DP headed by a numeral. More accurately.17 Borer proposes that the core feature of aspectual distinctions is [±quantity]. Borer proposes that the core functional projection of the clause is Aspect. Likewise. 1998. argument structure is computed on the basis of the syntactic structure. Thus.14. As mentioned in chapter 3. see Kratzer 2004 and Kiparsky 1998. 2000). Borer (2005) elaborates on the intimate relationship between quantification and telicity or lack thereof. I introduce some points of the work by Borer (2005). 2005) defends the idea that Case-checking processes are satisfied through aspectual projections. Assuming the parallel distinctions between mass/count nouns and atelic/telic verb interpretation. Borer (1994. 2001b).16 For Borer. 15 For further discussion about the relationship between case and event structrure. 14 . This head has the property of “quantity” (AspQMAX or Quantity). Event-roles are not assigned by a particular head but are simply derived as an entailment from the aktionsart of the whole event.Aspectual Alternations in IL Predicates 113 and functional structure is interpreted as event structure. That is. that is. she argues that aspectual properties of a determined event All the authors cited make proposals in a similar spirit in this regard. 2001. These works are framed in the Minimalist Program (Chomsky 1995.4). an object that can be marked as [–quantity] correlates with atelicity. with telicity. in turn. It is the syntactic structure. I will refer to her book (Borer 2005). the quantification properties of the DP direct object make a direct difference regarding telicity.

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

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

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

(b) When an animate DP subject and the PP are present. Let us begin by discussing point (b) concerning the aspectual contribution of the preposition. I will address two theoretical points. As I advanced. it has become clear that the account for the cruel-type constructions has to capture. I argue.3. I propose that the source of dynamicity and its related properties (agency etc. 2004). In support of this proposal.) in the active cruel-type constructions is a projection (the PP) that combines with the “basic” structure of the AP. all of them very likely emerging from animacy itself: the DP has to be able to behave as a controller agent. Stowell (1993). but it correlates with the existence of the PP relational complement. As I intimated before. I propose that the dynamic properties become a part of the construction through the preposition heading the PP. the construction has characteristics proper of processes.6. That is. To capture (a)—that is. I will take such a correlation seriously and I will propose that the source of the aspectual properties (specifically.1 Prepositions as Aspect Encoders I have shown that cruel-type APs. the aspectual alternation (state/activity) is not unpredictable at all.6. . 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. I will argue that the stative version of the cruel-type APs corresponds to their most “basic” structure. 2000. which gains dynamic properties from the preposition heading the PP denoting the DP goal. I will discuss the syntactic treatment of the opposition state/activity. behave as activities.Aspectual Alternations in IL Predicates 117 4. the PP makes them a dynamic predicate. Point (a) is discussed in section 4. the following two properties: (a) When there is a relational PP. among others. From the description in the previous sections. at least. since their intervention in the event structure is not obvious (given that they are oblique arguments). As I develop the proposal. based on Hale (1984). as soon as they combine with a (relational) PP. and I will also be concerned with the aspectual treatment of the arguments expressed via a PP. and Demirdache and UribeEtxebarría (1997. the dynamic aspectual behavior) actually resides in the preposition itself. that prepositions can be conceived of as heads encoding aspect properties. the DP subject has to involve a particular set of properties. To capture (b).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. 4.

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

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

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

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

In (149b) the TT is located “after” the span of time the event of cradling takes place. which establish a temporal ordering between time-denoting arguments. In (149). 2000. the time the speaker is talking about is a concrete time. Some clarifying examples appear below. the kind of the relations being either of central coincidence (progressive aspect) or noncentral coincidence (perfect and prospective aspect). Maria had cradled the baby [----------------]///////// c. 1996) proposal that the semantics of tenses can be described in analogy to prepositions like before or after. Based on Klein’s (1994. 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. Demirdache and Uribe-Etxebarría (1997.122 Individuals in Time Demirdache and Uribe-Etxebarría take as a point of departure Stowell’s (1993.” The discontinuous line represents the event and the slashes the TT. namely. 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. 2000) proposal inspired in Hale 198424 and Stowell 1993. Demirdache and Uribe-Etxebarría argue that Aspect can be conceived in parallel terms. Following Reichenbach (1947). when he entered the room. 2004) argue that Aspect can be conceived as a head establishing a temporal ordering relationship between two times. When I entered the room. Perfect AspP 2 c. and in (149c). (149) a. When I entered the room. 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”). he characterizes tense distinctions as relations of priority. the TT is captured “before. When I entered the room. Maria was cradling the baby [----------////////------------b. It is conceived as “contained” in the span of time the event of cradling takes place in (149a). Progressive AspP 2 b. (150) a. This time is the Topic Time (TT).

paying special attention to the APs I am taking care of in this chapter. 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: . Preposition con establishes a central coincidence relation between the figure and the ground (in the terms introduced before). I am going to argue that the preposition introducing the (affected) goal carries inner aspectual properties affecting the aspectual nature of the whole construction.2. con-PPs can also appear with motion verbs. denoting the goal. ¡corre con tu papa! Pablito. This preposition is associated with stative locational meanings. 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. that prepositions have aspectual meaning).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.Aspectual Alternations in IL Predicates 123 Assuming that aspectual meaning is reducible to prepositions (and. the preposition con. it is conceivable that prepositions can also convey inner aspect properties. 4. and.6. con ‘with’.6. as seen in the examples above. 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. conversely. and assuming with Demirdache and Uribe-Etxebarría that all temporal relations can be reduced to the same set of primitives. The preposition introducing the PP in Spanish is. 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. I propose that this is the case. like in examples such as the following: (153) Pablito. 4.1 Con: A Locative and a Directional Preposition.

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

Aspectual Alternations in IL Predicates


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.



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. 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


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).



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


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. 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.

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

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 . The two of them are related by a preposition meaning ‘toward’. As a directional preposition.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. They are. the DP subject can be considered as the figure and the DP inside the PP as the ground. we saw that Demirdache and Uribe-Etxebarría (1997. therefore. The two terms are put under a specific relationship according to the meaning of the preposition. it designates the movement of a figure toward a place or ground. and (183) represents it graphically.) (182) (183) TOWARD: preposition that establishes a centripetal noncentral coincidence relation between the figure and the ground. (182) summarizes the description of the preposition.6. I consider that the relation between the DP subject Juan and the goal Pedro is not very different from the one existing between Juan and Pedro in (181). 2004) proposed that centripetal noncentral prepositions translate as “before” in the temporal domain. associated with ‘future’ in the realm of Tense and with ‘prospective’ in the domain of outer aspect. (181) Juan went to Pedro In the terms of figure and ground presented before. (The bullet stands for the figure and the little house for the ground. it locates the figure in its trajectory toward the ground. …● ⌂ In section 4. the discontinuous dotted line for the trajectory. that is.1. 2000.

