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

............................... 1995) .....2....1..40 3.......2.....................1 Proposals about the Individual-Level/Stage-Level Dichotomy ................1.............2 ..............5 Summary of Section 2.... The Structure of Copular Constructions .............................5 2.............. Inner Aspect and Event Types .................................59 3........................................11 2.............2.......................5 2................33 2............................xi Chapter 1 Presentation of the Study ..................1........1.............................61 3..........39 3.................................................2.............................................................................8 2...3 Summary of Section 3...2............15 2......3 Summary of Section 2.........53 3......................................1 ....................3 Aspectual Differences among Individual-Level Predicates .........1 Distribution of Copular Verbs in Spanish .......................................25 2.....................2 ...................................................16 2..........4 Summary of the Chapter ..................... When There Is More Than One Copular Verb: Spanish Ser and Estar...........1 .......................39 3......14 2.......................................2...............38 Chapter 3 Event Classes and Individual-Level Predicates ...............................2........3 Summary of Section 3.......................5 2....................39 3.................2 Event Types and Event Structure.........1 Inner Aspect...... by Tim Stowell .. Introduction to the Tests Discerning the IL/SL Status............................4 The IL/SL Contrast as a Categorical/Thetic Distinction...................................................2 The IL/SL Dichotomy as a Syntactic Distinction....3 IL Predicates as Inherently Generic Predicates......................1 The IL/SL Distinction as a Semantic Distinction........................ Chierchia (1995) ..53 3.2....1............. Milsark (1974) and Carlson (1977)..........Table of Contents Acknowledgments......................................................1...................................1 A Cluster of Notions ..61 ........32 2....3.... Higginbotham and Ramchand (1996) ............................1 Chapter 2 Individual-Level Predicates ......................51 3............2 Agents in Event Structure .......................................................................................1................1...................2 Ser and Estar as an Aspectual Differentiation ..ix Foreword.......................................... Raposo and Uriagereka (1995)............... A Brief Stop at “Agentivity”......12 2................................................................................... Kratzer (1988................1....

.......7.........143 4..............110 4.......................................2 Summary of Section 4...........3 The Relational PP with Other APs .90 4........ States and Activities: A Grammatically Relevant Difference?....................................72 3....4 Summary of Section 4..................................... 69 3........................2 On the Optionality of the (Affected) Goal PP ..............137 4.................................................................................................1 Similarities between Activities and States ...................................3 Some Confusing Arguments in Defense of a State/Activity Distinction............2 Summary of Section 4.............................1 Prepositions as Aspect Encoders ...............3..98 4.5 ............1 Lexical Semantic Classes of Adjectives ......2 Differences between Activities and States .......4 The Issue: Aspectual Alternation in IL Copular Clauses ......3 The DP Subject of Relational MPs ............................84 4..............3 Adjectives Predicated of an Implicit Event Argument .........105 4.3.......4...............3................................... The Relational PP Complement....1......................84 4..............133 4.................4..68 3....2 The Preposition Introducing the “(Affected) Goal”: An Activity Inductor123 4.........................1 The Two-Copulas Hypothesis .............................................143 ..117 4.6...........91 4.............2............80 3.......4 Summary of Section 3...............................106 4.....................1......5...........6 An Account Based on the Relational PP Complement ........................................................3 ....6.......7....................1 On the Interpretation of the Relational PP .......................3 Relational Mental Properties.......2 .................... Summary of the Chapter ....................................96 4....86 4...............85 4....4 Summary of Section 4...83 4.......................116 4.......... The Cruel-type as a Small Clause Individuals in Time 3...6...........................96 4..........2...........................1 Aspectual Tests on IL Predicate ............5..........4. Summary of the Chapter ...2 Be as a Copula Shifting a State into an Activity .....5 Justifying the Approach .....6.........................................................1 Cruel-type Small Clauses Taken by Verbs Other Than the Copula ..............................................7 ................................................3....................2 Summary of Section 3..........135 4...........3 ...............4...6 ......................................8......1............3.........108 4....................1 Previous Explanations of Nonstative Copular Clauses ........................2 Syntactic Approaches ...81 Chapter 4 Aspectual Alternations in Individual-Level Predicates ....................95 4....4 ..............7..........................5....76 3.........................................107 4.................................................3 Summary of Section 4.......................69 3...........................................117 4.....................................138 4.5.....................................108 4.....................................................3............4...1 Lexicalist and Logico-Semantic Approaches ......62 3...2 Adjectival Predicates Showing an Activity-like Behavior .................................................................

1 Permanent IL Predicates ..........................194 6.......................................................................157 5........................2 Differentiating Temporal Extensions of IL Predicates ............................ 192 Chapter 6 Tense and Individual-Level Predicates .................. 1996) .3.....152 5..............171 5..................3 Brief Notes about the Complement of Mental Properties APs .........................................2 Nonpermanent IL Predicates.....1 Tense.....................202 6..............................3.......................................210 6..........................205 6......................2 Context Associated to Individuals .................199 6..............................2............................4.......4 The Determination of the TT Content and Lifetime Effects ............ Kratzer (1988.....2 ..... 176 5..........................................4...................3...3 The Progressive and Adjectival IL Predicates .... 180 5....................................2.2................1 The Imperfect and Adjectival IL Predicates ...3 Inner and Outer Aspect .....2 Aspect as a Quantifier over Occasions .5..................215 6.......1 Habitual Heterogeneous Predications...... Stowell (1993.........................................................2..................................... 1995) ...................4 Summary of Section 6............6......5 Adjectival Individual-Level Predicates and Viewpoint Aspect .................213 6...................................5 The Lifetime Reading in Compound Sentences .............. Proportion..2.................................................................... Summary of the Chapter ........................4..163 5..... 1997) ....................................2 Perfective Homogeneous Predications .2 The Habitual Interpretation: Iteration...........................................................................................Table of Contents vii Chapter 5 Outer Aspect and Individual-Level Predicates ..... 179 5..........................147 5..........1. 187 5.............218 6.....2...............................5.....3 On the Relation between the TT and the Habitual Q<occ> ....2............1 When the Subject Is a QDP ..............................193 6.......3 The Arising of Lifetime Effects ..... A Brief Summary of Aspect Notions .....................4 Summary of Section 6...................................................148 5.1 Necessary Conditions for Lifetime Effects .............................1 Quantifying over Occasions ...........173 5.................................................................3 The Content of the TT in the Arising of Lifetime Readings ........3....1................ Aspect as an Ordering Predicate ..................................................................2 The Internal Argument of Tense: The External Argument of Aspect..4 ............................................2 Introducing the Determining Role of Contextual Factors..............1 Temporal Interpretation as a Consequence of Argument Structure.......... 186 5.... Musan (1995.................5..............199 6..............................205 6..................................... and Systematicity ...............205 6......................2 The Perfective and Adjectival IL Predicates ......3................4..220 6.............................208 6...197 6...............................................148 5.................. 174 5.............3 Articulating the Account........ 177 5.......................222 .........1 Tense and Aspect as Ordering Predicates .....4................................158 5...........212 6...........

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

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

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

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

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

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


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

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

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

chapter 7 summarizes the conclusions of the study and offers some considerations about the appropriate description of the IL/SL dichotomy. and second. paying special attention to whether the perfective plays any role in making atelic eventualities telic. the imperfect. focusing on Lifetime Effects. the influence of contextual factors on the arising of the lifetime reading is derived. I also discuss the relationship between inner and outer aspect. I will put forth two claims: first. and the progressive. the lifetime reading is a salient reading only with lifetime IL properties. the perfective.e. ordering and quantification over occasions). Chapter 6 centers on the temporal interpretation of IL predicates. the lifetime interpretation depends on the content of the Topic Time. I show that inner aspect properties are independent from outer aspect ones (i. In this respect. Since contextual factors influence the content of the Topic Time. Finally.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. 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 will critically revise those analyses. the predicates of (1) are allowed. NP). 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. and introduce some of the concrete aspects for which this work makes alternative proposals. and pragmatic distinctions.1 Proposals about the Individual-Level/Stage-Level Dichotomy 2. according to which the copular verb selects for a small clause (SC) containing the nominal argument and the predicate (AP.3. There are several policemen available b. ser and estar. arguing for an aspectual distinction. a part of this chapter presents the differences between the two Spanish copulas (ser and estar). 2. Section 2. but those in (2) are excluded. since I am chiefly concerned with copular cases. Milsark (1974) and Carlson (1977). I also introduce proposals by Heycock (1994) and Becker (2000). the one offered by Stowell (1978.4 summarizes the chapter. I will first present the main descriptions of this notion found in the literature. There are several policemen in the corner a. I concentrate on the most influential proposals about the contrast between individual-level (IL) and stagelevel (SL) predicates. Since the data in the work are from Spanish. In section 2. *There are several policemen insane . according to most authors. (1) (2) a. Section 2.Chapter 2 Individual-Level Predicates In this chapter I introduce the concept of the individual-level predicate. For example. *There are several policemen intelligent b.2 is devoted to the differences between the two Spanish copulas. After that. specifically discussing arguments for semantic. 1981). the presence of functional projections) and its relationship with the IL/SL distinction. In section 2.1. who discuss the constituency of the SCs that copular verbs take (in particular.1. Likewise. which. syntactic. IL and SL. I will introduce the syntactic analysis that I will assume for copular sentences. lexically distinguish the two types of predication. This chapter is organized as follows. I present other two aspects basic for my purposes in this work. PP.1 The IL/SL Distinction as a Semantic Distinction. I present the syntactic analysis I assume for copular cases—namely. indicating some of the points on which my proposal differs.

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

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

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

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

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

(28) and (29) are excluded. 1995) hypothesis. Chierchia (1995) In Kratzer’s (1988. if these properties generally hold of the individual. the Davidsonian argument has to be locally bound by a generic operator (Gen).6 The difference between SL and IL predicates is that. it is due to the fact that. According to him. He defines IL predicates as predicates that ascribe permanent or tendentially stable properties to their subject. the possession or lack of the Davidsonian argument is what draws the line between SL and IL predicates. Chierchia (1995) argues that all predicates (including states) project a Davidsonian argument ranging over occasions/eventualities. it makes no sense to link them to a specific situation in which one observes such a property. in IL predicates. in the line of authors such as Parsons (1990). . Rather. taken from Chierchia (1995:207). This. he submits.Individual-Level Predicates 11 2. Taking a neo-Davidsonian approach. (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.3 IL Predicates as Inherently Generic Predicates.1.7 Chierchia (1995) argues that the hypothesis that IL predicates are inherent generics can account for the properties they typically display. generics express tendentially stable properties. according to which a generic/habitual interpretation is only possible with eventive verbs. which is also the main characteristic of IL predicates. Enç 1991b). representation. In Chierchia’s hypothesis. Hence Gen is defined as a null quantificational adverb over situations. the Gen operator ranges over situations that are arbitrarily located—that is. given that tallness (in (29)) is clearly perceivable and (29) is as ungrammatical as (28). The judgments are his. over situations not restricted to a concrete space or time—which explains the oddity of the locative modifiers (in italics) in sentences (25)–(27). Regarding the exclusion of IL predicates from the coda of there-sentences. Chierchia explains it as a subcase of the “quantification restriction” (Milsark 6 Chierchia assumes that this variable is projected in the semantic. is sufficient for ungrammaticality. but not in the syntactic.g. too. (25) ??John is a linguist in his car (26) ??John is intelligent in France (27) ??John knows Latin in his office For similar reasons.. 7 This proposal is the opposite to the one expressed by other authors (e. The Davidsonian argument is present only in SL predicates.

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

Nevertheless. scopes out to a particular functional position (F). contextualized in a concrete event. they differ in what the sentence is about.” and point of view in general. respectively. geniality is predicated of a decontextualized champ. these authors assume that the topic of a sentence is the argument that.” “contrast. by contrast. the DP is not linked to any particular event but “decontextualized” and an IL reading is derived. Thus. they correspond to “thetic judgments” in Kuroda’s classification.Individual-Level Predicates 13 report an event. hosting “topics. and an SL reading is borne out. Thus. as they claim (and as shown in section 2.2). genial el campeón . then geniality will hold of him at all events he participates in. When it is the DP. When. at Logical Form. its transient character being derived. in the first part of (35) (el campeón es genial). (36) a. in their view. the event scopes over the DP. F 1 el campeón T′ 1 T VP 5 F 1 genial T′ 1 T VP 5 b.” “emphasis. in the second part (no está genial). (35) is not contradictory in Spanish. That is. 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. the event is understood in relation to the DP. geniality is predicated of an individual under the context of a particular event. If geniality holds of the champ regardless of his being in a given event. Following Uriagereka (1994). genial. The example they use to put forth their point is that reproduced as (35). IL and SL predicates are differentiated by virtue of what gets such an F position: either the individual denoted by the DP subject or the event. Raposo and Uriagereka account for this fact by arguing that.” “focus. which is conceived as a scope-taking process. Their proposal would look roughly like (36). However. the DP is interpreted in the context of that particular event. SL and IL clauses differ in defining the “topic of the predication”. Raposo and Uriagereka (1995) assume that the topic of a sentence is established via a process of topicalization.