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

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

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

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

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

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. it had a directional meaning (‘to’). The structure I have proposed for deadjectival verbs exploit the participation of a preposition to account for their dynamic properties. Both cases are similar regarding the elements that participate in them: a preposition. since it is by virtue of a preposition that an adjectival construction becomes dynamic. the preposition had a locative (stative) meaning (English ‘in’). the paraphrases of both being ‘to’ (The Latin preposition in could be followed by a nominal either in ablative case. I suggest that such dynamic properties come from the semantics conveyed by the prepositions that participate in their formation. If followed by a nominal in ablative case. (v) en–negre–c–(er) ? ? ? ? P A v (infinitive-ending) ? ? ≈? con cruel ser . black. they come from the Latin “in” and “ad”. an adjective and a verbal piece. 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. The prepositions participating in the cases above are en (en-negreció) and a (a-largó). which renders parallel deadjectival verbs and cruel-type cases. Etymologically. contra Borer’s (2005) idea that statives are due to the presence of specialized functional structure (a sort of a “stativizer”). 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. or in accusative case. red) and refer to dynamic processes. 4. a `centripetal noncentral coincidence preposition. That is. whereas if followed by an accusative.Aspectual Alternations in IL Predicates 137 structure to emerge.

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

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

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

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

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

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

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

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


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

Lillooet Salish). posteriority.e. I turn my attention to adjectival IL predicates and show the different interpretations available with each type of viewpoint mentioned.. I show that states are compatible even with the perfective. posteriority. At that point. based on Zagona 1990. Therefore. by the same token. therefore. In particular. as Tense and Aspect are considered ordering predicates. It turns out that this cannot be described in terms of presence or absence of quantity. inner aspect can be as well. atelic). aspect properties can serve as indicators of temporal properties (a perfect action is computed as finished and. However.148 Individuals in Time In section 5.1 Tense and Aspect as Ordering Predicates In this section I deal in greater detail with the conception of Tense and Aspect as ordering predicates. which I mentioned briefly in chapter 4 (section 4. I show that. specifically. like to. this being why the eventuality is computed as “dynamic and unbounded” (i.1. as something that took place in the past).1).. 2 .3 Authors such as Zagona (1990) and Stowell (1993. I argued that dynamicity properties (that activities involve) can be described through prepositions. I argued that inner aspectual properties could also be conceived in ordering terms. and Eventuality Time). and simultaneity). 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. 5. in Spanish at least. inner-aspect characterization may work as an aspect indicator (an accomplishment can be taken as a perfect action. Since the line of this work is not formal semantics I do not discuss them here. I tried to show that all aspectual properties could be reduced to the same set of primitives. the progressive seems to impose restrictions on the type of eventuality. and inclusion are semantically defined as partial order relations fulfilling a number of properties. but in dynamicity versus lack of dynamicity. in some languages (e.g.1 Tense. and Wall 1993. that establish centripetal noncentral coincidence relations between a figure and a ground. Consistent with the conclusion that viewpoints do not interfere with quantity properties. Stowell (1993.5. For an introduction of such notions.2 One of the facts I alluded to was that. Speech Time. 3 Anteriority. Reichenbach gave a formalization of tenses in natural languages based on three intervals (Reference Time. 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. ter Meulen. 5. for example) and. The idea is depicted in (1). the interested reader is referred to Partee. These predicates translate into the temporal domain as ‘before’. and the possible relationships that can be established among them (anteriority.6.

(3) a. Tense takes two time-denoting arguments. The novelty of this approach is that temporal arguments are represented in the syntax. The ZP is conceived as a temporal quantifier phrase.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. whose structure and working is analogous to the structure of DPs. Elaborating on this idea. Stowell (1993.4 The external argument corresponds to the RT and the internal argument to the Eventuality Time (ET). as Enç (1987) had already noted. 1996) argues for a more complex structure (2). called Zeit-Phrases (ZP). as (3) schematizes. (2) TP 3 ZP (PRO) T′ 2 T ZP 2 Zi VP 2 ei VP 2 Subj V′ 2 V … In Stowell’s proposal. . which means the interference of syntactic constructions on temporal interpretation can be directly accounted for. DP 2 Di NP 2 NP ei b. ZP 2 Zi VP 2 ei VP 4 From the German Zeit ‘time’.

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

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

I call the interval the sentence is about (or “the interval the speaker talks about”) “Topic Time” (TT). as Klein (1994) pointed out: Tense does not order the ET with respect to RT. the speaker is asserting that María has been washing the car some time in her life. the obvious question is: what tool picks out the interval the sentence makes an assertion about? Following Smith (1991) and Klein (1994). According to (9). when someone utters (8). (10) seems the most accurate paraphrase of (8). at that time. It is clear that it seems desirable to establish that the interval located in time is not the whole span of time the event is taking place. (10) wants to capture a different assertion: the speaker has a concrete time in his mind and he is asserting that. The observation that Tense can refer to a concrete interval traces back to Partee (1973. Smith (1991) makes an insightful comparison between aspect and the lens of a camera. María was involved in the task of washing the car. The lens of a camera. in principle. 1984b) and was also noted in Stowell’s (1993.1. but the interval the speaker is referring to. The former would correspond with readings of the sort in (10). whereas the latter with those along the lines of (9). However. (10) There is a specific time at which Maria was involved in the task of washing the car. It orders the time the speaker talks about with respect to a time of reference. That time is in the past. Following Klein (1994). Aspect as an Ordering Predicate The temporal interpretation of a sentence like (8) can.152 Individuals in Time as an ordering predicate. That is. Depending on the focus . 1996) work. I assume that the element that does this job is Aspect. Now. the total interval during which María has been washing the car is left in the dark. 5. be paraphrased as (9) or (10). This leads to a big conclusion.2 The Internal Argument of Tense: The External Argument of Aspect. The analogy between DPs and ZPs enables Stowell to distinguish between “definite” and “indefinite” (with just pure existential force) ZPs. In (8). (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. 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. makes them visible for the receiver to be able to catch them in a photograph. Between the two. he intends (10) rather than (9). for example. by focusing objects.

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

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

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

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

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

and second. In this section I am chiefly concerned with the perfective. The values of such a quantifier. what the relationship is between the number of occasions alluded to by each viewpoint and the quantity nature of predicates (that is. I will consider two points—first. how many occasions each viewpoint alludes to. the habitual imperfect. The continuous imperfect does not count the number of occasions a determined predicate takes place.5. The number of occasions the progressive and the perfective form allude to is just |1|. inner aspect. as in (26). In this latter regard.1. here in section 5.2. who observed that the difference between the progressive and the habitual rests on the number of occasions they refer to. a quantificational component. As I will repeat later in section 5. Q<occ> can also have an existential value. giving rise to another interpretation. and the progressive. However. whose head is a quantifier over occasions (Q<occ>). (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. as the reader may have conjectured already. It appears . I address these points in turn.3. besides ordering. the three forms differ in the number of occasions they allude to.158 Individuals in Time of Quantity. can be either |1| or |>1|.2 and next in section 5. (25) ----------------------x---------------------(26) ------x------x--------x-------x---------x-In the spirit of Verkuyl (1999).1 Quantifying over Occasions As just mentioned. the imperfect in (24) refers to a plural number of occasions that can be described as greater than one (|>1|). Let me start by considering sentences (22)–(24). besides the ordering component. the nonhabitual or “continuous” imperfect. also contain information about the number of occasions an eventuality occurs. aspectual viewpoints. 5. I propose that aspectual viewpoints involve. which I analyze as complex phrases containing a quantifier over occasions and ordering information. as shown in (25). However. the functional projection which makes the eventuality heterogeneous and whose absence leaves it homogeneous.