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

A large part of the population uses Bambara as its mother tongue. I expand the discussion regarding what seems to be an adequate description for the IL/SL dichotomy. see also Rapoport 1987 and Rothstein 1995. In the past-tense cases. those clauses where the topic is an individual are instances of IL predications. I have introduced the idea that equating permanency of a property and ILhood is inaccurate. Spanish and Portuguese have two lexically different copular verbs.Individual-Level Predicates 15 are instances of SL predications. and as secondary language it is employed to communicate nationwide. In the next section. where the semantics of the two Spanish copulas are introduced. hinging on whether the predicate expresses an inherent or definitional property of the subject (38) or not (39) (Greenberg 1994).8 distinguish three types of copular verbs depending on the category the predicate belongs to (37). 2. When the predicate is an AP → copula “ka” b. When the predicate is an NP → copula “don” c. Matushansky 2000). 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. Other languages. 1996. such as Bambara (Koopman 1992). I take up this issue also in chapter 7.9 In Russian. ser and estar. the postcopular NP (the predicate) alternates in case (nominative or instrumental) depending on whether the property is permanent (40) or nonpermanent or noninherent (41). In turn. . 9 For further discussion on the overtness of Hebrew copula. (37) a. The Hebrew copula alternates between overt or null. the copula is null in the present tense but overt in the past tense (Kondrashova 1995. Bambara is one of the officially declared languages of Mali. found across most of western Africa.2 When There Is More than One Copular Verb: Spanish Ser and Estar A good number of languages show copular alternations.

if not a mistake. How can the distribution of the two copular verbs be described? As traditionally observed. However. which led to the claim that ser corresponds to permanent properties and estar to temporary ones. whereas estar is used to refer to those properties considered accidental. 11 See also comments by Milsark (1974) and Carlson (1977) about the same predicate.2. discussed in section 2.11 which. or ser. as . the copular verb must be ser (42). if the subject is a physical entity (43).1. obviously designates a nontemporary property.1).1 Distribution of Copular Verbs in Spanish Some traditional grammarians (Bello 1847. if the subject denotes an event (44).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. Roldán 1974) describe the distribution of the two copulas ser and estar on the basis of the Aristotelian dichotomy “essence” versus “accident. and accidental properties as temporary. In the case of Spanish. the copula used can be estar. there is no general agreement in the literature on how the copula alternation should be characterized. I would like to note Demonte’s (1979) observation that this distinction poses an asymmetry in the account of copula distribution. when the predicate is an NP. Este animal es/*está that animal ser/estar-PRES-3SG a goat “That animal is a goat” professor c. 2.12 10 Note that these descriptions fall in line with the ones given by Milsark (1974) for “properties” and “states. other authors observed that the distinction between permanent and temporary was. insufficient to characterize the copular alternation. Pedro es/*está Pedro ser/estar-PRES-3SG teacher “Pedro is a teacher” When the predicate is a PP.” The copula ser is used to refer to properties that are considered essential of the individual. The classic counterexample in this respect is estar muerto ‘be dead’.” as cited previously (see section 2.1.10 Essential properties are understood as permanent. un ordenador de 1960 (42) a. 12 Although I will not work out this difference here. if. despite being ungrammatical with ser. Ese artefacto es/*está that artifact ser/estar-PRES-3SG a computer from 1960 “That artifact is a computer from 1960” una cabra b.

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

When the adjective is referred to the place where one is born. Bosque and Picallo 1996. and those in (51) combine with both copular verbs. see Bosque 1993. However. For more details about classifying adjectives. *The trip was presidential). it is not gradable. 15 13 . 14 Note. (b) impossible modification by a degree adverb (48).13. the copular clause sounds degraded: (i) the Cuban guy → the guy is Cuban (ii) the Cuban fight against the embargo → ??the fight is Cuban The reading where African is understood as a particular way of behaving is not relevant here. that classifying adjectives are excluded as copulative predicates (The presidential trip vs. one of the properties that Schmidt (1972) and Bache (1978) provide to distinguish classifying adjectives from qualifying ones is. 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. and Demonte 1999. and (c) impossible participation in binary correlations (49). Further discussion is presented in chapter 7 (see section 7.6). that their acceptability as predicates in copular constructions seems to be confined to those cases where the subject is a physical entity. also.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. These tests are (a) impossible participation in comparatives (47).18 Individuals in Time alemana (46) Jelena es/*está Jelena ser/estar-PRES-3SG German “Jelena is German” Origin adjectives do not evaluate the DP they refer to. If the subject is a resultative nominal. precisely.

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

I am dealing with an IL predicate. Consider the next group of examples. not ser— that is.20 Individuals in Time nada gracioso. where a perception verb (note) has a SC (NP + AP) as a complement. 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. throughout this work. Thus. where the pound symbol indicates when the paraphrases are not appropriate. (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” . pero está moreno (54) Pablo es Pablo ser-PRES-3SG very light-skinned but estar-PRES-3SG dark-skinned “Pablo is very light-skinned. 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. Consider (59)–(62). pero está muy guapo (53) Pablo no es Pablo not ser-PRES-3SG handsome but estar-PRES-3SG very handsome “Pablo is not handsome. but he looks very handsome” muy pálido. all of the inflected versions in (55)–(58) that we can construe as paraphrases have to be done by using estar. Interestingly.1. as we already know from (51). (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. whenever the copular verb is ser. the copula designing SL-hood. 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. I will consider that. the copula estar will be considered a lexical sign of an SL predicate. Correspondingly.1.1). With the copular distinction in Spanish in mind.

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

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

and so on that contribute to the acceptability of the sentence. (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. 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’). subject pronouns need not be overt in Spanish. the corresponding paraphrases contain estar. the verb see distinguishes between states and events. and disambiguate between verbal forms (cantaba ‘sing-IMPF’ can refer to the first and the third person). they appear to contrast (Tu hermano puede venir al cine.Individual-Level Predicates 23 such as (70) and (71) are those that combine with estar (cf. pero tú te quedas ‘Your brother can come to the theater. Other adjectives. rather than IL/SL-hood. emphasize (Quiero que vengas tú ‘I want you [and not other] to come’). rather than ser (72). improve the sentence. Juan {guapo/pálido/moreno/gracioso} (i) *Percibí/advertí a I-perceived/noticed Juan {handsome/pale/tanned/funny} al contrincante {muy tranquilo/nervioso/apagado} (ii) Percibí I-perceived the opponent {very quiet/nervous/spiritless} As mentioned in footnote 16. 19 With infinitive complements. where a state such as know languages is excluded. or both types of perception) and the role of modifiers like muy ‘very’. or gracioso. 50 above) and. the semantics of perception verbs deserves a deeper investigation than what I can offer here. when the adjective is combinable with the two copulas. 20 . however. moreno. Consider the following contrast. the sense of the verb is that of sub- ‘notice’ are odd with predicates such as guapo.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. not all the results are so clear cut. physical. bastante ‘quite’. Roughly speaking. it is interesting to note the role that overt subjects (pronouns) play in the meaning of this perception verb in Spanish. as in (ii). In this regard. pálido.

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

Individual-Level Predicates 25 In sum. In turn. they select estar. I will discuss work by Luján (1981). a delimited process. there are many other factors intervening in the semantics of these constructions.2. Schmitt (1992). Estar expresses that a predicate is true of an individual for a delimited time period. where a beginning or end is assumed. When the predicates are used to refer to imperfective states—that is. She establishes a parallelism between estar and predicates like write a letter. but rather that it makes a different contribution to temporal reference.2. in general terms. Thus. Among those who have worked along these lines. ser expresses that a predicate applies to an individual during a stretch of time with no beginning or end assumed (76).” As I understand Luján’s proposal. as contento ‘glad’. (75) Perfective predicate A(x) at time tj 23 The cited date corresponds to the ninth edition. 2. states where no beginning or end is assumed—they select ser. or at least one of them is. Luján (1981) claims that all adjectives are stative. Although. Examples (75) and (76) are Luján’s (1981:177). Ser is conceived as a predicate parallel to write or admire.2. . they must bear the feature “–[–perfective]. the two copulas differ in the nature of the temporal reference they make.23 This author argued that the property distinguishing ser and estar is imperfective versus perfective. basically in the sense of ‘durative’ versus ‘punctual’. her account of the distribution of copular verbs (concretely in combination with adjectives) relies on two factors: the [±perfective] feature in the adjective. How can we explain those cases where an adjective can combine with both copulas? Luján argues that. both undelimited predicates. is already found in Gili Gaya 1961.2 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. if they combine with both. When they refer to perfective states. the work is from 1945. whose beginning and end are assumed (75). but can be either [+perfective] or [–perfective] (Luján 1981:175).1 Luján (1981). Following Querido (1976). it is not that the copula ser is atemporal. 2. tests based on perception verbs should not be taken as totally safe auxiliary tests to distinguish between IL and SL predicates. and Fernández Leborans (1999). According to Luján. the results emerging from them are consistent with the distribution of the copulas in Spanish. and the [±perfective] feature in the copula.

PPs (when the subject is a mobile and immobile object) Estar + Table 2. old). Among the facts that Schmitt cites to demonstrate the internal structure of estar and the underspecification of ser are those enumerated in (a)–(d). swim). PPs (when the subject is a mobile object) Spanish NPs. and processes or activities are those events with duration but without an inherent culminating point (walk. since the distribution of Portuguese copulas nearly parallels those of Spanish [see Table 2. female. but.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. its underspecification in aspect. However.2 Schmitt (1992). its wider flexibility in combining with adjectives indicates that it does not involve internal temporal characterization. since there is no reason to consider in the garden as a delimited predicate. scenarios incompatible with a result state are excluded for estar. happy). States are eventualities with no duration or endpoint (be tall). APs (human. (More on this in chapter 7. APs (human. PPs (when the subject is an event (party)) APs (tired.1]. ser manifests a wider flexibility. Schmitt provides several scenarios where estar and ser contrast in their possibility of appearance and argues that the explanation rests on the aspectual properties of estar. PPs (when the subject is an event (party) or an immobile object) APs (tired. her reasoning is as follows.2). Estar corresponds to a result state of an accomplishment verb. old). ser has no inherent temporal structure.2. Predicates like be in the garden are not covered by Luján’s generalization. happy).24 In a nutshell. nor is it an event or a process.1.1. Since estar refers to a result state. round. (Schmitt works on Portuguese. Portuguese and Spanish copular verbs The concrete claim of Schmitt (1992) is that estar involves aspectual properties.26 Individuals in Time (76) Imperfective predicate A(x) at times tj… tj+k However. In the spirit of Luján. which suggests. female. .) 2. That is. In contrast. Schmitt (1992) argues that the distinction between ser and estar is aspectual in nature. whereas ser is underspecified with respect to aspect. It is not a state. round. Accomplishments are those predicates that refer to a process with a inherent endpoint (write a letter or build a house). for this author. 24 All these concepts will be adequately presented in chapter 3 (see section 3. I take it that her conclusions about the topic can be applied to Spanish.) Copula Ser + Portuguese NPs.

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

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

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

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

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

” Along similar lines.). Since the descriptions of the readings fall in line with Carlson’s (1977) distinction IL/SL. Fernández Leborans (1999) argues that the IL/SL distinction is a primary parameter of lexical aspect. 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. More concretely. it can be said. when a predicate appears in combination with ser. known as “inner aspect. as I advanced. after having introduced a formal description of inner aspect and other temporal domains. which is that the IL/SL distinction is not a matter of inner aspect. The terms in which Luján treats the dichotomy IL/SL refer to the internal temporal properties of the predicates. others that combine just with ser.32 Individuals in Time endpoint is involved. Whereas NP predicates obligatorily combine with ser. lining myself up with Bosque (1990). whereas estar involves aspectual properties (it corresponds to a result state of an accomplishment verb). I have concluded that the opposition ser/estar is adequately described in terms of the IL/SL split. 2. mean. I have considered the native intuitions about the different interpretation of such pairs. I undertake this point in chapter 7.2 In this section I introduced the distribution of copular verbs in Spanish. yielding minimal pairs. is the topic of chapter 4. It is premature to introduce and develop a discussion about whether the real differentiation between IL and SL concerns inner aspect. I have critically reviewed her arguments. whereas estar predicates are taken as parallel to write a letter.” or “aktionsart. and others that are able to combine with both copular verbs. that there are some that combine just with estar. the subject is categorized as belonging to the class denoted by the predicate. I will describe ser-predicates as classificatory. 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. 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. which. a process where a delimited point is involved. What I can advance here is the conclusion I will draw. paying special attention to her discussion on the behavior of ser in combination with certain adjectives (cruel. Therefore. depending on which verb is used as the leading hint to define the opposition ser/estar. rather than states. and Escandell-Vidal and Leonetti (2002). APs present a more complicated paradigm. Schmitt (1992) considers that the difference between ser and estar is aspectual in the sense that.2. Fernández Leborans (1999). etc. Although I will add some remarks in chapter 7. That is. the property is associated to a concrete circumstance (also along the lines of Higginbotham & . I will assume that with estar (SL) predicates.” “lexical aspect. for the time being. based on Escandell-Vidal and Leonetti 2002. Demonte (1999). ser is underspecified with respect to aspect.3 Summary of Section 2.