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

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

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

and. several factors seem to play a role. 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. Compare (i) to (ii). the sentence is. simply. nor an account for these cases. I would like to note that in the iterative interpretation with the perfective and for-adverbials. (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. refers to a plural number of occasions. (iii) Leyó el periódico durante dos horas he-read-PERF-PRET-3SG the paper for two hours . 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. their compatibility with for-adverbials gets explained. as she conceives habituals to be. 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 this interpretation is much more difficult to get. the point I want to make with (27) is that. as Tim Stowell (p. in and of itself. whereas the imperfect habitual. For example. 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). the iterative interpretation seems possible when the DP object can be understood as different instances of the same object description. there is no need to interpret that the paper has been read completely or the sonata has been played entirely. where.) points out. Once accomplishments are states. it is interpreted as a progressive (ii). Consider (38).15 the perfective and the progressive can never denote a plural number of times. One can be reading just sections of the paper and playing just parts of the sonata. that cake is not physically the same cake year after year.c.162 Individuals in Time Regarding the number of occasions that the different viewpoints can allude to. this does not mean that the number of occasions that an eventuality in the perfective takes place cannot be modified by an adverbial. where the predicate is understood as iterated: (i) Hice ese pastel durante años (como postre de navidad) I-made that cake for years (as a Christmas dessert) De Swart (1998) accounts for these cases by arguing for the existence of a coercer operator. In (i). as a consequence. However. which turns the event (make the cake) into a state.16 Likewise.

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

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

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

rather. Juan fumaba. but. 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. to my understanding. 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. contextual information participates in establishing one of the members of the proportion the quantifier refers to. this is not always the case. Juan used to smoke. we can paraphrase John smokes by using ‘John is a smoker’. the semantic intuition about examples like (47). (49) En las fiestas de la empresa había marihuana. this does not fully capture. despite the fact that. Consider (49) and (50) as samples of other possible cases. . at the parties of the company was marijuana Juan smoke-IMPF-PRET-3SG “At the company parties there used to be marijuana around. which give us quantities of individuals. in this case.2. 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. 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.” (50) En verano los niños jugaban al tenis en el jardín. in principle. then. However. A certain event can be described as habitual if the cardinality of occasions that a subject performs it (|(A ∩ B)|) is asymptotically equal to the cardinality of times the event takes place by a contextual parameter (c). As before with many. the number of occasions that people smoke (in general). As I mentioned in section 5.166 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. rather than discussing the semantics of the present or imperfect preterit. the contextual parameter captures the average ratio relevant in each case.1. As shown. my main point in this section is to describe habituality.2. The cited contextual parameter would capture. in a simple case like (47). aspectual quantifiers give us the number of instances of a particular event. In any event. In principle.

Outer Aspect and Individual-Level Predicates


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.


Individuals in Time


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



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.) 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


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.


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. 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


(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

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

Outer Aspect and Individual-Level Predicates 173 found. I will argue this is not the case for Spanish since Quantity co-occurs with imperfect-habitual forms. In He was dead. I am going to insist on it with the aim of discussing the . If. and imperfect nonhabitual or continuous) and Quantity. which suggests that the quantification over occasions and the occasions designated by the TT are not the same. there are no occasions to enumerate or calculate. 5. in habitual cases. (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. The TT of the sentence (the time of “finding John in the bathtub”).” That is. as desirable. none of them can be taken as the realization of Quantity. that the number of times an event takes place and the TT are two different notions that should be separated in the formal representation. imperfect) are aspectual operators that can modify the inner aspect properties of eventualities. The representation of this would be like (81). The Spanish data show that these aspectual viewpoint forms contribute information other than inner aspect. Depraetere 1995. which does not refer to a particular number of occasions. therefore. as has been proposed for the perfective in Slavic languages (Filip 2000. we treat all aspect forms and cases uniformly. there is no correspondence between the quantificational properties involved in the imperfect form and those of the TT. That is. In particular. imperfect habitual. progressive. we do not have several TTs. however. I argue against the perspective that viewpoint values (perfective. I conclude. and he asserts that this time is included in the time at which John is dead. and nonquantity predicates are compatible with the perfective. can legitimately count as “one occasion. among others). Although this has already been noted by several authors (Delfitto & Bertinetto 1995.3 Inner and Outer Aspect In this section I discuss the relationship between aspectual viewpoints (perfective. Borer 2005). but that the plural quantification over occasions comes from a different component inside Aspect. we can claim that.

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

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

it does not mean that he has been walking at every single moment of that interval. if we say John had that car from 1974 to 1985.176 Individuals in Time 5. In the same vein as before. draw a conclusion in the same direction. Contrary to Mourelatos (1978). Piñón (1995). and others. 24 23 . I will argue that. he has had that car at every single moment between 1974 and 1985.4). (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. 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. a perfective form does not make a homogeneous predicate heterogeneous.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.2 Perfective Homogeneous Predications I turn my attention now to homogeneous predicates appearing in perfective form. Bach (1986).” as mentioned in the previous section. sentences such as these show that the presence of the perfective viewpoint does not contribute telicity. de Swart (1998). In contrast. (95) *Pablo nadó en una hora Pablo swim-PRET-PERF-3SG in an hour “Pablo swam in an hour” Other authors. at least in Spanish. such as Bertinetto (2000). a perfective form does not interfere with the quantity nature of the predicate. I argue that a perfective form does not trigger telicity effects on predicates.3. With the perfective. The meaning of these sentences is that Pablo was once involved in the task of swimming (91) and walking (92). As discussed back in chapter 3 (section 3. who conceive the perfective as a modifier that makes an eventuality telic. activities are not 100 percent homogeneous at a micro-level. the TT is located ‘after’ the ET. unlike states. In other words. However. If we say John walked from 2 to 3. which leads to the inference of “completeness” and “boundedness.24 as subinterval and additivity proofs show.