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

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

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

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

Because there is not a thematic subject. Regarding the internal constitution of the SCs taken by the copular verb. 1995) in that SL predicates do not involve the projection of just an event argument (here an Event Phrase). for Becker. it cannot be said to have been generated inside the VP but inserted at a higher level within the perception-verb complement. whereas Stowell (1981) originally conceived SCs as lexical projections. argued for the existence of two homophonous verbs be. Felser proposes the extra layer to be an Aspect projection. the contribution of this book will be twofold. but the projection of two nodes. in particular to copular sentences. Becker establishes a correspondence between the presence of Aspect and SL-hood. assuming that to occurs in the T head (Emonds 1976. and Becker (2000)29 have put forth the presence of a functional node (Aspect). In the first place. In particular. She just needs to assume one copular verb. as pointed out in section 2. Felser concludes. Thus. Likewise. the difference between SL and IL copular constructions resides in the complement of the copula. Pollock 1989). (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. differing in this respect from Carlson (1977). . Event Phrase and Aspect Phrase. unlike Schmitt. other authors have argued for the existence of functional projections inside them too. suggests that the perception-verb complement is not a TP either. see (105). Whereas SL predicates are linked to aspectual content. Thus. Furthermore. Summarizing. However.Individual-Level Predicates 37 (108) I wouldn’t like to see [there be so many mistakes] Her argumentation goes as follows. The reasons to discard a CP or TP projections come from the ungrammaticality of sentences like (109) and (110). who. Heycock (1994).1. since both establish a relationship between copular distinction and aspectual content. 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. the exclusion of a to-infinitive. such an aspectual content is absent in IL predicates. perception verbs take Aspect Phrases. Becker considers that an aspectual distinction does not exclude in and of itself a split conceived as IL versus SL. Becker’s (2000) idea differs with Kratzer’s (1988. The proposal defended by Becker (2000) shares the spirit of Schmitt’s (1992). Chomsky 1986.1. As can be appreciated from (105). authors such as Schmitt (1992). which are presented as codependent. I argue that IL-hood 29 See also Guéron and Hoekstra 1995.

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

e. I will show that not all cases in which the copula ser participates share the same inner-aspect properties. First.1 Inner Aspect and Event Types In this section I deal with two points. I start by investigating. Other more specific issues. whereas another can be justified as dynamic predicates (i. refers to the internal temporal properties of events in sentences. To accomplish this task. in this chapter. Once we have become familiar with the different event types.4—namely. Such a contrasting behavior raises a number of questions and invites reflection on some general issues. Specifically. Because permanency and stativity are temporal concepts in nature. their internal temporal properties (inner-aspect properties). Inner aspect has to do with properties such as duration. I described individual-level (IL) predicates as predicates that are said of their subject independently from a given situation. I first introduce the notion of “inner aspect” and the different types of events that can be distinguished according to it. I will bring the proofs onto IL predicates. according to a set of tests. the criteria to distinguish them. I also emphasized that this should not be taken to mean that IL predicates are stative and permanent predicates. I then present a survey of the distinct types of predicates traditionally distinguished in the literature. also called “aktionsart” or “situation aspect” (Smith 1991).Chapter 3 Event Classes and Individual-Level Predicates In chapter 2.1. According to classical and well-established aspectual tests. Differing from most previous literature. 3. I give a few introductory notes about inner aspect and its relevance in the description of the properties of sentences. 1993). the extent up to which the dichotomy between states and activities is real and grammatically relevant. I will examine the temporal properties of these predicates. and the tests to diagnose them. I defend a finer grained distinction among IL predicates based on the different inner-aspect properties they show. as has been widely assumed across the literature. or whether there is more than one (IL) copular verb. or delimitation. I present one of them in section 3. I will demonstrate that copular adjectival constructions can be classified in two groups.1 Inner Aspect “Inner aspect” (Verkuyl 1989. such as where the stem of the contrast resides (in the copular verb or in the adjective itself). Such properties tell us whether the event inherently tends . as activities). are relegated to the next chapter. culmination. One group tests out as states. 3..

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

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

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

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

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

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

(13) In + x time a. Pablo viajó durante tres semanas Pablo traveled for three weeks c. *Pablo viajó en tres semanas Pablo traveled in three weeks c. They express for how long someone has been engaged in a particular activity or for how long a particular state has held of an individual. *Pablo estuvo enfermo en tres semanas Pablo was ill in three weeks b. *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. as (13) shows. the entailment of a delimit 5 Note that this test does not distinguish between IL and SL predicates. . in-adverbials refer to the period of time it takes to complete a certain event. 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. see Piñón 1999.46 Individuals in Time ending point. which is the reason why they are compatible just with eventualities entailing a delimit point. For a recent account about the different readings available in such sentences. As the ungrammaticality of (14) shows. Pablo estuvo enfermo durante tres semanas Pablo was ill for three weeks b. Pablo construyó una casa en tres semanas Pablo built a house in three weeks d. *Pablo construyó una casa durante tres semanas 6 Pablo built a house for three weeks d.5 (12) For + x time a. 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.

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

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

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

get differentiated from activities and accomplishments.50 Individuals in Time The possible interpretations of almost put in contrast accomplishments with other event types. an achievement. the tests in (22)–(25) can be used to diagnose stativity. 1970). To conclude this survey of aspectual tests. 3.1. ¡Pasea! Take a walk! d.2 gives a more precise notion of “agent. (22) Occurrence in command imperative form a. yielding ungrammatical results. Pablo casi construyó la casa Pablo almost built the house b. an activity. Pablo casi caminaba con ocho meses Pablo almost walked when he was eight months c. Pablo casi se dio cuenta de que su madre tenía razón Pablo almost realized that his mother was right Sentence (21a).2. To the extent that an agent is conceivable only if a proper event (a nonstate) is at hand. (21) Interpretative entailments from the adverb almost a. can have two interpretations. which are grammatical. an accomplishment. can only have the second interpretation. the subject did not get to walk or realize. states and achievements. too. ¡Construye la casa! Build the house! c. *¡Estate enfermo! Be sick! b. as van Voorst (1988) and others point out (see also section 3. However. As can be appreciated. In the other. the subject did not even start out the process of building.5 Agentivity Tests.5 in this regard). I want to introduce another group detecting the presence of an agentive subject (Ryle 1949. ??¡Date cuenta de que estás confundido! Realize that you are wrong! . Lakoff 1966. and (21c). in (22)–(25). (21b). (Section 3. the subject was involved in the process of building up a house but did not finish it. that is. In one of them.”) The roles that DPs play in the event and the analysis of the event itself usually overlap.

Lo que hizo fue pasear What he did was walking d. *Pablo forzó a Juan a estar enfermo Pablo forced John to be sick b. I also introduced the idea that inner aspect does not depend on the inner properties of the verb itself. which pattern together because of the stative feature they share. *Lo que hizo fue estar enfermo What he did was be sick b. I presented different tests usually employed to diagnose which event type a concrete predicate belongs to. where other elements like the object play a determining role. Lo que hizo fue construir la casa What he did was build the house c.1. but it is a compositional matter. Pablo forzó a Juan a pasear Pablo forced John to walk d. Pablo paseó deliberadamente Pablo walked deliberately d. Table 3. the good behavior of activities and accomplishments in agency contexts is due to the property of dynamicity the two of them involve. This accounts for the ungrammaticality of both states and achievements. Pablo construyó la casa deliberadamente Pablo built the house deliberately c. In the second part. Pablo forzó a Juan a construir la casa Pablo forced John to build the house c. *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 3.2.Event Classes and Individual-Level Predicates 51 (23) Pseudocleft with hacer ‘do’ a.2 summarizes the tests and the behavior of event types. *Pablo estuvo enfermo deliberadamente Pablo was sick deliberately b. *Pablo se dio cuenta de que estaba confundido deliberadamente Pablo realized that he was wrong deliberately (25) As a complement of force a.3 Summary of Section 3. A “+” . 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.1 The first part of this section presented the notion of inner aspect.

the results of the tests (9)–(11). and a “%” if the event type shows a positive behavior in the test but just under a certain interpretation. Tests for event types Among other things. as pointed out above in the description of each test.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. and (22)–(25) reveal that activities and accomplishments share an important part of their . (20). The bracketed numbers in the leftmost column refers to the number of the test as they appear in the text above. a “–” when it cannot. (17). States – – – + + / – + – / Activities + + + + + + + % + – Accomplishments + + + – + – – + – + + Achievements % + + – + – / – % – – (9) Progressive form (10) Complement of ‘happen’ (11)Interpretation as habitual in present tense (12) For + x time (13) In x time (14) and (15) Assertion of noncompletion (16) Perfect entailment from the progressive (17) Complement of parar de (18) Complement of dejar de (20) Complement of ‘finish’ (21) Ambiguous interpretation with the adverb ‘almost’ (22) Command imperative form (23) Cleft with ‘do’ (24) Combination with ‘deliberately’-like adverbs (25) Complement of ‘force’ – – – + + + + + + – – – – + + – Table 3.2. The symbol “/” means that the test does not apply to the event type for some reason.

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

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

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

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

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

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

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

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

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

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

1.Event Classes and Individual-Level Predicates 63 remarks). its acceptability in (46b) is expected. as the contrasts between (46c) and (47c) show. and cruel. 17 Much like do-pseudoclefts (see footnote 10). in-adverbials refer to the period of time it takes to complete a certain event.18 states and activities from accomplishments and achievements. . but the oddness in (46a) calls for further explanation. It does so as well with IL predicates. It is interesting to note. as shown earlier.3. we see that one group comprises APs showing a state nature (a. Since it appears that the AP in (c) behaves like an event. If.b)17 and the other comprises one kind of AP manifesting eventive properties (c). In turn. As I introduced in chapter 2 (and will discuss further in chapter 6). (45c) gets a habitual interpretation. rather than a state. predicates like Eskimo refer to properties that. If we compare these results with the ones obtained in the previous section. Even in the absence of the habitual adverbial. which is why they are compatible only with eventualities entailing a delimit point. por eso sabe hablar Inuit “What happens is that her cousin is Eskimo. on the other. although it gives ungrammaticality for one of the predicates that above tested out as a state (Eskimo). pseudoclefts with happen are acceptable with states when present tense is involved. 3. for-adverbials express how long someone has been engaged in a particular activity or a state has held of an individual. once again. states are durative and for-adverbials are durative adverbials. on the one hand.2.2. 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). that is why she can speak Inuit” 18 As mentioned in section 3. the reasons for its ungrammaticality are independent from the test itself. Consider a scenario like the following: (i) Lo que sucede/pasa es que su primo es esquimal. hold for that person’s entire lifetime. The test in (46) distinguishes. as reasoned in the literature. a clear contrast emerges among different types of APs in combination with the copula ser. The main point of (46) and (47) is to show that the AP manifesting eventive-like properties (cruel) pattern with activities rather than with accomplishments or achievements. which is why the limitation of its duration with a for + x time complement sounds weird. These three tests show that not all APs in combination with ser behave the same way. Rather. Actually. the next step is to find out exactly what kind of event it is. when they hold of an individual.1.2 Activities and States versus Achievements and Accomplishments.

*Juan fue esquimal durante tres semanas Juan was an Eskimo for three weeks b. In principle.3 Activities and Accomplishments versus States and Achievements. since this is only possible with eventive predicates.1.1. 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. tests like (49) divide activities and accomplishments from states and achievements. Juan ha sido cruel con Pedro Juan has been cruel to Pedro 3.3. (48b) is a licit entailment from (48a). *Juan fue rubio en dos horas Juan was blond in two hours c.64 Individuals in Time (46) For + x time a. only activities. and not accomplishments. a predicate like ser + AP would be unexpected as a complement of parar de ‘stop’. Once again.2. where there is no endpoint. yield a perfect entailment from the progressive form. (48) Perfect entailment from the progressive form a. *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). However. which argues in favor of considering the predicate (ser cruel) as an activity.3. As shown in section 3. *Juan fue esquimal en dos horas Juan was an Eskimo in two hours b.2. the combination of ser and an AP such as cruel ‘cruel’ looks good. Juan fue rubio durante tres años Juan was blond for three years c. such a predicate shows a nonstate patterning.2. Juan está siendo cruel con Pedro Juan is being cruel to Pedro b. . As mentioned in section 3.1.

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

*Juan terminó de ser esquimal. Juan finished being an Eskimo b. Tests (52) and (53) show some more contrasts consistent with the results from (49). Only ser cruel is possible under a command imperative form (54) or in combination with adverbs like deliberately (55). *Juan casi fue esquimal Juan almost was an Eskimo 3. it contrasts with other APs with ser. except for cruel. and (b) that the subject did not even start out the process.1. ?Juan terminó de ser cruel con su adversario (en el minuto 10 del debate) Juan finished being cruel to his opponent (in the 10th minute of the debate) The following sentences with almost. whose combination with almost is impossible.1. the latter is with activities.1. 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. 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. like (53c).2.4. Although be cruel predicates do not give perfect results after finish.4 Accomplishments versus All the Rest. which suggests that dynamic properties are somehow involved in these predicates. Juan no empezó a ser cruel con su adversario Juan did not start being cruel to his opponent c.20 Although (53a) does not sound extremely natural. *Juan terminó de ser rubio Juan finished being blond c. 20 Recall that. All tests give the same results. confirm be cruel as an activity.66 Individuals in Time 3. as mentioned in section 3. .5 Agentivity Tests. None of the AP predicates with ser. they clearly contrast with the other two. fit in canonical agentive contexts. since the unique licit entailment from (53a) is (53b). The former reading is available with accomplishments. (53) Interpretative entailments from the adverb almost a.3.3. which also test accomplishments versus all the other eventualities. ?Juan casi fue cruel con su adversario Juan almost was cruel to his opponent b.

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

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

Since part of this work deals with the opposition between states and activities. as a consequence. Pablo viajó durante tres semanas Pablo traveled for three weeks c. and another group that cannot.1 Similarities between Activities and States Section 3. push a cart.g.Event Classes and Individual-Level Predicates 69 lifetime properties. states and activities share properties that make them pattern together. belong) lack an inherent endpoint. I devote more attention to their interpretation in chapter 6. I consider the aspectual difference observed here (those between states and activities) from a broader point of view. *Pablo construyó una casa durante tres semanas Pablo built a house for three weeks d. and. as shown in the aspectual tests. be sick. I consider their similarities and differences in this section. This fact has led a large number of logicians and semanticists to claim that activities and states constitute a natural class. both are atelic).1 pointed out some similarities between activities and states. Such a property makes them pattern together in contexts involving for-adverbials and in-adverbials. For convenience. Both activities (e. swim.. They do not advance toward a culminating point that delimits the temporal contour of the eventuality (i. Pablo estuvo enfermo durante tres semanas Pablo was ill for three weeks b.. *Pablo se dio cuenta de que su madre tenía razón durante tres semanas Pablo realized that his mother was right for three weeks . In the next section. They are distinguishable in contexts that make an endpoint explicit (51).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. know.e.g. there is no real distinction between them. (62) For + x time a.. (62) and (63) reproduce these tests. 3. walk) and states (e. 3.4. As noticed earlier.