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

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

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

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

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

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

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

2. 30 For the reasons why the progressive is excluded with Eskimo and blond. when a complement that can act as a “distributee” is added (125). rather than the habitual (cf. see section 5. they acquire activity-like properties. although the progressive form is correct with cruel.3. as noted in section 5. in Spanish. (119) and (120)). which. the progressive paraphrase of the imperfect is more natural in these contexts. as noted above. 29 . this is just marginally available with be cruel (cf. (122)). 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. However. Likewise.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.184 Individuals in Time It is worth noticing that. make the habitual reading available. accordingly. the progressive is not a paraphrase saliently available from the imperfect form.5.1.30 Finally. whereas activities in imperfect (117) can also have a progressive paraphrase (118). when a relational complement is added (123). a habitual interpretation emerges. the preferred interpretation of adjectives denoting intellectual aptitudes (121) is the continuous reading.

Juan era in his reincarnations Juan ser-IMPF-PRET-3SG Chinese habitually “In his reincarnations. nonstative IL (128). and any eventive predicate. the habitual viewpoint can be said to be compatible with any kind of predicate—stative IL (127). or achievements (132). activities (130).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. stative SL (129). the typo) are subject to referential variability for habitual interpretation to be possible. . accomplishments (131). chino habitualmente (127) En sus reencarnaciones. 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 objects (the house.

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

the progressive in English works differently in some contexts. consistent with the description of the perfective above.31 The progressive viewpoint does not seem to work like the imperfect and the perfective. states hold in time but do not take time.5. in the sense that the progressive is not compatible with every kind of predicate. SL (141) and IL (142) statives are excluded with the progressive. 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). according to which it is compatible with basically any type of predicate.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. .3 The Progressive and Adjectival IL Predicates I now turn my attention to the progressive viewpoint in Spanish. As can be seen. 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. it is also possible to have the stative version of these adjectives together with the perfective. 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. and therefore reject the expression of progression in time. the progressive is possible with other class of homogeneous predicates. 31 As mentioned in chapter 3.

progression in time is not possible for those eventualities that are “dense”—that is. In fact. 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. it is difficult to find an explanation in more technical terms. Some authors.1) for the introduction of this concept.” Dynamic eventualities progress in time. In some sense. as a result. nevertheless.33 Landman (1992). See also chapter 6. unexplained under this view. a third point can be established. a sickness. followed by Bertinetto (2000). footnote 4. However.188 Individuals in Time The property usually invoked to explain the differences regarding progression (advancement) in time is “dynamism. for example. nondynamic eventualities do not. 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. and. examples such as the following. can be argued to lack any input of energy. However. that is. if we can point at it and say ‘it’s the same event in a further stage of development. 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. nevertheless. since estar enfermo ‘be sick’ and estar preocupado ‘be worried’ are states and. strictly speaking. argue that progression in time is not possible for those eventualities lacking internal development. In a similar vein. they are excluded in the progressive form. Aside from the intuitive understanding of the correlation between dynamism and progression in time.32. an event is a stage of another event if the second can be regarded as a more developed version of the first. An eventuality can appear in the progressive if it is divisible in stages. does not allow us to use the progressive. where the verb empezar ‘start’ refers to the onset of the predicate remain.”34 alludes to “temporal stages” to describe the distribution of the progressive. According to Landman. eventualities with such a temporal structure that between every two temporal points. . we can distinguish different stages in. such as Landman (1991). which. accordingly. as the ungrammaticality of cases like *Juan estaba estando enfermo ‘Juan was being sick’ shows. The input of energy correlates with the onset of the predicate. According to these authors. 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. based on Carlson’s (1977) notion of “stage.

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

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

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

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

which has consequences for their temporal interpretation. The last section summarizes the conclusions. IL predicates offer different inner-aspect properties.5 discusses the nature of the TT in greater detail. As I have shown.3 presents the discussion of lifetime effects and describes the properties the predicate has to involve to trigger them. As mentioned in chapter 2. it is not the case that Tense has the same interpretive impact on all individual-level (IL) sentences. Throughout the discussion I will be chiefly concerned with the interpretation of IL sentences in the past tense. . in particular. the domain of Tense. In section 6. and future) have on copular sentences with adjectival predicates. 1995). Section 6. This chapter is organized as follows. formerly discussed by Kratzer (1988. by exploring the interpretation of IL predicates in compound sentences. in 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. As it appears. These two features suppose a difference with respect to previous accounts of IL predicates in general and. as a consequence. 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.4 I show how contextual factors intervene in the rising of this temporal interpretation. since this tense form gives rise to one of the most important interpretive effects with IL predicates—namely. to previous accounts of the temporal interpretation of IL predicates. although it is not only the type of predicate that counts. Section 6.2. In the first section I introduce one previous account for the particularities of the temporal interpretation of IL predicates. I propose a division among (adjectival) IL predicates that correlates with different temporal interpretations. Likewise. structurally. I will be concerned with the interpretive contribution that tenses (present.Chapter 6 Tense and Individual-Level Predicates In this chapter I study another temporal realm—namely. it is the content of the Topic Time (TT) that ultimately decides the temporal interpretation of the sentence. lifetime effects refer to the interpretation whereby the referent of the DP subject can be interpreted as ‘no longer alive’. past. In this regard. Second. First. those dubbed “lifetime effects” (Musan 1995). I assume tenses are syntactically represented in the node of Tense. is located higher than the Aspect node. which. The two main points I would like to make here are the following. it does not influence the ultimate temporal interpretation of the sentence. I will make special reference to the work by Musan (1995.

a spatiotemporal or eventive (<e>) argument (Davidson 1967). SL predicates involve an extra semantic operation.1. The configurational difference between IL and SL predicates can be depicted roughly as in (1). see Anderson 1973. As said before in chapter 2 (section 2. 1995) is the first in establishing an explicit correlation between the temporal interpretation later dubbed “lifetime effects” and IL predicates. That is.1 I will briefly summarize her line of reasoning and show that there are cases it cannot account for. . as Diesing 1992 proposes. just aims to capture Kratzer’s idea on the location of the external argument with respect to the maximal projection of the predicate. IL predicates IP 1 I′ 1 I VP 1 θ-subj Regarding temporal interpretation.2). she proposed that IL predicates applied directly to the subject. I will mention two points of this proposal. in the same vein as Carlson (1977).”2 Specifically. 3 As I also said in chapter 2.3 (1) a. 1995) Kratzer (1988.1). SL predicates IP 1 I′ 1 I VP 1 <e> b. whereas permanent ones lack such an argument. Kratzer proposes that the spatiotemporal argument is syntactically represented. the representation in (1).194 Individuals in Time 6.1. the “realization function. Kratzer proposes that the differentiation between SL and IL predicates resides in their argument structure. For Carlson (1977). Furthermore. 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. putting in direct correlation the presence of a spatiotemporal 1 2 For an earlier brief discussion on this phenomenon. First.1 Temporal Interpretation as a Consequence of Argument Structure: Kratzer (1988. she argues that the eventive argument is the external argument of (SL) predicates.” Likewise. whereas SL predicates needed something “extra. Kratzer argues that SL predicates involve an extra argument in their argument structure—namely.” See chapter 2 (section 2. In particular. This author assumes that the factor that draws the line between the two types of predicates is “being temporary” versus “being permanent. temporariness and permanency is encoded in the argument structure of predicates: temporary predicates involve a spatiotemporal argument.

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

perfective viewpoint does not trigger lifetime effects. In examples like (7). On the other hand. That is. despite being parallel to the one mentioned by Kratzer Henry was French. the past tense does not seem to locate in the past either the situation of his being from California or his lifetime. Consider the following examples as an illustration. 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. In section 6. Intuitively. lifetime effects arise more easily when the predicate is just temporally limited .3 and section 6. On the one hand. predicts that the temporal reading of lifetime effects always arises. The reading disappears. Harry and I arrived in the USA. The second point I want to mention is that the lifetime reading does not arise with any type of IL predicate. First. the past tense of Harry was from California refers back to the moment at which the speaker and Harry arrived in the USA. 1997) noticed. so he did not have to go through the immigration process. In the next sections I will argue that the reasons for this differ. I advanced that equating permanency to IL-hood is not accurate. However. these interpretive effects can be easily neutralized. (7) That day. a purely syntactic approach. In chapter 2. based on the argument structure. which is why they do not appear in (8). Throughout this work I have suggested that not all IL predicates can be considered alike.4. Harry was from California. the second part—that is. I give an account for the way contextual factors can contribute to block the lifetime reading. 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. but in none of them is a lifetime reading available. as Musan (1995. Harry was from California— does not trigger a reading like ‘he is no longer alive’. such as hers.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. In chapters 3 and 4 I focused on the inner aspectual differences among them. Consider (7) as an example.

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


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


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


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


(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

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.


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

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

As I intimated in chapter 2.204 Individuals in Time To the contrary. (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. 6 . when the copular verb is the SL estar. like that in (i). (i) Al ser tan tarde. 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. this discussion at least shows the nontemporal nature of these clauses. it is necessary to point out that this is not the only meaning the al + infinitive clause has with estar. al + infinitive clauses are interpreted as temporal as discussed above or causal clauses. It also has an interpretation very close to the one exhibited with ser.6 Although I cannot investigate further the status and meaning of al + infinitive clauses with ser here. no quise llamarte In being so late. Al acompañarme a casa. which would be along the lines of a causal adjunct. With estar. 7 The meaning of the al + infinitive clause in these cases is not easy to capture. whose paraphrase appears in (ii). it is typical of estar. According to Hernanz (1999). 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. Pedro estuvo muy amable In walking-me home. 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.7 However. the al + infinitive clauses can behave as true temporal adjuncts. I did not want to ring you up (ii) No quise llamar porque era muy tarde I did not want to ring you up because it was very late However.

The 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. in their lexical entry. Second. Musan (1995) does. I want to briefly enumerate the conditions that are necessary (not sufficient) for the lifetime reading to appear. and (c) the past form should be in the imperfect form. as. Differing from Kratzer (1988. can neutralize the lifetime effects. 6.2 I have made two main points in this 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. I do not consider that those predicates encode. Musan (1995. literally. 1995). That is. in the line of Musan (1995. necessarily produce an inaccurate overgeneralization. Ph.2.g. e. such as the presence of another past tense around. In support of the first claim. according to which the temporal interpretation known as lifetime effects derives from the argument structure of IL predicates. nevertheless.3.). In the following section. for lexical reasons.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. these three conditions have to be met: (a) the predicate has to be an IL lifetime predicate. I am going to argue. More accurately. I argue that lifetime effects are a salient option in those cases where the predicate. As already mentioned. for example. 1997).4 Summary of Section 6. First. do not have to be permanent properties. I argued that those accounts. (b) the predicate has to appear in the past form.D. 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. I focus on another restriction for this temporal interpretation. according to independent grounds (such as their combination with a particular lexical choice of the copula in languages like Spanish). compare the following sentences: . For the interpretation of the DP subject as ‘no longer alive’ to be an option (in languages like Spanish).. since there is a large number of predicates that. equating IL-hood to property stability is inadequate. 6. 1997) observed that contextual factors.Tense and Individual-Level Predicates 205 6.1 Necessary Conditions for Lifetime Effects First. can be argued to be IL and. whether they denote a lifetime property. that lifetime effects are subject to certain contextual conditions.

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

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

I return to this point in section 6. This author proposed to consider in contrast sentences like (52). as Kratzer’s examples were. is exactly the same. since it was a consequence of their argument structure. Westerståhl (1984) and Lewis (1986) observed that domains of quantification are contextually restricted. the lifetime reading does not necessarily come out. Both examples here are from Musan (1995).208 Individuals in Time other verbs appearing in lifetime predications. roughly. despite the fact that the sentence. Kratzer’s (1995) account predicts a lifetime reading to be available with IL predicates in all circumstances. I would like to emphasize that these are not sufficient conditions for the lifetime reading to arise. like (53). Gregory was from America and Eva-Lotta was from Switzerland. it is not in (53). 1997). Then. I was introduced to Gregory and Eva-Lotta. The sentence Gregory was from America is given a specific temporal context in which it is interpreted temporally. In this section I show that even when the predicate denotes a lifetime property. To account for this fact. 8 As mentioned in the previous chapter. Musan (1995.8 Thus.4. whereas a lifetime reading is available in (52).2 Introducing the Determining Role of Contextual Factors. and others with a previous context. 6. as any other operator.1. Precisely. tense is taken as a sentence operator. ‘the day I was introduced to Gregory and Eva-Lotta’. it can be restricted by the context through a variable C.3. . such as have brown eyes or have long legs. “out of the blue” cases. (52) Gregory was from America (53) On that day. The third condition is that the viewpoint form must be the imperfect. Musan’s idea can be informally depicted as in (54). This was first noted by Musan (1995. Musan argues as follows. I will show a number of situations where such a reading does not arise. 1997) As noted earlier. the temporal operator of Gregory was from America in the second case is restricted by the previous context. Gregory was from America. The second condition is that the tense form has to be the past. Although I am abstracting away from the mechanics of the semantic formalization. In the framework she assumes. As Musan observes. in the next sections. 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.