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

consider (72). However. much is a quantifier proper for mass nouns. meaning that ‘Pablo drawing a circle takes place very often’. For more detailed discussion of frequency readings. 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. with activities. Likewise.2.23 In this respect. building a house is not the result from summing portions. (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. Mourelatos (1978). a table is not the result of minor portions of tables. . Quine (1960). Carlson (1981). see section 5. 24 I thank Tim Stowell for bringing these cases to my attention. For example. the subdivision and the sum of the parts of a count noun and the result are not of the same nature. “water” can be divided into parts. (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). There are other state predicates that also allow for such a frequency reading. (71)). or it can also mean that Pablo goes walking very often. or subintervals. noticed the close parallel between the mass-count distinction in the nominal domain and the aspectual classification of predicates. which yields bad results with count nouns (cf. In (67) the adverbial mucho refers to the intensity with which Pablo loves Mary. and Bach (1986). and the sum of portions of water is always water. As is known. each of which is water. among others. Pablo is able to walk 45 miles without stopping). if we sum subintervals of walking or being sick. it can mean that the amount of walking is big (e. it is ambiguous: in (68).. However. The legs of a table are not a table. of building a house.Event Classes and Individual-Level Predicates 71 For example. However. we always obtain walking or being sick as a result. Observing such properties.g.

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

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

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

26 I have stopped knowing her/getting to know her I have nothing to say. the different interpretation of predicates with modal verbs diagnoses agency versus lack of agency. epistemic and deontic. the oddness of (xiv)) or an inchoative one (xv). rather than stativity versus dynamicity. Roughly described. Martha. The same can be said for accomplishments (80) and achievements (81).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. (xiv) *This table is missing a leg more and more (xv) *This table is getting to miss a leg 26 On my view. (xii) This table is missing a leg (xiii) *This table misses a leg This pair cannot be explained either by alluding to a scalar interpretation (cf. Piñón (1995) notes that modal verbs have different interpretations with dynamic and stative predicates. however. is usually involved in. about examples from English such as those pointed out by van Voorst (1988). On the former. 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. the speaker makes a guess about Martha’s state. On the latter the speaker expresses a command that Martha has to effectuate. with an activity such as walk around the park. the modal has two meanings— namely. however. the modal has just an epistemic reading. (78) Marta debe estar enferma Martha must be sick (79) Marta debe pasear todos los días por el parque Martha must walk every day around the park (80) Marta debe trazar un círculo ahora mismo Martha must draw a circle right now (81) Marta debe llegar al final de la carrera Marta must arrive at the end of the race Note that with the stative verb be sick. the speaker expresses a guess about a particular activity she thinks the subject. However.

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

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

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

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

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

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

it gets excluded from the dyadic usage (16). of an individual and an event simultaneously. age. mean. The speaker would be adding something unnecessary.86 Individuals in Time Quantity and Manner. as in (11). Stowell (1991) suggests that the active behavior observed in certain copular clauses is due to an implicit event argument. this proposal leaves unexplained why it is a concrete type of adjectives which activates the set of dynamic properties. for the preference of one reading over the other. or dyadic. intelligent) can be predicated of an individual. of an event (12). I will mention four points. the adjective must be able to be predicated of individuals. 4.1. as the account of the two copular entries. (11) John was cruel (12) To punish the dog was cruel So. color. as in (9) and (10). shape. and. With respect to the dyadic usage. (13) *To invest in the bonus market was wide/green/squared/old (14) John was wide/old to invest in the bonus market Second.3 Adjectives Predicated of an Implicit Event Argument. (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. Finally. when the adjectives are predicated of two arguments at the same time (an individual and an event).) can appear in dyadic sentences. Thus. First. etc. to be able to participate in the so-called dyadic constructions. the adjective must be able to be predicated of events. at best. no adjective referring to physical properties (dimension. optionally. If the adjective cannot be predicated of an individual (15). Stowell studies the set of adjectival predicates. Rothstein’s Gricean explanation does not account for the fact that (7) is ambiguous but just. whose behavior I described in chapter 3 as proper to activities in English constructions such as those in (9) and (10). kind. in cases such as (11) and (12). and therefore creating confusion to the hearer. MP adjectives can be monadic. when the adjectives are predicated of just one argument. (15) *John was premeditated (16) *John was premeditated to invest in the bonus market .

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

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

as (26) illustrates. 1995).4 he takes it that the presence of such an argument is proper of SL predicates. I take up the discussion related to the interpretation of the “infinitival argument” in Spanish and its status (as argument or adjunct). this proposal would entail that in all instances of APs followed by a relational PP. necessarily. he argues that the action denoting argument is an overt version of the (Davidsonian) eventive argument. and the contrast between the examples with ser and those with estar (27) can be captured by paraphrases such as the one in (28). estar. in chapter 6 (section 6. I repeat one of the examples here in (29). According to Stowell. to admit there is an eventive argument that is implicit amounts to admitting that (25) is a SL predication. the event argument represented by the infinitive is implicit in simpler copular cases such as the ones under investigation here. . the copular verb should be. (28)). Following Kratzer (1988. First. 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. in his account. 1995) account.Aspectual Alternations in IL Predicates 89 Stowell (1991) establishes a correspondence between the monadic version of these adjectives and individual-level (IL) predication. In Spanish. that is not case. see chapter 2 (section 2.1.1) regarding the APs that can appear with either copula (cf.2. which is the lexical marking for SL-hood.2). 4 For details of Kratzer’s (1988.3). I will make two observations. Ser is completely grammatical. and between the dyadic one and stage-level (SL) predication. (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.2. However. but he is not cruel to him usually” Later. (25) John was cruel to Peter Regarding Stowell’s analysis. Specifically. Note that the contrast between ser and estar in this case falls in line with the contrasts studied in chapter 2 (section 2.

we observed that there were certain copular combinations that behave like states and others that behave like activities.3).2 Adjectival Predicates Showing an Activity-like Behavior In chapter 3 (section 3. the event argument Stowell proposes corresponds to an accomplishment (21) or activity (22). these nonstative copular clauses pattern with activities and not with accomplishments. since they are not the same. I will consider these sentences as instances of IL predication. More specific- . I have argued in the previous section that the activity-like behavior should not be attributed to properties of the copular verb. I will then offer an explanation regarding the stative reading also available with such predicates (cf. as a consequence. which is proved by their acceptation of during-adverbials and the rejection of in-adverbials. to account for nonstative IL copular clauses. As shown in the previous chapter. I will first elaborate an explanation as to what specific kind of adjectives are the ones that show an activity-agentive patterning and why. 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 purpose in the remainder of the chapter is twofold. Thus. my proposal will not be based on properties of the copular verb. 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. (8) above). as shown above. since it is not the copula that varies in the minimal pairs that can be construed (be Eskimo/be cruel) and. 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. it cannot be said it is the copula that induces the active features. the properties of copular cases such as (25) (the active-agentive characteristics) cannot be attributed to or derived from those of the implicit event argument. but he looks very handsome in that suit” Given the grammaticality of ser. Since the activity-like behavior depends on the adjectival predicate combining with the copula. 4. Second.90 Individuals in Time guapo. These two facts lead me to discard this hypothesis based on the existence of implicit event arguments.

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

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

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

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

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

One can. 1994] and . 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. set on fire and bother. 1988. even if we finally concluded that the PP is an “affected argument”.” This captures his intuition that such a DP refers to the individual that gets affected by the (underlying) action undertaken by the subject DP. I will show that its syntactic presence correlates with the aspectual patterning of the construction. 4.3. focusing on two aspects: its interpretation and its optionality or obligatoriness.1 On the Interpretation of the Relational PP Stowell (1991) suggests that the DP inside the relational PP. offend or regale. 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). offend someone else either by saying something unpleasant. it is understood that Juan was the agent of an action that had Pedro as its affected argument. Thus. and such an action is qualified as cruel.3 Relational Mental Properties: The Relational PP Complement As just described. In this section. 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”).96 Individuals in Time correlation with their aspectual characteristics. and discuss all the consequences derivable from this. in italics in (58). (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. to name just a few. I investigate the nature of the PP complement. for example. abuse. can be considered as an “affected goal. or by acting in a certain way. although it seems intuitively clear that the PP specifies the goal addressed by the subject’s action. Likewise. 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. harass. 4. whether or not the PP expresses an “affected argument” depends on the concrete type of action undertaken by the subject. Incidentally.

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

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

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

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

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

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

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

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

it is not inferred from the context or interpreted as generic. but just a “theme. I consider these cases with inanimate subjects as evidence that the PP complement cannot be an obligatory complement of cruel.” if we use traditional vocabulary. which enables agency. In the next sections. I will consider that the presence of the PP correlates with a subject involving particular properties and.3. which are not totally excluded. although these adjectives referring to human (or animate) aptitudes can appear in combination with a relational PP. conclude that the structure of cruel-type APs does not always involve a PP complement and aspectually behave as states in principle. If the PP complement is not overt. (103) Juan es inteligente Juan is intelligent . Consider (101) and (102). with particular characteristics of the construction. can also take a relational PP complement. while the subject of cruel is not an agent. they do not have the same relationship with the PP. 4.3 The Relational PP with Other APs There are other MPs. This is because they describe a property that can be understood in relation to another animate entity. namely. (104) is not the interpretation of (103).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. Summarizing. as mentioned before. kind. as it is possible with the other adjectives (cruel. since. as Violeta Demonte makes me notice.). etc. I therefore. not included in the group of relational MPs in the table above which. otherwise. In the first place. dynamicity. its impossibility to appear with certain subjects would remain unaccounted for. also. I account for the relation between the two inner aspect forms. (101) ?Juan fue estúpido con el entrevistador (y no le contestó a ninguna pregunta) “Juan was very stupid to the interviewer (and did not answer any question)” (102) ?Juan fue muy inteligente con su jefe (y consiguió lo que quería) “Juan was very intelligent to his boss (and got what he wanted)” I will make two brief considerations in this respect. since it needs to bear the required properties for that (and animacy is the most basic one).

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

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

(107) El delantero avanzó hasta la portería The foremost advanced until the goal (108) a.2) the contrast between animate and inanimate DPs in pairs like (109) and (110). the thematic interpretation of the subject and particular aspectual properties. we would be forced to assume that each aspectual behavior of adjectival copular cases (stative and nonstative) corresponds to different (but homophonous) lexical items.1 Lexicalist and Logico-Semantic Approaches The observed correlations among the presence of the PP relational complement.5 Justifying the Approach 4. In the remainder of the chapter I will be concerned with such a contrast in the realm of adjectival copular clauses. syntactic presence of a PP) as correlations since. The pair in (108) from Borer (2005) shows an alternation stative/dynamic which correlates with the [±animate] properties of the DP subject. two different entries (stative and dynamic) for each adjective (cruel. kind. by definition. From a lexical point of view.5.4). both atelic eventualities. Along the lines of Borer (2005) (among others). In (107) it is a PP which is exerting a crucial aspectual role in making the event an accomplishment. . 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.).” Recall in this regard from chapter 3 (section 3. whose contrast has been debated in the literature (see section 3. since there are syntactic and semantic facts suggesting its actual substantiation. María tocó la verja (dos veces) María touched the fence (twice) The alternation here is of states versus activities. it does not allude to the other elements present in the syntactic configuration. etc. 4. according to which aspect properties are fully encoded in the lexical items. I argued that such a contrast is grammatically relevant. That is. La pared tocaba la verja The wall touched the fence b. I begin by discussing which approach to event representation can make sense of the observed correlations among the syntactic presence of a constituent.108 Individuals in Time tioned thus far. mean. This perspective cannot capture any of the other properties (interpretation of the subject. However. 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.

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

1989. the notions of origin and termination give the characterization of events (113)–(115) and the temporal points identifying the beginning and the end correspond to physical entities (the arguments): that is. Tenny (1987. there is no obvious way to distinguish between the two constructions in point (stative and dynamic copular clauses). (113) •---------------------------------------object of origin/actualization (114) •----------------------------------------• object of origin object of termination (115) ----------------------------------------• object of termination Activities Accomplishments Achievements (116) Event Correspondence Rule (Van Voorst 1988) Subject NP Direct object NP •----------------------------------------• event-object of event-object of termination origin/actualization . As van Voorst puts it. both would contain the term “hold. As discussed earlier (see section 3. 4.2. In particular. these authors emphasized the role of the internal object as an event delimiter.1 Event Roles.” Second.” given the fact that both are atelic and do not “culminate. the dynamic or stative properties. since. 1994) and van Voorst (1988) elaborate both on the observation that arguments grammaticize the points that give the temporal contour of an event. this approach does not establish any relationship between the aspectual term (Hold). among others. in principle. 1994). and the presence or absence of other elements in the sentence (the PP complement in our cases) that.110 Individuals in Time This approach leaves several questions unanswered.5. authors such as Verkuyl (1972).1. each entity corresponds to an aspectual notion. First. which opens the possibility of establishing a direct relationship between the thematic properties of the arguments present in a sentence. and van Voorst (1988).5. as captured in the Event Correspondence Rule (116). 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. arguably.2 Syntactic Approaches 4. Tenny (1987. correlate with the aspectual outfits of the construction. showed the relevance of arguments in the inner aspect properties of a sentence. its syntactic position and its role in the determination of the aspectual properties of the sentence. Dowty (1979).1). 1989.

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

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

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

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

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

accomplishments. and the aspectual properties of the construction. but it is the syntactic structure which gives the type of aspectual interpretation. provides the skeleton where lexical items merge and their interpretation is obtained as a by product. differing from Borer.116 Individuals in Time Regarding this specialized stative structure. “default.5. 2000). I assume that eventualities are not specified from the lexicon as states.3.3 Summary of Section 4. differing from Borer’s idea. the conclusion I draw.” and dynamic properties come induced from separate projections. . achievements or activities. I assume that it is syntax itself which. I assume that the interpretation of arguments emerges as an entailment from the aspectual structure. Likewise. in particular. which has been proved to be essential in explaining aspect properties. I also assume that heterogeneous and homogeneous eventualities are distinguished by a concrete syntactic projection (Quantity). I will discuss the claim that adjectival predicates can only be stative. their interpretation. 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. this complement is not obligatory. My discussion of inner aspect properties of IL copular sentences in Spanish takes Borer’s (2005) work as a frame of reference. 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. The dynamic properties exhibited by the adjectival constructions studied here (those of the cruel-type) directly challenge this hypothesis. by creating the structure based on functional nodes. and. that states are the type by default. by the intervention of the PP complement that some adjectives can have: John was cruel to Pedro. Regarding the main aspectual issue I am concerned with (the state/activity contrast observed in IL copular clauses). in the spirit of Borer (2005) and Ritter and Rosen (1996. More in the spirit of the seminal work by Hale and Keyser (1993).5 In this section I introduced the approach I will take for the analysis of the aspectual properties of the IL copular constructions. 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. is that the stative status is. I have suggested. Since. and its absence correlates with the stativity of the construction. Specifically. Such a perspective is in itself an account of compositionality. in some sense. 4. Focusing on adjectival copular cases. as argued in section 4.

Stowell (1993). at least. the PP makes them a dynamic predicate. Let us begin by discussing point (b) concerning the aspectual contribution of the preposition. I will address two theoretical points. I will argue that the stative version of the cruel-type APs corresponds to their most “basic” structure.6. I will take such a correlation seriously and I will propose that the source of the aspectual properties (specifically. . To capture (a)—that is. As I advanced. and I will also be concerned with the aspectual treatment of the arguments expressed via a PP. the construction has characteristics proper of processes. the DP subject has to involve a particular set of properties. since their intervention in the event structure is not obvious (given that they are oblique arguments). all of them very likely emerging from animacy itself: the DP has to be able to behave as a controller agent. the following two properties: (a) When there is a relational PP. 2004). As I develop the proposal. From the description in the previous sections. that prepositions can be conceived of as heads encoding aspect properties. I propose that the dynamic properties become a part of the construction through the preposition heading the PP. the aspectual alternation (state/activity) is not unpredictable at all. I will discuss the syntactic treatment of the opposition state/activity. 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. it has become clear that the account for the cruel-type constructions has to capture. 2000. based on Hale (1984). I propose that the source of dynamicity and its related properties (agency etc. I argue. 4.Aspectual Alternations in IL Predicates 117 4. which gains dynamic properties from the preposition heading the PP denoting the DP goal. as soon as they combine with a (relational) PP. (b) When an animate DP subject and the PP are present. As I intimated before. Point (a) is discussed in section 4. In support of this proposal. the dynamic aspectual behavior) actually resides in the preposition itself. behave as activities. among others. To capture (b).1 Prepositions as Aspect Encoders I have shown that cruel-type APs.3. That is.6. but it correlates with the existence of the PP relational complement.) in the active cruel-type constructions is a projection (the PP) that combines with the “basic” structure of the AP. and Demirdache and UribeEtxebarría (1997.6 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.

I argued that DPs like the one in italics in (134) should be analyzed as “goal/destination-objects. the presence of a complement and particular aspectual properties are both due to the preposition. That is. the directional enclitic of central coincidence -yi. Davis. In other words. I offer a formal account for these predicates. “affected-goals. Hale (1984) notices a strong correlation between the spatial and the temporal system in Warlpiri (spoken in the central western part of Northern Australia). as an activity. (135) *Juan was very kind Pedro (136) Juan regaled Pedro The preposition is needed to make the (affected) goal DP available. expressing a close meaning. One realm where this can be noted is its determiner system. but they have to be inside a PP. For example. As will be shown. spatial oppositions appear replicated in the set of enclitic elements constituting the aspectual system. In the absence of specific temporal marking. 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. prepositions are seen as carriers of aspect.” which can be. 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. Demirdache 1997. (Matthewson 1996. based on the distinction of visibility/proximity relative to the speaker versus 20 . he explains. temporal interpretation is derived from the meaning of locatives. In the previous section. and also. I will motivate the preposition as a predicate conveying aspect information.118 Individuals in Time I will first consider the role of the preposition to make the (affected) goalcomplement syntactically available. among others). I will then study the relation between prepositions and aspect. In what follows. as the oddity of (136) shows. language spoken in the southwest of British Columbia. The mutual resemblances between prepositions and temporal content have been noted by many authors. Finally. has the meaning of ‘durative’.20 Spatial marking is used to express time relations also in Lillooet Salish. it introduces an important aspectual change: whenever the preposition is present. whose complements can be added directly. This makes adjectives differ from verbs. the predicate unambiguously behaves as a dynamic predicate. in preparation. morphologically belong to the spatial clitic system. potentially. in Warlpiri. Consider (136). He notices that.

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

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

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

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

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

as (157) summarizes. the only difference between con ‘with’ and a ‘to’ is the nature of the DP acting as their complements. or hurt them… [Alfonso X : Siete partidas. contra (‘against’). Consider first the following examples taken from the corpus of Spanish from the Illinois State University (Davies 1999). . S. he was very cruel against whom did not deserve it [Alfonso X: General Estoria V. in previous centuries. the prepositions introducing the relational complement in cruel-type cases were of same characteristics as con. Svenonius 2004). XIII] 25 I would like to thank Isabel Pérez for bringing this source into my knowledge. XIII] (159) (…)& guardandosse segunt su aluedrio fue muy cruel contra los que lo non merecían & keeping himself according to his will. Most were locative predicates that can also have a directional meaning. the complement of adjectives of the type of cruel appears introduced by prepositions such as a (‘to’). in previous periods of Spanish. S.124 Individuals in Time (155) ¡Corre a la farmacia antes de que cierren! Run to the pharmacy before they close! (156) Volví a la tienda a devolver los zapatos I went back to the store to return the shoes Thus. con (‘with’) and a (‘to’) can be said to share properties and can be both analyzed as prepositions denoting a trajectory with a destination. It is interesting to note that. S. (157) Con + [+animate] DPs A + [–animate] DPs Therefore. 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. with a goal.25 As reported in the corpus. 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. 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. 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.

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

The two of them are related by a preposition meaning ‘toward’. The two terms are put under a specific relationship according to the meaning of the preposition. 2004) proposed that centripetal noncentral prepositions translate as “before” in the temporal domain. 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). we saw that Demirdache and Uribe-Etxebarría (1997. the DP subject can be considered as the figure and the DP inside the PP as the ground. (182) summarizes the description of the preposition.) (182) (183) TOWARD: preposition that establishes a centripetal noncentral coincidence relation between the figure and the ground. the discontinuous dotted line for the trajectory. therefore. They are. it designates the movement of a figure toward a place or ground.1.6. and (183) represents it graphically. 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 . (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. 2000. associated with ‘future’ in the realm of Tense and with ‘prospective’ in the domain of outer aspect. it locates the figure in its trajectory toward the ground. 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. that is. …● ⌂ In section 4.

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

This fact suggests two things. and we could not explain why it cannot appear in sentences like (186). when the DP subject is inanimate.Aspectual Alternations in IL Predicates 133 construction is decided.2. As shown in section 4. when the subject is inanimate. when adjectives such as cruel are predicated of an inanimate subject. 4. the PP complement would be possible independently from the properties of the DP subject. I repeat the contrasts below. (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. eventuality properties are obtained from the very syntactic structure.3 The DP Subject of Relational MPs As I mentioned in the beginning of the section. as a complement of verbs like force or in combination with volition adverbials: (190) Occurrence in command imperative Juan/*imagen. the relational PP complement cannot appear. And second. 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. the construction cannot appear in contexts proper of dynamic eventualities. the construction cannot appear in agency scenarios either. be cruel!” (191) As a complement of force El director forzó a Juan/*la imagen a ser cruel “The director forced Juan/the image to be cruel” . If that were the case. 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. it suggests that the PP is not a plain complement of the adjective.6.3. First. In sum. such as the progressive or as a complement of parar de ‘stop’. ¡sé cruel! “Juan/image. it suggests that the properties of the noun in the DP subject and the PP stand in a codependent relationship. such as the imperative form.

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

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

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

red) and refer to dynamic processes. the paraphrases of both being ‘to’ (The Latin preposition in could be followed by a nominal either in ablative case. The prepositions participating in the cases above are en (en-negreció) and a (a-largó). Both cases are similar regarding the elements that participate in them: a preposition. If followed by a nominal in ablative case. 4. Etymologically. contra Borer’s (2005) idea that statives are due to the presence of specialized functional structure (a sort of a “stativizer”). The structure I have proposed for deadjectival verbs exploit the participation of a preposition to account for their dynamic properties. or in accusative case. a `centripetal noncentral coincidence preposition. That is. they come from the Latin “in” and “ad”. 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. an adjective and a verbal piece. since it is by virtue of a preposition that an adjectival construction becomes dynamic. I propose that it is by virtue of the combination of the adjective with the temporal content contributed by the preposition that the structure gains dynamic features: (iv) PP 2 P la tela 2 P A en negro This would explain the aspectual properties of deadjectival verbs. it had a directional meaning (‘to’).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. black. whereas if followed by an accusative. (v) en–negre–c–(er) ? ? ? ? P A v (infinitive-ending) ? ? ≈? con cruel ser . which renders parallel deadjectival verbs and cruel-type cases. I suggest that such dynamic properties come from the semantics conveyed by the prepositions that participate in their formation.Aspectual Alternations in IL Predicates 137 structure to emerge. the preposition had a locative (stative) meaning (English ‘in’).

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

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

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

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

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

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

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

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


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

and inclusion are semantically defined as partial order relations fulfilling a number of properties. posteriority. Therefore. I argued that inner aspectual properties could also be conceived in ordering terms.1 Tense. For an introduction of such notions. like to. and Eventuality Time). which I mentioned briefly in chapter 4 (section 4. Since the line of this work is not formal semantics I do not discuss them here. as Tense and Aspect are considered ordering predicates. I argued that dynamicity properties (that activities involve) can be described through prepositions. aspect properties can serve as indicators of temporal properties (a perfect action is computed as finished and. 3 Anteriority. for example) and. the interested reader is referred to Partee. I show that.3 Authors such as Zagona (1990) and Stowell (1993. However. atelic). as something that took place in the past). inner aspect can be as well.g. I turn my attention to adjectival IL predicates and show the different interpretations available with each type of viewpoint mentioned.. I tried to show that all aspectual properties could be reduced to the same set of primitives. in Spanish at least. 5.2 One of the facts I alluded to was that. by the same token.1.1). These predicates translate into the temporal domain as ‘before’. therefore. this being why the eventuality is computed as “dynamic and unbounded” (i. in some languages (e. Reichenbach gave a formalization of tenses in natural languages based on three intervals (Reference Time. At that point.. specifically. It turns out that this cannot be described in terms of presence or absence of quantity. based on Zagona 1990. posteriority.148 Individuals in Time In section 5. Lillooet Salish). and Wall 1993. I show that states are compatible even with the perfective.6.5. The idea is depicted in (1). and simultaneity). 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 1947.1 Tense and Aspect as Ordering Predicates In this section I deal in greater detail with the conception of Tense and Aspect as ordering predicates. Stowell (1993.e. but in dynamicity versus lack of dynamicity. that establish centripetal noncentral coincidence relations between a figure and a ground. inner-aspect characterization may work as an aspect indicator (an accomplishment can be taken as a perfect action. 5. Consistent with the conclusion that viewpoints do not interfere with quantity properties. In particular. and the possible relationships that can be established among them (anteriority. the progressive seems to impose restrictions on the type of eventuality. Speech Time. 2 . ter Meulen. 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.