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

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

1). since from the perspective I propose here it is not the case that the reading arises and then it becomes neutralized. but the interval the speaker is referring to.1. it seems more accurate to talk about the arising or not of the interpretation. as Stowell (1993) proposes for the other two (see chapter 5. in the past).e. In the spirit of Musan (1995). the interval corresponding to the arrival in the USA. In other words. since we have proved that it is the content of the TT that matters. As I introduced in chapter 5 (section 5. the interval at which Maria and Harry arrived in the USA. Likewise.” However. is located within the span of time at which being from California holds and after the RT. In this case. 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.9 What needs to be defined is which content gives rise to the lifetime interpretation. 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. From this. this interval is called “Topic Time” (TT) (Klein 1994). lifetime effects can be reasonably considered a phenomenon of temporal interpretation. Tense is locating in the past not the whole interval the eventuality (be from California) holds.1..Tense and Individual-Level Predicates 211 sentence—namely. 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. here assumed to be the Utterance Time (UT) (i. Thus. I thank Tim Stowell for this remark. I assume that the nature of the TT is not different from other intervals like Reference Time (RT) and Eventuality Time (ET). . section 5.2). no lifetime reading is available. I assume that the TT is a time-denoting phrase (ZP).

the lifetime issue is reduced to the working of temporal interpretation in general. that the TT takes its content from the previous interval mentioned (I will refine this account in section 6. In this sense. For sentences inserted in a context with a previous past interval mentioned. focusing on sentences with lifetime predicates. we have seen two types of cases: “out of the blue” cases and sentences inserted in a context. 6.4 The Determination of the TT Content and Lifetime Effects Thus far. who. 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). a lifetime reading arises (or not) depending on the content of the TT. I argue that these two cases can be explained uniformly. Thus. For “out of the blue” examples. I claimed.3). in (61). I conceive the lifetime reading as an interpretive outcome derived from the location in the past of a TT with a specific content—namely. it gets its content from salient elements in the conversation. that referring to the lifetime span of the individual referred to by the DP subject. also in the line of Musan. the reading that arises is a lifetime reading. I will discuss the role of the contextual information conveyed by the DP subject itself in particular. In what follows.212 Individuals in Time (61) RT (UT) T′ 2 T AspP (after) 3 TT Asp′ (Harry’s existence interval) 2 Asp VP (within) 2 ET VP (being from California) Since. In sum.4. that the TT takes its content from the lifetime interval of the individual referred to by the DP subject. as I said before. and Tense restriction for contextual ones). I proposed. in this way differing from Musan. Put in these terms. accounts for them by appealing to different resources (Gricean Maxima of Quantity for “out of the blue” cases. it is the whole span of time of Harry’s existence that gets located in the past. I argue that the temporal TP 2 . based on Musan (1995). The next sections discuss the sources for the TT. The leading idea in this regard will be that the TT is topic sensitive—that is.

I will show that the TT is influenced by the discourse topic.Tense and Individual-Level Predicates 213 interpretation of these two types (with and without context) varies (i.e. I assume that every natural language discourse takes place in a context. The common ground determines the options that are live at any point in the conversation. 1979). nevertheless. 6. and von Fintel (1994). Grice (1975).3 I will show how my proposal (to account for all the cases uniformly) works in a systematic way.4. following von Fintel (1994). Kratzer (1977. they are anaphoric to distinct “discourse topics. In the next two sections.” Following widespread proposals in semantics and pragmatics. 1981). where. As defined by Stalnaker (1972. I define “discourse topic” as the subject of immediate concern in the conversation. In section 6.1 When the Subject Is a QDP I am going to examine a set of examples without any previous explicit context. 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 . Consider the following sentences. Regarding the specific relationship between the discourse topic and the elements influenced by it. 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. 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).4. a set of propositions that form the shared background of the participants in the conversation. Among the elements constituting the discursive common ground. lifetime effects do not arise. the content of their TTs is different) because their TTs get their values from distinct “discourse topics”. that it can be described as an anaphoric relationship.. Based on Keenan-Ochs and Schieffelin 1976. I assume. among many others. I assume that sentences are uttered against this background.

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

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

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

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

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

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

As Kratzer (1988. 6. Since the DP subjects of copular sentences can be considered “topical elements. 1997). Actually. the value it takes by default is the individual himself and the interval associated to him.220 Individuals in Time crucial importance in the licensing and interpretation of a past form. (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. (82) Gregory was from America (83) Topic of the sentence that may establish the connections to a previous background = Gregory (84) Gregory c= undefined. Finally. Result: a past form is allowed in the copular clause and. Bearing in mind that the TT is a ZP context sensitive. Since the contextual variable directs us to a previous context containing a past interval. I have argued that the establishment of the relevant context is brought about through the free variable DPs (more specifically. these typically are out of the blue sentences. 1995) and Musan (1995) assume. Result: lifetime arises With Musan (1995.4 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. let us consider a sentence where a lifetime reading can arise. (…) Topic of the sentence that may establish the connections to a previous background = Harry. Once a context is fixed. it is the information associated to the DP subject that counts.4 Summary of Section 6. since they are the same in all relevant respects—all of them are copular sentences in context. a lifetime effect does not arise. that in (59) it is the contextual information associated to the DP Harry that establishes the context that counts. I propose. then. there is no reason to consider (59) different from (72) and (73). when there is no specific context the free variable is to be linked to. Gregory himself TT → the whole interval overlapping Gregory’s lifetime span. The interpretive procedure is always the same and works the following way.” they establish the relevant context. the ZP . I assume that.4. (80) Harry was from California. by default. since it refers to the time of the arrival. no lifetime effect arises. the determiners) have.

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

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

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

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

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

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

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

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

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

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

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

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

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

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

así que no tuvo que pasar la aduana Maria and Harry arrived in the USA.1.Tense and Individual-Level Predicates 235 (59 ) María y Harry llegaron a EEUU. etc. the lifetime reading does not arise. I then argued that the TT is an interval (arguably. in the RC of (109). and (b) those denoting properties that are not necessarily permanent. coming from the antecedent DP a guy. I argued that equating IL-hood with permanency of the property is inadequate. I divided IL predicates into two groups: (a) those denoting permanent or lifetime properties. Stowell (1993) proposed that the content of the RT (PRO) was defined by the “closest” c-commanding ZP. Thus. considered as a “temporal PRO” affected by control. differing from Kratzer (1995). The temporal bounds of lifetime properties (Eskimo. the content of the TT is not primarily subject to syntactic conditions. the relevant background (where the subordinate TT finds its content) is built up through the contextual properties present in the relative pronoun. that one where the guy was kissed. therefore. First. by virtue of its properties as “specific.) coincide with the initial and final temporal bounds of the . but by the discursively most prominent ZP. namely. TT is an interval sensitive to discourse prominence. the context acting as relevant background is the one the DP is linked to. This makes the kissing time the most likely antecedent for the TT in the RC. I suggest that. The TT is not sensitive to syntactic parameters (such as closeness) but to the salience an interval has in discourse. In section 6.6 Summary of the Chapter In this chapter I dealt with the temporal interpretation of IL predicates and made two points. which. I argued against the idea that lifetime effects arise with any kind of IL predicate and under any circumstance. That is. Opositive. Harry era de California. independently from their c-commanding relation. Since.2. The analysis of lifetime predicates in RCs suggests that the TT is not determined by the closest ZP. as Stowell (1993) proposes for the RT.” reaches a topical position by being scoped out. However. I analyzed the temporal interpretation of IL predicates in several contexts. but to conditions related to discourse topicality. the TT is not the interval over which the lifetime of the individual extends over. 6. the TT of the lifetime predicate in the RC refers to the context containing the TT of the kissing time. data like the RCs discussed here show that the content of the subordinate TT can refer to the main TT. As surveyed in chapter 5. as native intuitions confirm. so he did not have to go through immigration As in (59). In support of these two points. In section 6. a ZP in nature) whose content is not determined by syntactic distance factors. This leads us to the conclusion that the content of the TT is not determined in the same terms proposed for the RT. Harry ser-IMPF-PRET-3SG from California.