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

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

yielding the reading according to which Maddi’s leaving takes place at 5. I examine this situation in chapter 6. the value of the external ZP is the UT. Modifiers become the evidence for modified arguments.) (iii) To the extent that time adverbials can be considered modifiers of time-denoting arguments. .M. if the adverbial modifies the TT (iii). In sum. For example. the adverb restricts the reference of the ET. These authors argue that time adverbials can be considered as modifiers of temporal arguments. Demirdache and Uribe-Etxebarría (2004) provide empirical evidence supporting their actual representation in the syntax. nothing else hinges on this. In turn.M. the grammatical existence of such ZPs is demonstrated..e. ZP(ET) 2 ET PP (at 5 P. with no additional stipulation.6 I will adopt this conception of Tense 6 As for ZPs.) ZP(TT) 2 TT PP (at 5 P. for the moment. 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. prior to 5). temporal relations are accounted for by arguing that Tense is a dyadic predicate that takes two temporal arguments (ZPs) plus general syntactic rules. In (ii).Outer Aspect and Individual-Level Predicates 151 (7) TP 2 ZP T′ (RT=UT) 2 T ZP (after) 2 Zi VP 2 ei VP 2 Mary V′ 2 take the book Because these sentences are matrix sentences. The different readings thus emerge depending on which temporal argument they modify. However.M. (i) (ii) Maddi had left school at 5 P. the two interpretations of (i) are derived from the different ZP modified in each case. where I explore the temporal interpretation of IL predicates in simple and compound sentences. 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 principle.152 Individuals in Time as an ordering predicate. It orders the time the speaker talks about with respect to a time of reference. The observation that Tense can refer to a concrete interval traces back to Partee (1973. According to (9). 1996) work. 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). The lens of a camera. The former would correspond with readings of the sort in (10). Now. I assume that the element that does this job is Aspect. but the interval the speaker is referring to. Depending on the focus . That time is in the past. That is.1. Following Klein (1994). 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. 1984b) and was also noted in Stowell’s (1993. when someone utters (8). Aspect as an Ordering Predicate The temporal interpretation of a sentence like (8) can. In (8). Between the two. (10) wants to capture a different assertion: the speaker has a concrete time in his mind and he is asserting that. as Klein (1994) pointed out: Tense does not order the ET with respect to RT. 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. However. 5. whereas the latter with those along the lines of (9). be paraphrased as (9) or (10). The analogy between DPs and ZPs enables Stowell to distinguish between “definite” and “indefinite” (with just pure existential force) ZPs. I call the interval the sentence is about (or “the interval the speaker talks about”) “Topic Time” (TT). he intends (10) rather than (9). (10) seems the most accurate paraphrase of (8). María was involved in the task of washing the car. makes them visible for the receiver to be able to catch them in a photograph. the total interval during which María has been washing the car is left in the dark. for example. This leads to a big conclusion. by focusing objects. Smith (1991) makes an insightful comparison between aspect and the lens of a camera. (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. (10) There is a specific time at which Maria was involved in the task of washing the car.2 The Internal Argument of Tense: The External Argument of Aspect. at that time.

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

that of ordering temporal arguments.3SG Trabajó he-work-PERF-PRET. The verb estar acts as an auxiliary and can appear in any tense and aspect form. Aspect nomenclature (García 1999) As indicated earlier.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.1. That is. 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. Returning to the internal working of Aspect. the next question is how the singled-out portion of the situation (TT) relates to the rest of the event. whereas with the perfect. 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. the adverbial refers to the time at which the event takes place. the event has already taken place at the time denoted by the adverbial.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. Tense and Aspect consist in the same mechanism—namely. I restrict my attention to the forms called “imperfect” and “perfective” (and concretely to the perfective simple form) as well as with the progressive. (i) Perfective → the leaving occurs at 3 (ii) Perfect → the leaving occurs prior to 3 . The progressive is construed by the verb estar (locative be) plus the ing-form of the verb (Spanish -ndo). I mention them here only as clarifications. Thus. I follow Klein’s (1994) proposal that there is an ordering relationship between the TT and the ET. In the glosses for my Spanish examples. the difference between the composed forms of “have” corresponding to the perfective and the perfect is that.3SG Trabaja-ba he-work-IMPF-PRET. is an ordering predicate. 9 As many authors have pointed out. with the perfective. as Demirdache and Uribe-Etxebarría (2000) put it. which is the element in charge of establishing a relationship between these two intervals. More strictly speaking. Aspect.

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

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

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

In this latter regard. Let me start by considering sentences (22)–(24). I will consider two points—first. and second. (25) ----------------------x---------------------(26) ------x------x--------x-------x---------x-In the spirit of Verkuyl (1999).158 Individuals in Time of Quantity. the functional projection which makes the eventuality heterogeneous and whose absence leaves it homogeneous. what the relationship is between the number of occasions alluded to by each viewpoint and the quantity nature of predicates (that is. as shown in (25). a quantificational component. the three forms differ in the number of occasions they allude to. besides the ordering component. who observed that the difference between the progressive and the habitual rests on the number of occasions they refer to.3. the imperfect in (24) refers to a plural number of occasions that can be described as greater than one (|>1|). Q<occ> can also have an existential value. as the reader may have conjectured already. I propose that aspectual viewpoints involve. besides ordering. the habitual imperfect.5. which I analyze as complex phrases containing a quantifier over occasions and ordering information. giving rise to another interpretation. However. The continuous imperfect does not count the number of occasions a determined predicate takes place. here in section 5. can be either |1| or |>1|. It appears . (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. However. also contain information about the number of occasions an eventuality occurs.2 and next in section 5.2. I address these points in turn.1 Quantifying over Occasions As just mentioned. The values of such a quantifier. whose head is a quantifier over occasions (Q<occ>). As I will repeat later in section 5.1. how many occasions each viewpoint alludes to. as in (26). inner aspect. the nonhabitual or “continuous” imperfect. 5. The number of occasions the progressive and the perfective form allude to is just |1|. In this section I am chiefly concerned with the perfective. and the progressive. aspectual viewpoints.

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

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

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

whereas the imperfect habitual. the point I want to make with (27) is that. which turns the event (make the cake) into a state. in and of itself. there is no need to interpret that the paper has been read completely or the sonata has been played entirely. as Tim Stowell (p. 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. Once accomplishments are states. Compare (i) to (ii). it is interpreted as a progressive (ii). 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). refers to a plural number of occasions. it is interesting to consider accomplishment predicates in perfective followed by a “for + time” adverbial. the sentence is. 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). In (i). Consider (38).) points out. (iii) Leyó el periódico durante dos horas he-read-PERF-PRET-3SG the paper for two hours . simply.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. For example. 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. where this interpretation is much more difficult to get. as she conceives habituals to be. nor an account for these cases. and. Although I will discuss neither the process of coercion itself. I would like to note that in the iterative interpretation with the perfective and for-adverbials.162 Individuals in Time Regarding the number of occasions that the different viewpoints can allude to. that cake is not physically the same cake year after year. the iterative interpretation seems possible when the DP object can be understood as different instances of the same object description. as a consequence. However. where.16 Likewise. this does not mean that the number of occasions that an eventuality in the perfective takes place cannot be modified by an adverbial. their compatibility with for-adverbials gets explained. (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.c.

(40) Juan fuma Juan smokes (41) Juan va a Nueva York Juan goes to New York Suppose we take a year as a reference unit. 5.2 The Habitual Interpretation: Iteration. (43) Juan lives in Madrid and goes to New York four times a year (iv) Tocó la sonata durante una hora he-play-PERF-PRET-3SG the sonata for an hour 17 Girona International Summer School of Linguistics class notes.Outer Aspect and Individual-Level Predicates 163 progressive. In this subsection I deal with the properties of one such quantifier that gives rise to habitual interpretation. Specifically. the exact number of event instances is not specified.2. I will consider that habituality is encoded in the “quantificational floor” of Aspect. Pablo estaba llamando a la puerta tres veces In the two occasions I heard him. Consider the contrast between (40) and (41) as a starting point. Now suppose we interpret (40) and (41) against the following scenarios in (42) and (43). I consider that habituality corresponds to a quantifier over occasions denoting a plurality of occasions of a particular eventuality. . However. and propose that it is expressed through functional quantificational structure. Pablo was knocking at the door three times. whereas the progressive and the perfective (in the absence of adverbial modification) refer to a singular occasion.2. 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. (42) Juan smokes four times a year.1 Iteration. The speaker’s assumptions about the actual number of instances of a habitually quantified event depends on heterogeneous (pragmatic) factors. 5. and Systematicity Verkuyl (1999) and Stowell (2000).17 among others. as described in (27c). As just mentioned. define habituality as an iteration of instances of a given eventuality distributed in time. I assume that the meaning of habituality is lexically expressed in quantifier adverbials such as habitually or usually. Proportion. Following the quantificational-based perspective defended by Verkuyl.2. the habitual viewpoint refers to a plural number of occasions by itself.

2. In other words. If five is more than the average ratio of students that get an A. and (b) a certain proportion with respect to the number of times the action at stake is usually performed. (41) is very likely to be judged as appropriate. As Westerståhl (1984) and Partee. among others.” In turn. (40) does not reach the category of “habitual. Sentence (44) would be judged inappropriate for such a scenario. In particular. ter Meulen. Suppose further that the number of students who got an A is five. I propose that habitual Q<occ> behaves similarly to canonical quantifiers such as many. Whereas (40) is very likely to be judged as inappropriate. therefore. and Wall (1993). it seems that. as such. The following pair is from Partee. ter Meulen. (44) should be judged as a sentence that properly describes the situation.164 Individuals in Time Although the implicated number of times is the same (four). given the situations of (42) and (43). some notion of “average” regarding the number of occasions that an action is performed is taken into account. If the average ratio of right-handed people is (statistically) higher than that. what counts as “many” depends on different contextual parameters. (40) and (41) are judged differently. that in the determination of “habituality” there are. when judging whether a habitual form is appropriate. two intervening factors: (a) iteration. since the eventualities take place more than once. . is affected by contextual parameters (Westerståhl 1984). (44) Many students got an A (45) Many students are right-handed Suppose that the set under consideration is a class of 20 students. (45) does not properly describe the situation. 5.2 Proportion. It seems. Thus. and Wall. at least. a habitual sentence is considered appropriate for depicting such a scenario. (41) is judged as appropriate for the situation in (43). Specifically. 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 number of times people usually smoke tobacco in our culture highly exceeds four times a year.” which is established by external information. we conclude that the crucial notion in describing many resides in the “average ratio. I will review the behavior of this quantifier in the nominal realm briefly. (46) expresses this formally. The number of times considered “average” is established taking into account extralinguistic information. How does such pragmatic information get integrated into the aspectual encoding? I propose that aspect Q<occ> behaves like regular quantifiers and. Suppose now that the number of right-handed people in the class is also five. note. whose definition heavily depends on contextual basis.2.

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

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). Consider (49) and (50) as samples of other possible cases. 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. despite the fact that. in principle. the number of occasions that people smoke (in general). .1. rather. the contextual parameter captures the average ratio relevant in each case. the semantic intuition about examples like (47). In any event. then. As I mentioned in section 5. Juan fumaba. we can paraphrase John smokes by using ‘John is a smoker’. As shown. The cited contextual parameter would capture. contextual information participates in establishing one of the members of the proportion the quantifier refers to. Juan used to smoke. to my understanding. which give us quantities of individuals. in this case. to the number of occasions that Juan smoked marijuana at the parties with place. at the parties of the company was marijuana Juan smoke-IMPF-PRET-3SG “At the company parties there used to be marijuana around. it would not be appropriate to use (47) to describe a situation where John has smoked only four times in a year.2. rather than discussing the semantics of the present or imperfect preterit. but.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. 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. this does not fully capture. 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. aspectual quantifiers give us the number of instances of a particular event. in a simple case like (47). In principle. As before with many. (49) En las fiestas de la empresa había marihuana. However.” (50) En verano los niños jugaban al tenis en el jardín. my main point in this section is to describe habituality.2. this is not always the case.

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

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

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

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

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

he has had that car at every single moment between 1974 and 1985. However.176 Individuals in Time 5. I argue that a perfective form does not trigger telicity effects on predicates. who conceive the perfective as a modifier that makes an eventuality telic. As discussed back in chapter 3 (section 3. 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.23 Observe these examples: (91) Pablo nadó Pablo swim-PRET-PERF-3SG “Pablo swam” (92) Pablo caminó Pablo walk-PRET-PERF-3SG “Pablo walked” Verbs such as swim and walk test out as homogeneous predicates. In other words. unlike states. In contrast. The meaning of these sentences is that Pablo was once involved in the task of swimming (91) and walking (92).2 Perfective Homogeneous Predications I turn my attention now to homogeneous predicates appearing in perfective form. at least in Spanish. sentences such as these show that the presence of the perfective viewpoint does not contribute telicity. With the perfective. If we say John walked from 2 to 3. such as Bertinetto (2000). Bach (1986). which leads to the inference of “completeness” and “boundedness. draw a conclusion in the same direction. it does not mean that he has been walking at every single moment of that interval. 24 23 . de Swart (1998).3.24 as subinterval and additivity proofs show. (95) *Pablo nadó en una hora Pablo swim-PRET-PERF-3SG in an hour “Pablo swam in an hour” Other authors. I will argue that.” as mentioned in the previous section. a perfective form does not interfere with the quantity nature of the predicate. and others. a perfective form does not make a homogeneous predicate heterogeneous. (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. Contrary to Mourelatos (1978). In the same vein as before. the TT is located ‘after’ the ET. if we say John had that car from 1974 to 1985. activities are not 100 percent homogeneous at a micro-level.4). Piñón (1995).

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

When the quantifier denotes a number roughly describable as “greater than one. 26 This aspect form is involved with stative SL predicates (She was sick). Specifically.25 The existential quantifier corresponds to the continuous imperfect. |>1|. but these predicates are also compatible with the habitual quantifier (She was sick usually). I assume to be present with every type of predicate (SL and IL). following Stowell (1993. and not only with SL ones. this is borne out. ∃) should be possible with any type of predicate. I argued that. the ordering component takes the quantified eventuality and orders it with respect to the TT interval. 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). In the last section. As will be shown in section 5. (98) ZP 2 Zi VP 2 ei VP (99) AspP 2 TT Asp′ 2 Asp0 ZP (after/before/in) 2 VP Z0 (|1|. presented in (3) above (repeated here as (98)) about the internal constituency of ZPs. Elaborating on Stowell’s (1993) proposal. which does not make the sentence habitual.5.1 for further discussion. ∃) 2 e VP Thus.26 The quantifier over occasions binds the eventive variable (e) which. 25 . 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). and makes the prediction that the quantifiers (|1|. 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). in Spanish.” perfective or progressive. where there are no occasions to enumerate (think of stative homogeneous predicates such as Peter is tall). |>1|. if “one. This unifies the working of Aspect. See section 5.4. independently from the kind of predicate. 1996).” habitual imperfect is at stake. as Kratzer (1988) originally proposed.

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

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

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

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

That is. 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. the habitual reading is the most salient. 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).Outer Aspect and Individual-Level Predicates 183 (una persona) muy cruel. as in (113). When the PP complement is present. Habitual AspP 2 TT Asp′ 2 Asp Q<occ> (within) 2 >1 b. por eso lo detestaban (111) Juan era Juan ser-PRET-IMPF-3SG a person very cruel for this him hated sus hermanas his sisters “Juan was a very cruel person. 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” . that is why his sisters hated him” Contexts (110) and (111) represent scenarios where the property is attributed to the person as a whole. Nonhabitual (continuous) AspP 2 TT Asp′ 2 Asp Q<occ> (within) 2 ∃ Consider now the viewpoint interpretations when the PP complement is present. when the complement inducing the activity-like properties is entered into the structure. in a similar way as we say John was blond or John was Eskimo.

(119) and (120)).30 Finally. (122)). although the progressive form is correct with cruel. make the habitual reading available. they acquire activity-like properties.184 Individuals in Time It is worth noticing that. 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.3.2. 30 For the reasons why the progressive is excluded with Eskimo and blond. However.5. 29 . see section 5.1. a habitual interpretation emerges. whereas activities in imperfect (117) can also have a progressive paraphrase (118). accordingly.29 (117) Pablo leía mientras María paseaba Pablo read-IMPF-PRET-3SG while Maria walk-IMPF-PRET-3SG (118) Pablo was reading while Maria was walking (119) Pablo era cruel con Pedro mientras este era simpático con él (durante la entrevista) Pablo ser-IMPF-PRET-3SG cruel to Pedro while he ser-IMPF-PRET-3SG nice to him (during the interview) (120) ??Pablo estaba siendo cruel con Pedro mientras este estaba siendo simpático con él (durante la entrevista) Pablo was being cruel to Pedro while he was being nice to him (during the interview) That is. Likewise. in Spanish. as noted above. rather than the habitual (cf. as noted in section 5. the progressive is not a paraphrase saliently available from the imperfect form. which. this is just marginally available with be cruel (cf. the preferred interpretation of adjectives denoting intellectual aptitudes (121) is the continuous reading. when a relational complement is added (123). when a complement that can act as a “distributee” is added (125). the progressive paraphrase of the imperfect is more natural in these contexts.

nonstative IL (128).Outer Aspect and Individual-Level Predicates 185 ingenioso en sus bromas/inteligente (125) Juan era Juan ser-IMPF-PRET-3Sg cunning at his jokes/ intelligent en los negocios at his business (126) “Juan used to be cunning at his jokes/intelligent at his business” Summarizing. the typo) are subject to referential variability for habitual interpretation to be possible. or achievements (132). and any eventive predicate. the objects (the house. accomplishments (131). 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. . chino habitualmente (127) En sus reencarnaciones. activities (130). Juan era in his reincarnations Juan ser-IMPF-PRET-3SG Chinese habitually “In his reincarnations. stative SL (129). the habitual viewpoint can be said to be compatible with any kind of predicate—stative IL (127).

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

31 As mentioned in chapter 3. consistent with the description of the perfective above. the progressive is possible with other class of homogeneous predicates. and therefore reject the expression of progression in time. according to which it is compatible with basically any type of predicate. in the sense that the progressive is not compatible with every kind of predicate. states hold in time but do not take time. the progressive in English works differently in some contexts.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. it is also possible to have the stative version of these adjectives together with the perfective. whereas processes take time and progress in time (although this progression in time does not mean they advance toward a culminating point). . 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. (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. SL (141) and IL (142) statives are excluded with the progressive. As can be seen. 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).5.3 The Progressive and Adjectival IL Predicates I now turn my attention to the progressive viewpoint in Spanish.31 The progressive viewpoint does not seem to work like the imperfect and the perfective.

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

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

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

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

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

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

194 Individuals in Time 6. the representation in (1). IL predicates IP 1 I′ 1 I VP 1 θ-subj Regarding temporal interpretation. First. see Anderson 1973. In particular. 1995) Kratzer (1988. 3 As I also said in chapter 2. 1995) is the first in establishing an explicit correlation between the temporal interpretation later dubbed “lifetime effects” and IL predicates. just aims to capture Kratzer’s idea on the location of the external argument with respect to the maximal projection of the predicate. This author assumes that the factor that draws the line between the two types of predicates is “being temporary” versus “being permanent. I will mention two points of this proposal. . as Diesing 1992 proposes. a spatiotemporal or eventive (<e>) argument (Davidson 1967). Kratzer proposes that the spatiotemporal argument is syntactically represented. SL predicates involve an extra semantic operation. That is.2). The configurational difference between IL and SL predicates can be depicted roughly as in (1).3 (1) a. temporariness and permanency is encoded in the argument structure of predicates: temporary predicates involve a spatiotemporal argument. Kratzer argues that SL predicates involve an extra argument in their argument structure—namely. Furthermore. For Carlson (1977). Kratzer proposes that the differentiation between SL and IL predicates resides in their argument structure.1). she argues that the eventive argument is the external argument of (SL) predicates.1.1. the “realization function.” Likewise. 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.1 I will briefly summarize her line of reasoning and show that there are cases it cannot account for. As said before in chapter 2 (section 2.1 Temporal Interpretation as a Consequence of Argument Structure: Kratzer (1988. whereas permanent ones lack such an argument. SL predicates IP 1 I′ 1 I VP 1 <e> b.”2 Specifically.” See chapter 2 (section 2. whereas SL predicates needed something “extra. in the same vein as Carlson (1977). she proposed that IL predicates applied directly to the subject.

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

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

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


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

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

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

since there is a large number of predicates that. literally. do not have to be permanent properties.4 Summary of Section 6. 6. 1997). whether they denote a lifetime property. I argued that those accounts. 1995). Musan (1995. such as the presence of another past tense around. in their lexical entry. necessarily produce an inaccurate overgeneralization. (b) the predicate has to appear in the past form. compare the following sentences: .3. Second. 6.. First.Tense and Individual-Level Predicates 205 6.). 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.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. In support of the first claim. In the following section. Ph. 1997) observed that contextual factors. I do not consider that those predicates encode. can neutralize the lifetime effects. As already mentioned. and (c) the past form should be in the imperfect form. I am going to argue. I focus on another restriction for this temporal interpretation. More accurately.g. That is.1 Necessary Conditions for Lifetime Effects First. e. in the line of Musan (1995. 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. nevertheless. 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. equating IL-hood to property stability is inadequate. Differing from Kratzer (1988.2. that lifetime effects are subject to certain contextual conditions. for lexical reasons. as.2 I have made two main points in this section. I argue that lifetime effects are a salient option in those cases where the predicate. according to independent grounds (such as their combination with a particular lexical choice of the copula in languages like Spanish). according to which the temporal interpretation known as lifetime effects derives from the argument structure of IL predicates. these three conditions have to be met: (a) the predicate has to be an IL lifetime predicate. I want to briefly enumerate the conditions that are necessary (not sufficient) for the lifetime reading to appear. can be argued to be IL and.D. Musan (1995) does. for example. For the interpretation of the DP subject as ‘no longer alive’ to be an option (in languages like Spanish).

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

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

208 Individuals in Time other verbs appearing in lifetime predications. as any other operator. ‘the day I was introduced to Gregory and Eva-Lotta’. Westerståhl (1984) and Lewis (1986) observed that domains of quantification are contextually restricted. 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 can be restricted by the context through a variable C. . In this section I show that even when the predicate denotes a lifetime property. Both examples here are from Musan (1995). is exactly the same. the temporal operator of Gregory was from America in the second case is restricted by the previous context. “out of the blue” cases. Musan argues as follows. Gregory was from America and Eva-Lotta was from Switzerland. Although I am abstracting away from the mechanics of the semantic formalization. To account for this fact. and others with a previous context. The third condition is that the viewpoint form must be the imperfect. roughly. In the framework she assumes.3. Precisely. Kratzer’s (1995) account predicts a lifetime reading to be available with IL predicates in all circumstances. This author proposed to consider in contrast sentences like (52). I will show a number of situations where such a reading does not arise. 1997). like (53). (52) Gregory was from America (53) On that day.2 Introducing the Determining Role of Contextual Factors. despite the fact that the sentence. such as have brown eyes or have long legs. the lifetime reading does not necessarily come out.4. The second condition is that the tense form has to be the past. 8 As mentioned in the previous chapter. tense is taken as a sentence operator. This was first noted by Musan (1995. 6. Musan’s idea can be informally depicted as in (54). I was introduced to Gregory and Eva-Lotta. as Kratzer’s examples were. The sentence Gregory was from America is given a specific temporal context in which it is interpreted temporally. it is not in (53). I would like to emphasize that these are not sufficient conditions for the lifetime reading to arise. Gregory was from America. in the next sections. Then. I return to this point in section 6.1. whereas a lifetime reading is available in (52). As Musan observes. since it was a consequence of their argument structure. 1997) As noted earlier.8 Thus. Musan (1995.

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

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

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

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

I define “discourse topic” as the subject of immediate concern in the conversation. The common ground determines the options that are live at any point in the conversation.” Following widespread proposals in semantics and pragmatics.1 When the Subject Is a QDP I am going to examine a set of examples without any previous explicit context.3 I will show how my proposal (to account for all the cases uniformly) works in a systematic way. I assume. 6. 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 content of their TTs is different) because their TTs get their values from distinct “discourse topics”. Consider the following sentences.e. 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). Regarding the specific relationship between the discourse topic and the elements influenced by it. I assume that every natural language discourse takes place in a context. following von Fintel (1994). 1979). and von Fintel (1994). Grice (1975). In section 6. lifetime effects do not arise. nevertheless. where. Kratzer (1977. (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 .4.4. that it can be described as an anaphoric relationship. In the next two sections. a “context” contains a “common ground”—that is. a set of propositions that form the shared background of the participants in the conversation.Tense and Individual-Level Predicates 213 interpretation of these two types (with and without context) varies (i. I will show that the TT is influenced by the discourse topic. among many others. 1981). As defined by Stalnaker (1972. Among the elements constituting the discursive common ground. they are anaphoric to distinct “discourse topics. Based on Keenan-Ochs and Schieffelin 1976. I assume that sentences are uttered against this background.

where the reading can be defended to exist as a possibility. in these cases we cannot allude to the presence of a previously mentioned past interval to account for the mitigation of the interpretation.” (69) The boys a. Higginbotham 1983. In (70). Then. Lewis 1986). the set that determiners quantify over is restricted by the context (Westerståhl 1984. The context provides a set that the NP intersects with. 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 meaning of (70) is that of (71). as mentioned in chapter 5.214 Individuals in Time (68) Ningún chico era esquimal No guy was Eskimo In none of these examples is the lifetime reading available. constitutes the quantificational restriction. and von Fintel (1994). D x [ x is N] g restrictive clause Dx [x(N)] Furthermore. quantifiers bind the open argument (variable x). determiners/quantifiers are interpreted with respect to the set contextually provided. and these examples resides in the nature of the subject DPs. As is classically argued (Heim 1982. among others. QP/DP 2 Di NP the 2 xi N′ g boy b. . all together. Stump (1981). but just over the contextually relevant set of plates. Thus. The contextual relevance gets into play through the contextual variable appearing as a subindex of the determiner. However. I am going to argue that the very nature and working of quantifiers is what precludes the arising of the reading. the simplest hypothesis is to suppose that the difference in the availability of the reading resides in the different nature of the DPs. boys in (69). say. Longobardi 1994). Since the difference between (52) (Gregory was from America). in the world. and this. every is not quantifying over the whole set of plates. That set is called the “restrictive clause. In essence. Partee (1984a). propose that such a restriction is formalized through a free variable in the determiner. which the noun is a predicate of (69a). Stowell 1989.

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

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

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

Elaborating on Musan 1995. 4. 3.3 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. Beghelli & Stowell 1996). how this is articulated. the context acting as background constitutes the set where anaphoric elements and free variables find their antecedents.4. A specific interval is an interval that is linked to a previous context. where the individual referred to by João is present.218 Individuals in Time situation. This seems to suffice to convert the utterance situation as the background that counts as relevant for the antecedent of the TT. more systematically. 5. 2. as Musan (1995) suggests. and. the utterance situation is the most salient context. following von Fintel (1994). That is. (77) Salient context " TT . The context relevant to find the “antecedent” for the TT does not have to be another past interval necessarily. The occurrence of lifetime effects depends on the content of TT of the sentence. 6. I have argued that this is brought about through the context variable restricting quantifiers (free variable C). lifetime effects do not arise when the interval relevant in the sentence is other than the lifetime span of the individual referred to by the DP subject. In other words. Conversely. The leading ideas introduced thus far are the following: 1. along the same lines that specific DPs are generally understood (Pesetsky 1987. The TT refers to a specific interval. The relationship between the contextual background and the TT or the free variable C has been taken as anaphoric. which contains contextual information associated to the individual referred to by the DP. Enç 1991a. so that all the surveyed cases can be accounted for in a uniform way. The relevant context can be brought to the scenario by the DP subject itself. I want to show now. I also argued that proper name DPs involve a similar context variable. it becomes the source for antecedents. 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. 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. as a consequence.