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

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


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

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

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

I have considered that each of them is syntactically represented (quantity.242 Individuals in Time of them describe SL predicates as “involving aspectual properties. 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. I understand that the difference consists of something describable along the lines of the rightmost column. Thus. Table 7. 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 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). their definitions. SL predicates are conceived as episodic.D. In contrast. I distinguished three temporal levels (inner aspect. once “acquired.” “temporally bound. if we say that SL and IL predicates differ in certain temporal terms. One of the difficulties in these descriptions is the vagueness as to what exactly the authors mean by such terms. Chierchia 1995. 7. tense). which lack all such characteristics). nonstable predicates. if we . However. aspect.” it lasts for the remainder of the individual’s lifetime (Ph. I have argued that to attribute permanency to a property amounts to assigning it a determined temporal interval that the property extends over.3 The IL/SL Dichotomy Is Not a Permanent/Episodic Distinction IL predicates have been defined in the literature as permanent predicates (Kratzer 1988. Different number of occasions eventualities occur and different orderings between the Eventuality Time (ET) and the Topic Time (TT) intervals. In this work. among others). Differentiation between homogeneous and heterogeneous predicates Outer aspect Inner aspect Table 7.).” or “temporally anchored” (in contrast to IL predicates.2. and the distinctions that can be made according to each one.2 summarizes the temporal units. and gave concrete definitions to each. 1995. outer aspect. To understand a property as permanent means either that the property holds for the entire lifetime span of the individual (intelligent) or that.

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

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

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

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

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

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

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

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

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

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

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

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

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

those peculiar properties. This perspective also allows us to avoid. among other things. Ser classifies the subject by categorizing it into the class denoted by the predicate following the copula. 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 . The definition of ser given explains why nouns combine with ser (40) (since nouns denote classes). the obligatory predication with estar is explained. Since a location is not a class. I proposed that the peculiar properties to the constructions these APs participate in emanate from the particular properties the APs themselves involve. although episodicity can be one of the consequences of the linking to a particular situation. As we already know. but does not account for why they can appear just with ser and cannot be conceived as linked to a particular situation. such as agency. it seems that both notions can be conceived apart from each other. the cruel-type. en España (41) Madrid está Madrid estar-PRES-3SG in Spain Since a location is not a property of any individual. but something external to it. the linking to a particular situation is independent from the restriction of a predicate to a determined temporal interval: (42) Pedro fue camarero dos meses Pedro was a-waiter two months Finally. Thus. In chapter 4.256 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. Consistently with this hypothesis. the description of estar as ‘episodic’. unlike adjectives. which would leave (41) without explanation since in Spain is not a temporary property of Madrid. the ungrammaticality of Juan es en Brasil is explained. 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.

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

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

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


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

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

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

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

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

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

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

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

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

MA: MIT Press. Somerville. Pinker (eds. dissertation. . and the structure of IP. 1995. Pustejovsky. H. J. CA.). Ph. Verb movement. Piñón.). 1995. Washington. In The Projection of Arguments. Piñón. 179–206.. Rapoport. 1987. Quine. W. Wall. Cambridge: Cambridge University Press. Nominal. ter Meulen and R. 1960. Cambridge. Mood at the Interface. Levin and S. Universal Grammar. and S. Piñón. E. Dordrecht: Kluwer. A Mereology for Aspectuality. W. Proceedings of the Berkeley Linguistic Society 26:435. E. 1998. Rosen. Happening gradually. 343–366. Pollock. G. Stanford. 1993. 1987. B. 1989. A. New York: Academic Press. Delimiting events in Syntax. 2003..). New York: The Free Press. Butt (eds. Reinhart. Rosen. 1991. Reichenbach. Pustejovsky. The syntax of event structure. 1995. J. 47–81. Wh-in-situ: Movement and unselective binding. and Small Clauses: A study of Israeli Hebrew. CA: CSLI Publications. T. 98– 129. Stanford. A. Linguistic Analysis 26:29–62. Carnie. University of Utrecht. In Current Studies in Romance Linguistics. Ritter. Interpreting mood. Stanford University. In The Representation of (In)definites. Cambridge.-Y. Hensey (eds. Tenny (ed.D. 1999. Ramchand. First phase syntax. E. MA. Mathematical Methods in Linguistics. C. V. MIT. Haugen and P. 2001. Linguistic Inquiry 20:365–424. University of Utrecht.). Guasti (eds. The semantics of copulative constructions in Portuguese. Quer. 2000. Ph. Elements of Symbolic Logic. 420–433. Ritter. In Small Clauses [Syntax and Semantics 28]. Durative adverbials for result states. Norquest (eds. C. MA: Cascadilla Press. Cambridge. 1947. Ph.). The theta system: syntactic realization of verbal concepts. Quer. Raposo. Word and Object. Oxford: Blackwell. 1976. Cambridge.D. and S. Berkeley. 135–164. MA: MIT. DC: Georgetown University Press. J. Two types of small clauses (Toward a syntax of theme/rheme relations.). and J. 1998. 1988. Strong and weak predicates: Reducing the lexical burden. Pustejovsky. Luján and F.D. dissertation. In Proceedings of WCCFL 18. The geometry of events. dissertation. Probus 13:81–111. Ms. J. Uriagereka. J. UiL OTS Working Papers. T. 2000. 1996. D. S. B. Bird. Pesetsky. C. 19–39.270 Individuals in Time Partee. University of California. Copular. MA: MIT Press. A. J. Querido. C. Greuder and M. Reuland and A. A. University of Oxford. In Lexical and Conceptual Structure.). J. M. The Generative Lexicon. Cardinaletti and M-T. In Studies in Generative Approaches to Aspect [Lexicon Project Working Papers 24]. E. ter Meulen (eds.

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

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

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


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

86–90. 194. M. R. 245 Martin. 88 Lemmon. 138. 135 Kiparsky. 150 Reinhart. 124. 215 Stalnaker. 33. 118 Matushansky. 156 Tenny. 119. 140. 3. 252. 77 Roldán. S. 32. L. 119. 29 Talmy. 164. P. 162. 220. E. 240 Kondrashova. 3. 45. 115. 188 Keenan-Ochs. E. 31. ter 148. 1. 191 Partee. 43. 43. 50. 75. M. A. 116. 38. 94. 163. W. 8. 110– 112 Torii. 165. 98. 153. C. R. V. H. 178. 102 Matthewson. 119. B. P. 120– 123. M. J. 211. 154. P. 214 Svenonius. 235. L. 32. 176. 43. A. 3. 216 Reichenbach. de 48 Milsark. 61. 213. 11. 15. 216 Uribe-Etxebarría. T. 12–14. 79. 59. 148. 207. 84. 106. 214 Luján. 55. J. 3. R. de 30. 40–44. 29. 250 Ladusaw. 107. J. M. 56. 28. 240. 112. 214 Longobardi. 213 Kennedy. R. 254 Lewis. A. 78. 5. E. 5. 102. 89. 217–220. 40–43. M. 233 Querido. 26. 10. P. 194 Kratzer. S. 37 Pustejovsky. 152. 46. 22 Kenny. 152. J. 155. 85. 113 Kitagawa. R. 122. C. 213 Schmidt. 86. 244 Stump. 126 Morimoto. 144. C. C. 8–12. 25. 38 Mithun. 215. 243 Travis. C. 44. T. 144. 47 Keyser. C. 251 Sportiche. 58. 222. 251 Krifka. S. D. 127 Uriagereka.-L. S. 205. 144. 205. 143 Russell. 112 Tungseth. 189. 127. 29. F. 233 Rohrer. 33. W. 237. A. 191 Larson. 40. 80 Ritter. 172. 236 Naumann. 6. M. 164 Miguel. 3. D. 218 Phillips. 191 Picallo. H. 208. E. 120 Pesetsky. M. 208–212. 151. 213 Stowell. 162. 22 Meulen. D. 251. 138. 36–38. 71. 61. 207. 70. 18 Schmitt. J. 145. 117. 38. 148. 71 Ramchand. 136. 97. 41. L. 15 McConnell-Ginet. Y. 3. G. 167. 16 Rosen. 191 Ogihara. O. 12–14. 37. S. T. 120 Parsons. M. 113. 127. 40. B. 178. 232.276 Individuals in Time Kearns. 8. 93. 214 Perkins. 12. C. T. 240 Rivero. 43. 193–196. 225. 176 Szabolcsi. 225. 222. B. 225. 232 Ryle. 112 Schein. 150.-Y. 38. 106. 156. G. 242–244. W. T. A. 14. 251 Quer. 188. 96. 60. 113. 226. 12–14. 117. 43. 131. 55. 11. 194 Klein. 109 McNally.-Y. 25 Quine. 147. 38. 41. C. 74. 202. 148–150. S. G. 145. A. 144. 195 Leonetti. B. 192. 15 Koopman. 94 Morera. 208. 191 Pollock. J. 232–235. 147. 94 Musan. 211. 176 Munro. 71. M. 143. 61. 55. 25–32. G. Y. E. 193. T. 126. 15 Raposo. 250 Rapoport. 214. 77. 196. 15. 122. J. 84. 73. 50 Landman. 114 Schieffelin. G. M. 80. 109. 136. 114 Kuroda. J. 232 Lakoff. S. 81. H. 49. 189. 130 Mourelatos. 56 Sanz. 109. N. G. 37. 70. 154. 61. 122. L. M. 8. 14. 152. 70. 34. 55 Rosen. 113. 127 Swart. 240 . 218. 112. 40. W. 57 Masullo. 145. 240 Rothstein. 245. 192. P. 143. R. 84. 116. 18 Piñón. 190. 116. 240. 96. 16. 97. 226 Pagliuca.

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


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

137 Subinterval property 70. see Temporal interpretation Russian 15 Salish 118. 233. 202 Ground 119. 189. 131. 236 Theme 105. see also Event(ive) argument Speech Time 122. 129. 229. 191. 117. 134 Instigator 55. 215 Goal(Path) 96–98. 35 Figure 119. 174–178. 112. 11. 234 Situation aspect 39 Small v 55 Spatiotemporal argument/variable 8–10. 187. 128–131. 154. 181 Existential reading 6. 155. 184. 186. 179. 131. 57. 72. 226. 155–157. 159. 100. 159 Quantification restriction 11 Quantity 113–116. 246–248. 76. 227–229. 141. 132. 122. 12. 221–223. 31 Imperfective paradox see Progressive paradox Initiation. 153–160. 165. 240. 186. 246. 158–160. 224. 116. 176–178. 246. 147. 141. 158. 227 Simultaneous reading 76–79. 161. 160. 232. 233 Perfect entailment 47. 227. 72. 189. 136. 58. 14 Generic reading 6. 147. 167. 86. 168 French 121 Generic operator 11. 201. 132. 94 Margi 121 Maxim of Quantity 85. 207. 258. 136 Place/situational prepositions 120. 189. 180 Perfective (preterit) 31. 249 Relative clauses temporal interpretation of. 154–157. 195. 257. 67. 148–152. 211. 35. 55. 184. 248 Swahili 120 Temporal interpretation of Complement clauses 76. 6. 79. 183. 83. 136. 245. 111. 258 Hindi 120 Homogeneous (predicates) 70. 162. 221 Habitual 77. 142. 130. 180. 187–192. 72. initiator 61. see also Subinterval property and Additivity property Narration temporal interpretation of. 241. 165. 207. 222–225. 252 Realization function 6. 178. 123. 176. 148. 189– 192. 248. 249. 224. 244. 230 of Narration 77–80 of Relative clauses 223. 186. 194. 115. 246. 176. 158. 80. 144. 212. 156 Hawaii-Creole 120 Hebrew 15 Heterogeneous (predicates) 72. 14. 244. 158–160. 162. 113. 131. 231. 158. 116. 175. 129 Progressive 153–163. 177–180. 152. 250 . 178. 189. 172. 158. 189. 156 Focus 13. 257 Perform(er) 53. 94. 175–177. 80. 192 (Im)perfective contrast 25. 148. 112 Juba Arabic 120 Kinesis 40 Krio 120 Lakhota 55. 209. 134–136. 178. 229. 172–174. 123. 26. 186. 194 Reference Time (RT) 3. 226. 144. 79. 148–152. 192. 249 Existential coda 12 Existential quantification 8. 212 Mereological properties 3. 129. 192 Prospective 120–122. 258 Imperfect Continuous 76. 12. 183. 172.280 Individuals in Time Event-object 110. 125. 212. 222. 147. 171– 175. 230. 134 There-sentences 5. 147. 128. 37 Thetic (judgments) 12–14. 226. 258 Progressive paradox 191. 240. 187. 123– 126. 14. 148. 211. 192. 248. 184. 118. 186. 135. 242. see Temporal interpretation Past-shifted reading 76–78. 148 Sequence of Tense 226. 132. 148. 134–136. 173–180. 181. 208. 173. 230. 144. 64. 35. 88. 105. 122. 242. 148. 196. 157. 227. 150. 106. 242. 191 Event(uality) Time (ET) 3. 233–235. see also Utterance Time Stativizer 115. 94 Pomo 94 Prepositions Directional prepositions 127. 11.

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

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