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

Since the DP subjects of copular sentences can be considered “topical elements. 1995) and Musan (1995) assume. the value it takes by default is the individual himself and the interval associated to him. Once a context is fixed. Since the contextual variable directs us to a previous context containing a past interval. (82) Gregory was from America (83) Topic of the sentence that may establish the connections to a previous background = Gregory (84) Gregory c= undefined. (80) Harry was from California. 6.4 In determining the context relevant for the specification of the TT content (which is what finally decides the temporal interpretation of a sentence). since they are the same in all relevant respects—all of them are copular sentences in context. there is no reason to consider (59) different from (72) and (73).4 Summary of Section 6.4.220 Individuals in Time crucial importance in the licensing and interpretation of a past form. I have argued that the establishment of the relevant context is brought about through the free variable DPs (more specifically. the ZP . Result: lifetime arises With Musan (1995. (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. it finds its antecedents in the most salient context.” they establish the relevant context. no lifetime effect arises. The interpretive procedure is always the same and works the following way. As Kratzer (1988. I assume that. a lifetime effect does not arise. since it refers to the time of the arrival. by default. Bearing in mind that the TT is a ZP context sensitive. let us consider a sentence where a lifetime reading can arise. the determiners) have. these typically are out of the blue sentences. it is the information associated to the DP subject that counts. when there is no specific context the free variable is to be linked to. Actually. I propose. Finally. (…) Topic of the sentence that may establish the connections to a previous background = Harry. 1997). Result: a past form is allowed in the copular clause and. Gregory himself TT → the whole interval overlapping Gregory’s lifetime span. then. that in (59) it is the contextual information associated to the DP Harry that establishes the context that counts.

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

M. (Thus far. I focus now on the interpretation of lifetime IL predicates in another type of syntactic environment. if the RT of the subordinate clause is in the c-command domain of the ET. 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. and its eventuality will be located with respect to the eventuality of the main clause. in compound sentences. subordinate clauses (complement and relative). In this work. yielding different interpretations. According to him. I have just dealt with IL sentences which establish a link with a previous one paratactically. TT and ET can be modified by time adverbs.1) I described the way temporal interpretation is accounted for following Stowell’s (1993. by default. Following the logic of Stowell’s proposal discussed above. yielding the interpretation in (iii). Thus.1. giving rise to different temporal readings. the ordering between the TTs should depend on the specification of the embedded RT. whose nature I argue to be also a ZP. as in (i). the prototypical case of this is the modification of the past perfect. instead of with respect to the TT. . Specifically. namely.5 The Lifetime Reading in Compound Sentences For the sake of a more complete survey of cases. given that the ET is the closest ZP.15 However.M.222 Individuals in Time 6. this does not mean that ETs always remain in the dark for interpretation.14 but rather the TT.) I will deal with complement clauses and relatives in turns. As mentioned in chapter 5.1 Complement Clauses In chapter 5 (section 5. since the RT would be controlled by the main ET rather than by the main TT. The time adverb at 3 can modify the TT.M. the content of the RT is. so that it can be controlled by the ET of the main clause. the embedded TT would become ordered with respect to the main ET. this could lead us to a situation where the TTs are not ordered between each other strictly speaking. the subordinate clause will be temporally dependent on the main one. Peter’s leaving occurs at 3 P. (i) (ii) (iii) 15 14 Peter had left at 3 P. yielding the reading in (ii) or the ET. temporal interpretation is derived from the ordering relationship Tense establishes between the RT and the ET. As Hornstein (1990) and Demirdache and Uribe-Etxebarría (2004). as the subindexes of the ZPs in (i) suggest. have noted. However. starting by the former ones. among others. the UT. 6. This way. the subordinate RT can be affected by the ET of the main clause. 1996) work. I argued that what Tense locates is not the ET as such. In simple sentences. Peter’s leaving occurs prior to 3 P. to understand the temporal interpretation of a compound sentence is to understand the relationship holding between the two (or more) TTs.5.

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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


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

Specifically. the aspectual characteristics of the copular constructions are derived from their syntax. such as blond or young. as is the case with any other type of predicate in Spanish (e. Ritter & Rosen 2000.. there are two types of IL predicates regarding inner aspect: stative and dynamic predicates. gypsy. among others). this study is a contribution to a better understanding of the role of prepositions in the properties of event structure. The book proposes a uniform treatment of tense. Demirdache & Uribe-Etxebarría 2000). The investigation of Lifetime Effects (the interpretive phenomenon whereby the DP subject can be understood as ‘no longer alive’) in chapter 6 deepens our understanding of the concept of Topic Time. differing from previous influential proposals like Kratzer 1988 where. the temporal interpretation of IL predicates is derived. in Spanish. unlike what is commonly assumed. dynamicity is proposed to correlate with the presence of the relational complement and argued to be rooted in the Prepositional Phrase. which clearly pattern with activities in an inner-aspect classification. I have argued that the dynamic nature of copular constructions does not correlate with an SL classification. in the case of IL predicates. this book gives further empirical support to recent hypotheses according to which aspectual properties are syntactically encoded (Borer 2005. Finally. Thus. Crucially also. kind. Topic Time has been argued to be a time interval taking its value from the salient discursive context. contrary to general belief. IL predicates are not all stative. based on hypotheses that attribute aspectual content to prepositions (Hale 1984.240 Individuals in Time individual. As a result. tense takes and orders the nominal argument (the NP subject). Rather. ser and estar. the dynamic properties of clauses where adjectives referring to mental properties appear (such as cruel. I have concluded that those proposals arguing that ser-predicates should be considered SL under aspectual forms like the perfective and the progressive cannot be correct. no stative verb can appear in the progressive). from the content of the so-called Topic Time (Klein 1994). I have shown that ser-predicates can appear in any aspectual form. or color-blind. but it is independent from that: cruel mean/kind to someone can appear with both copulas. I have accounted for a variety of contextual situations illustrating the sources from . Third. In this respect. In particular. Ramchand 2003. Others.g. Second. are properties that can be true just of a limited period of time. mean) have been proposed to correlate with the presence of the relational complement of these adjectives (cruel to Mary). such as Eskimo. The restrictions between a ser-predicate and aspectual forms like the perfective or progressive are derived from the inneraspect properties of the predicate. however. I have proposed that the dynamic and active properties classically observed in certain copular constructions are due to the presence of a complement. as is the case with any other type of predicate.

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

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

the argument that Tense takes. In other words. and in (5) the copular ser-clause appears with another number of adverbial complements restricting the interval the property holds. (3) Juan dejó de ser rubio/guapo/dulce/generoso/altruista/ egoísta/atrevido/miedica/valiente/criticón/retorcido/sensible/ soberbio/envidioso/pesado/servicial cuando se hizo mayor Juan stopped ser-ing blond/handsome/sweet/generous/altruist/egotistical/daring/fearful/brave/faultfinding/twisted/sensitive/arrogant/ envious/tedious/helpful when he grew up Cuando era pequeño. in correlation with the wa/ga marking of subjects in Japanese. which can be tested by the nonvariation of the copular verb in Spanish. the distinction between IL/SL cannot be described in terms of permanent versus episodic.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 interval a certain property lasts seems independent from its nature as SL or IL. see Torii 2000.1 Another important respect in which the proposal defended here differs from earlier accounts (particularly Kratzer 1988. at least in temporal terms regarding the length of the interval. in direct relation to this. 1995) concerns the structure of IL and SL predicates and. 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}. permanency cannot be a defining characteristic of IL predicates. In (4). all the following cases (discussed in chapters 2 and 6) become unexplained. Since the length of the span of time properties hold does not define IL predicates. which is unexpected if IL predicates are forced to apply for the entire lifetime span of an individual. the IL/SL dichotomy is not rooted in temporal terms. since the interval IL predicates hold over can be restricted. an adverbial clause restricts the period the property holds. 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}. the copular clause appears as a complement of dejar de ‘stop’. but then he changed his attitude (4) (5) In conclusion. In (3). .

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

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

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

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

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

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

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

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

Strictly speaking. This explains their interpretation as IL when they act as modifiers inside a DP. I consider that adjectival predicates are conceived to be IL “by default. We have similar consequences with the copula ser. as soon as we add a degree quantifier such as muy ‘very’. since. That is to say.hood has been proposed to derive from the properties of a lexical head. When these properties are not met. The cases I have in mind are those like (23) or (24).g.” in the sense that in the absence of anything else in the structure. In this respect.252 Individuals in Time Escandell-Vidal and Leonetti (2002) consider that estar selects for a small clause (SC) with certain aspectual properties. whereas SL.9 However. they will be interpreted that way. inside them. I do not consider it necessary to talk about a pragmatic reinterpretation process when the verb estar enters into the picture. . 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). inner aspect). there is nothing producing the linking to a circumstance. The main difference between inner-aspect properties and SL ones. As suggested previously. in a similar vein as I have conceived other grammatical properties in this work (e. strictly compositional. then. Juan usually takes the Americans’ side” 9 This view of SL-hood is.. is that the former are built up out of a functional projection (Quantity). as I have sketched here. 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. Juan is taking the Americans’ side” muy demócratico (24) Juan está Juan estar-PRES-3SG very democratic “Juan is behaving very democratically” As the glosses show. Nevertheless. coercion converts the SC containing the adjective from IL to SL in order to meet the copula requirements. muy americano (25) Juan es Juan ser-PRES-3SG very American “Juan usually behaves in the way Americans do” “In his opinions. the sense of the adjective has undergone a slight change. coercion occurs and converts an IL SC (lacking the necessary aspectual properties) into an SL SC. these cases differ from (1) and (2). In principle. muy americano (últimamente) (23) Juan está Juan estar-PRES-3SG very American (lately) “Juan is behaving in the way Americans usually do” “In his opinions.

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

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

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

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

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

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

• The heterogeneous cases of estar have a meaning different than the one registered in the minimal pairs ser/estar.3. hinging upon the type of the predicate following the copula and the telic inducing adverbial in + x time. stative and dynamic Can appear in any outer aspect form The property is referred to the interval specified in the Topic Time Estar Situates individuals in a determined circumstance • Homogeneous constructions (stative and dynamic) and heterogeneous ones. Can appear in any outer aspect form The property is referred to the interval specified in the Topic Time Table 7.Conclusions and Final Remarks 259 Ser Classifies individuals Homogeneous constructions. Differences between ser and estar .


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

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

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

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

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

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

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

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

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

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

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

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

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


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

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

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


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

175–177. 158–160. 226. 226. 209. 222. 156 Hawaii-Creole 120 Hebrew 15 Heterogeneous (predicates) 72. 242. 176. 184. 211. 165. 129. initiator 61. 195. 35. 147. 123. 178. 189. 221 Habitual 77. 72. 229. 189. 11. 211. 134 Instigator 55. 174–178. 248. 232. 233. 234 Situation aspect 39 Small v 55 Spatiotemporal argument/variable 8–10. 156 Focus 13. 230. 134–136. 137 Subinterval property 70. 231. 181. 189. 249 Relative clauses temporal interpretation of. 122. 236 Theme 105. 192. 144. 226. 178. 187. 252 Realization function 6. 242. see also Utterance Time Stativizer 115. 168 French 121 Generic operator 11. 148. 202 Ground 119. 76. 112. 131. 123. 208. 183. 117. 134–136. 244. 116. 80. 129 Progressive 153–163. 249 Existential coda 12 Existential quantification 8. 148. 105. 258 Imperfect Continuous 76. 227. 221–223. 100. 158. 6. 136. 171– 175. 72. 173. 26. 158. 192 Prospective 120–122. 245. 79. 186. 158. 113. 172. 229. 246–248. 189– 192. 136 Place/situational prepositions 120. 242. 148–152. 191. 157. 94 Pomo 94 Prepositions Directional prepositions 127. 155–157. 158. 189. 207. 148. 12. 132. 131. 35 Figure 119. 12. 154. 31 Imperfective paradox see Progressive paradox Initiation. 222–225. 246.280 Individuals in Time Event-object 110. 165. 35. 144. 80. 212 Mereological properties 3. 248 Swahili 120 Temporal interpretation of Complement clauses 76. 147. 148–152. see Temporal interpretation Past-shifted reading 76–78. 11. 144. 153–160. 178. 192. 175. 148 Sequence of Tense 226. 14. 128–131. 94 Margi 121 Maxim of Quantity 85. 187. 132. 246. 240. 154–157. 129. 86. 55. 258. 258 Progressive paradox 191. 134 There-sentences 5. 176–178. 112 Juba Arabic 120 Kinesis 40 Krio 120 Lakhota 55. 116. 162. 159. 161. 215 Goal(Path) 96–98. 191 Event(uality) Time (ET) 3. 159 Quantification restriction 11 Quantity 113–116. 192 (Im)perfective contrast 25. 244. 172. 14. 233–235. 194. 125. 64. 248. 241. 148. 160. 180 Perfective (preterit) 31. 147. 186. see Temporal interpretation Russian 15 Salish 118. 130. 148. 37 Thetic (judgments) 12–14. 186. 172–174. 249. 240. 246. 177–180. 158–160. 94. 142. 250 . 131. 173–180. 88. 227 Simultaneous reading 76–79. 181 Existential reading 6. 257. 122. 155. 83. 123– 126. see also Subinterval property and Additivity property Narration temporal interpretation of. 136. 189. 224. 72. 115. 186. 118. 141. 128. 227. 183. 194 Reference Time (RT) 3. 258 Hindi 120 Homogeneous (predicates) 70. 257 Perform(er) 53. 152. 135. 227–229. 141. 212. 167. 150. 184. 186. 224. 176. 58. 184. 14 Generic reading 6. 111. see also Event(ive) argument Speech Time 122. 67. 230. 147. 230 of Narration 77–80 of Relative clauses 223. 187–192. 57. 196. 162. 233 Perfect entailment 47. 179. 201. 212. 207. 180. 132. 106. 79.

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

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